Atoll 3.3.0 Technical Reference Guide PDF

Version 3.3.
Technical Reference Guide

for Radio Networks
AT330_TRR_E1
AT330_TRR_E1
Atoll 3.3.0 Technical Reference Guide for Radio Networks

Release: AT330_TRR_E1 (March 2015)
© Copyright 1997-2015 Forsk. All Rights Reserved.
Published by:
Forsk
7 rue des Briquetiers
31700 Blagnac, France
Tel: +33 562 747 210
Fax: +33 562 747 211
The software described in this document is provided under a licence agreement. The software may only be used or copied under the terms and conditions of the licence agreement.
No part of the contents of this document may be reproduced or transmitted in any form or by any means without written permission from the publisher.
The product or brand names mentioned in this document are trademarks or registered trademarks of their respective registering parties.
The publisher has taken care in the preparation of this document, but makes no expressed or implied warranty of any kind and assumes no responsibility for errors or omissions.
No liability is assumed for incidental or consequential damages in connection with or arising out of the use of the information contained herein.
AT330_TRR_E1 Chapter : Introduction
Introduction
This Technical Reference Guide is aimed at radio network engineers with an advanced knowledge of Atoll and radio network
planning. It provides detailed information about the inner workings and formulas that are implemented by Atoll.
About Atoll
Atoll is a 64-bit multi-technology wireless network design and optimisation platform. Atoll is open, scalable, flexible, and
supports wireless operators throughout the network life cycle, from initial design to densification and optimisation.
Atoll’s integration and automation features help operators smoothly automate planning and optimisation processes through
flexible scripting and SOA-based mechanisms. Atoll supports a wide range of implementation scenarios, from standalone to
enterprise-wide server-based configurations using distributed and multi-threaded computing.
Atoll Microwave is a complete backhaul and microwave link planning solution based on the leading Atoll platform, which
includes a high performance GIS and advanced data and user management features. Atoll Microwave can share its site data-
base with Atoll radio planning and optimisation modules, thus allowing easy data consistency management across the oper-
ator organisation.
If you are interested in learning more about Atoll, please contact your Forsk representative to inquire about our training solu-
tions.
About Forsk
Forsk is an independent company providing radio planning and optimisation software solutions to the wireless industry since
1987.
In 1997, Forsk released the first version of Atoll, its flagship radio planning software. Since then, Atoll has evolved to become
a comprehensive radio planning and optimisation platform and, with more than 5000 installed licenses worldwide, has
reached the leading position on the global market. Atoll combines engineering and automation functions that enable opera-
tors to smoothly and gradually implement SON processes within their organisation.
Today, Forsk is a global supplier with over 300 customers in 100 countries and strategic partnerships with major players in the
industry. Forsk distributes and supports Atoll directly from offices and technical support centres in France, USA, and China as
well as through a worldwide network of distributors and partners.
Since the first release of Atoll, Forsk has been known for its capability to deliver tailored and turn-key radio planning and opti-
misation environments based on Atoll.
To help operators streamline their radio planning and optimisation processes, Forsk provides a complete range of implemen-
tation services, including integration with existing IT infrastructure, automation, as well as data migration, installation, and
training services.
Getting Help
The online help system that is installed with Atoll is designed to give you quick access to the information you need to use the
product effectively. It contains the same material as the Atoll 3.3.0 User Manual.
You can browse the online help from the Contents view, the Index view, or you can use the built-in Search feature.
You can also download manuals from the Forsk web site.
Printing Help Topics

You can print individual topics or chapters from the online help.
To print help topics or chapters:
1. In Atoll, click Help > Help Topics.
2. In the Contents tab, expand the table of contents.
3. Right-click the section or topic that you want to print and click Print. The Print Topics dialog box appears.
4. In the Print Topics dialog box, select what you want to print:
• If you want to print a single topic, select Print the selected topic.
• If you want to print an entire section, including all topics and sections in that section, select Print the selected
heading and all subtopics.
5. Click OK.
3
Chapter : Introduction © 2015 Forsk
About Atoll Documentation

The following PDF manuals are available for Atoll and Atoll Microwave and can be downloaded from the Forsk web site at:
http://www.forsk.com/support.
• Atoll 3.3.0 User Manual
• Atoll 3.3.0 Administrator Manual
• Atoll 3.3.0 Data Structure Reference Guide
• Atoll 3.3.0 Technical Reference Guide
• Atoll 3.3.0 Task Automation Guide
• Atoll 3.3.0 Model Calibration Guide
To read PDF manuals, you can download Adobe Reader from the Adobe web site at:
http://get.adobe.com/reader/
Hardcopy manuals are also available. For more information, contact to your Forsk representative.
Contacting Technical Support

Forsk provides global technical support for its products and services. To contact the Forsk support team, visit the Forsk
Support web site at:
http://www.forsk.com/support.
Alternatively, depending on your geographic location, contact one of the following support teams:
• Forsk Head Office
For regions other than North and Central America and China, contact the Forsk Head Office support team:
• Tel.: +33 562 747 225
• Fax: +33 562 747 211
• Email: support@forsk.com
Opening Hours: Monday to Friday 9.00 am to 6.00 pm (GMT +1:00)
• Forsk US
For North and Central America, contact the Forsk US support team:
• Tel.: 1-888-GO-ATOLL (1-888-462-8655)
• Fax: 1-312-674-4822
• Email: support_us@forsk.com
Opening Hours: Monday to Friday 8.00 am to 8.00 pm (Eastern Standard Time)
• Forsk China
For China, contact the Forsk China support team:
• Tel: +86 20 8557 0016
• Fax: +86 20 8553 8285
• Email: atollsupport@forsk.com.cn
Opening Hours: Monday to Friday 9.00am to 5.30pm (GMT+08:00) Beijing, Chongqing, Hong Kong, Urumqi.
4
AT330_TRR_E1 Table of Contents
Table of Contents
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3
1 Antennas and Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25
1.1 Antenna Attenuation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
1.1.1 Calculation of Azimuth and Tilt Angles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
1.1.2 Antenna Pattern 3D Interpolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
1.1.3 Additional Electrical Downtilt Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
1.2 Antenna Pattern Smoothing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
1.3 Power Received From Secondary Antennas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
1.4 Transmitter Radio Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

1.4.1 GSM Documents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
1.4.2 UMTS, CDMA2000, TD-SCDMA, WiMAX, and LTE Documents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
1.5 Repeaters and Remote Antennas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

1.5.1 UMTS, CDMA2000, TD-SCDMA, WiMAX, and LTE Documents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
1.5.1.1 Signal Level Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
1.5.1.2 Total Gain Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
1.5.1.3 Repeater Noise Figure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
1.5.1.4 Appendix: Carrier Power and Interference Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
1.5.2 GSM Documents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
1.5.2.1 Signal Level Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
1.5.2.2 EIRP Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
1.5.3 Donor-side Parameter Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
1.5.3.1 Azimuth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
1.5.3.2 Mechanical Downtilt. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
1.6 Beamforming Smart Antenna Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

1.6.1 Definitions and Formulas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
1.6.1.1 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
1.6.1.2 Downlink Beamforming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
1.6.1.3 Uplink Beamforming. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
1.6.1.4 Uplink Beamforming and Interference Cancellation (MMSE). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
1.6.2 Downlink Beamforming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
1.6.3 Uplink Beamforming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
1.6.4 Uplink Beamforming and Interference Cancellation (MMSE) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
1.7 Grid-of-Beams Smart Antenna Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
1.8 Adaptive Beam Smart Antenna Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
1.9 Statistical Smart Antenna Gain Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2 Radio Propagation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57

2.1 Path Loss Calculation Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
2.1.1 Ground Altitude Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
2.1.2 Clutter Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
2.1.2.1 Clutter Classes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
2.1.2.2 Clutter Heights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
2.1.3 Geographic Profile Extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
2.1.4 Resolution of the Extracted Profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
2.2 List of Default Propagation Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.3 Okumura-Hata and Cost-Hata Propagation Models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

2.3.1 Hata Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2.3.2 Corrections to the Hata Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5
Table of Contents © Forsk 2015
2.3.3 Calculations in Atoll . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64
2.4 ITU 529-3 Propagation Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64

2.4.1 ITU 529-3 Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64
2.4.2 Corrections to the ITU 529-3 Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64
2.5 Standard Propagation Model (SPM). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .65

2.5.1 SPM Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .65
2.5.2.1 Visibility and Distance Between Transmitter and Receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66
2.5.2.2 Effective Transmitter Antenna Height . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66
2.5.2.3 Effective Receiver Antenna Height . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .69
2.5.2.4 Correction for Hilly Regions in Case of LOS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70
2.5.2.5 Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70
2.5.2.6 Losses due to Clutter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70
2.5.2.7 Recommendations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71
2.5.3 Automatic Calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72
2.5.3.1 General Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72
2.5.3.2 Sample Values for SPM Path Loss Formula Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .73
2.5.4 Unmasked Path Loss Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .74
2.6 WLL Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75

2.6.1 WLL Path Loss Formula. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75
2.7 ITU-R P.526-5 Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76

2.8 ITU-R P.370-7 Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76

2.9 Erceg-Greenstein (SUI) Propagation Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78

2.9.1 SUI Terrain Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78
2.9.2 Erceg-Greenstein (SUI) Path Loss Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .79
2.10 ITU-R P.1546-2 Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .80

2.10.1.1 Step 1: Determination of Graphs to be Used . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .80
2.10.1.2 Step 2: Calculation of Maximum Field Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81
2.10.1.3 Step 3: Determination of Transmitter Antenna Height . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81
2.10.1.4 Step 4: Interpolation/Extrapolation of Field Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81
2.10.1.5 Step 5: Calculation of Correction Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83
2.10.1.6 Step 6: Calculation of Path Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84
2.11 Sakagami Extended Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84
2.12 Free Space Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86
2.13 Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86

2.13.1 Knife-edge Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86
2.13.2 3 Knife-edge Deygout Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87
2.13.3 Epstein-Peterson Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89
2.13.4 Deygout Method with Correction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89
2.13.5 Millington Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89
2.14 Shadow Fading Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .90

2.14.1 Shadowing Margin Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .95
2.14.1.1 Shadowing Margin Calculation in Predictions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .95
2.14.1.2 Shadowing Margin Calculation in Monte-Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97
2.14.2 Macro-Diversity Gains Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97
2.14.2.1 Uplink Macro-Diversity Gain Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98
2.14.2.1.1 Shadowing Error PDF (n Signals). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98
2.14.2.1.2 Uplink Macro-Diversity Gain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
2.14.2.2 Downlink Macro-Diversity Gain Evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
2.14.2.2.1 Shadowing Error PDF (n Signals). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
2.14.2.2.2 Downlink Macro-Diversity Gain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
6
2.15 Path Loss Matrices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

2.15.1 Calculation Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
2.15.2 Validity of Path Loss Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
2.15.3 Path Loss Tuning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
2.15.3.1 Transmitter Path Loss Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
2.15.3.2 Repeater Path Loss Tuning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
2.16 File Formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

2.16.1 Path Loss Matrix File Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
2.16.1.1 Pathloss.dbf File Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
2.16.1.2 Pathloss.dbf File Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
2.16.1.3 LOS File Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
2.16.2 Path Loss Tuning File Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
2.16.2.1 Pathloss.dbf File Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
2.16.2.2 Pathloss.dbf File Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
2.16.2.3 PTS File Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
2.16.3 Interference Matrix File Formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
2.16.3.1 CLC Format (One Value per Line). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
2.16.3.2 CLC File Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
2.16.3.3 DCT File Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
2.16.3.4 IM0 Format (One Histogram per Line) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
2.16.3.5 IM1 Format (One Value per Line, TX Name Repeated) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
2.16.3.6 IM2 Format (Co- and Adjacent-channel Probabilities) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
2.16.4 "Per Transmitter" Prediction File Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
2.16.4.1 <per_transmitter_prediction>.dbf File Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
2.16.4.2 <per_transmitter_prediction>.dbf File Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
2.16.5 Coverage Prediction Export and Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
2.16.5.1 Filtering Coverage Predictions at Export. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
2.16.5.2 Smoothing Coverage Predictions at Export . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
2.16.5.3 Examples of Prediction Export Filtering and Smoothing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
2.16.5.4 Coverage Prediction Reports Over Focus/Computation Zones. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
3 GSM GPRS EDGE Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125

3.1 Signal Level Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
3.1.1 DL Signal Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
3.1.2 UL Signal Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
3.1.3 Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
3.1.3.1 Profile Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
3.1.3.2 Reception Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
3.1.4 Signal Level-based DL Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
3.1.4.1 DL Service Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
3.1.4.1.1 All Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
3.1.4.1.2 Best Signal Level and a Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
3.1.4.1.3 Second Best Signal Level and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
3.1.4.1.4 Best Signal Level per HCS Layer and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
3.1.4.1.5 Second Best Signal Level per HCS Layer and a Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
3.1.4.1.6 HCS Servers and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
3.1.4.1.7 Highest Priority HCS Server and a Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
3.1.4.1.8 Best Idle Mode Reselection Criterion (C2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
3.1.4.2 Coverage Display. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
3.1.4.2.1 Coverage Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
3.1.4.2.2 Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
3.2 Interference-based DL Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

3.2.1 DL Carrier-to-Interference Ratio Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
3.2.2 Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
3.2.3 Interference-based DL Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
3.2.3.1 Service Area Determination. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
3.2.3.2 Coverage Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
3.2.3.2.1 Interference Condition Satisfied by At Least One TRX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
3.2.3.2.2 Interference Condition Satisfied by The Worst TRX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
3.2.3.3 Coverage Display. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
3.2.3.3.1 Coverage Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
3.2.3.3.2 Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
3.3 GPRS/EDGE Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136

3.3.1 Coding Scheme Selection and Throughput Calculation Without Ideal Link Adaptation . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
7
3.3.1.1 Calculations Based on C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

3.3.1.2 Calculations Based on C/I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
3.3.1.3 Calculations Based on C/(I+N) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
3.3.2 Coding Scheme Selection and Throughput Calculation With Ideal Link Adaptation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
3.3.2.1 Calculations Based on C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
3.3.2.2 Calculations Based on C/I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
3.3.3 Application Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
3.3.4 BLER Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
3.3.5 GPRS/EDGE Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
3.3.5.1 Service Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
3.3.5.1.1 All Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
3.3.5.1.2 Best Signal Level and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
3.3.5.1.3 Second Best Signal Level and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
3.3.5.1.4 Best Signal Level per HCS Layer and a Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
3.3.5.1.5 Second Best Signal Level per HCS Layer and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
3.3.5.1.7 Highest Priority HCS Server and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
3.3.5.1.8 Best Idle Mode Reselection Criterion (C2). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
3.3.5.2 Coverage Display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
3.3.5.2.1 Coverage Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
3.3.5.2.2 Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
3.4 Codec Mode Selection and CQI Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145

3.4.1 Circuit Quality Indicator Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
3.4.2 CQI Calculation Without Ideal Link Adaptation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
3.4.2.1 Calculations Based on C/N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
3.4.3 CQI Calculation With Ideal Link Adaptation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
3.4.3.1 Calculations Based on C/N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
3.4.4 Circuit Quality Indicators Coverage Predictions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
3.4.4.1.1 All Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
3.4.4.1.4 Best Signal Level per HCS Layer and a Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
3.4.4.1.5 Second Best Signal Level per HCS Layer and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
3.4.4.1.7 Highest Priority HCS Server and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
3.4.4.2.1 Coverage Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
3.4.4.2.2 Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
3.5 UL Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152

3.5.1 DL Service Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
3.5.1.1 All Servers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
3.5.1.2 Best Signal Level and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
3.5.1.3 Second Best Signal Level and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
3.5.1.4 Best Signal Level per HCS Layer and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
3.5.1.5 Second Best Signal Level per HCS Layer and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
3.5.1.6 HCS Servers and a Margin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
3.5.1.7 Highest Priority HCS Server and a Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
3.5.1.8 Best Idle Mode Reselection Criterion (C2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
3.5.2 Coverage by UL Signal Level. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
3.5.2.1 Coverage Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
3.5.2.2 Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
3.5.2.2.1 UL Signal Level (in dBm, dBµV, dBµV/m) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
3.5.2.2.2 Best UL Signal Level (in dBm, dBµV, dBµV/m). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
3.5.2.2.3 UL Total Losses (dB). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
3.5.2.2.4 Minimum UL Total Losses (dB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
3.5.3 Coverage by UL C/I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
3.5.3.1 Coverage Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
3.5.3.2 UL C/I Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
3.5.3.3 Coverage Area Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
3.5.3.4 Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
3.5.3.4.1 C/I Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
8
3.5.3.4.2 Max C/I Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

3.5.3.4.3 Min C/I Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
3.5.4 Coverage by UL Coding Schemes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
3.5.4.2 Coding Scheme Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
3.5.4.3 Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
3.5.5 Coverage by UL Codec Modes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
3.5.5.2 Codec Mode Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
3.6 Traffic Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

3.6.1 Traffic Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
3.6.1.1 Normal Cells (Nonconcentric, No HCS Layer) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
3.6.1.1.1 Circuit Switched Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
3.6.1.1.2 Packet Switched Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
3.6.1.2 Concentric Cells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
3.6.1.3 HCS Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
3.6.2 Calculation of the Traffic Demand per Subcell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
3.6.2.1 User Profile Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
3.6.2.1.1 Normal Cells (Nonconcentric, No HCS Layer) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
3.6.2.1.2 Concentric Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
3.6.2.1.3 HCS Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
3.6.2.2 Sector Traffic Maps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
3.6.2.2.1 Normal Cells (Nonconcentric, No HCS Layer) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
3.6.2.2.2 Concentric Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
3.6.2.2.3 HCS Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
3.7 Network Dimensioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

3.7.1 Dimensioning Models and Quality Graphs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
3.7.1.1 Circuit Switched Traffic. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
3.7.1.2 Packet Switched Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
3.7.1.2.1 Throughput . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
3.7.1.2.2 Delay. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
3.7.1.2.3 Blocking Probability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
3.7.2 Network Dimensioning Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
3.7.2.1 Network Dimensioning Engine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
3.7.2.1.1 Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
3.7.2.1.2 Outputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
3.7.2.2 Network Dimensioning Steps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
3.7.2.2.1 Step 1: Timeslots Required for CS Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
3.7.2.2.2 Step 2: TRXs Required for CS Traffic and Dedicated PS Timeslots . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
3.7.2.2.3 Step 3: Effective CS Blocking, Effective CS Traffic Overflow and Served CS Traffic . . . . . . . . . . . . . . . 178
3.7.2.2.4 Step 4: TRXs to Add for PS Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
3.7.2.2.5 Step 5: Served PS Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
3.7.2.2.6 Step 6: Total Traffic Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
3.8 Key Performance Indicators Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181

3.8.1 Circuit Switched Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
3.8.1.1 Erlang B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
3.8.1.2 Erlang C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
3.8.1.3 Served Circuit Switched Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
3.8.2 Packet Switched Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
3.8.2.1 Case 1: Total Traffic Demand > Dedicated + Shared Timeslots. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
3.8.2.1.1 Traffic Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
3.8.2.1.2 Packet Switched Traffic Overflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
3.8.2.1.3 Throughput Reduction Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
3.8.2.1.4 Delay. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
3.8.2.1.5 Blocking Probability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
3.8.2.1.6 Served Packet Switched Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
3.8.2.2 Case 2: Total Traffic Demand < Dedicated + Shared Timeslots. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
3.8.2.2.1 Traffic Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
3.8.2.2.2 Packet Switched Traffic Overflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
3.8.2.2.3 Throughput Reduction Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
3.8.2.2.4 Delay. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
9
3.8.2.2.5 Blocking Probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183

3.8.2.2.6 Served Packet Switched Traffic. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
3.9 Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184

3.9.1 Radio Resource Management in GSM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
3.9.1.1 GSM Simulation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
3.9.1.2 Servers Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
3.9.1.3 Codec Mode Assignment and DL Power Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
3.9.1.4 Coding Scheme Assignment, Throughput Evaluation and DL Power Control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
3.9.1.5 Subcell Traffic Loads Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
3.9.1.6 Half-Rate Traffic Ratio Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
3.9.1.7 DL Power Control Gain Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
3.9.1.8 DTX DL Gain Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
3.9.1.9 GSM Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
3.10 Automatic Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191

3.10.1 Neighbour Allocation for All Transmitters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
3.10.2 Neighbour Allocation for a Group of Transmitters or One Transmitter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
3.10.3 Neighbour Importance Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
3.10.4 Appendix: Calculation of the Inter-Transmitter Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
3.11 AFP Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196

3.11.1 The AFP Cost Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
3.11.1.1 Cost Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
3.11.1.2 Cost Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
3.11.1.2.1 Separation Violation Cost Component. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
3.11.1.2.2 Interference Cost Component . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
3.11.1.2.3 I_DIV, F_DIV and Other Advanced Cost Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
3.11.2 The AFP Blocked Traffic Cost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
3.11.2.1 Calculation of New Traffic Loads Including Blocked Traffic Loads. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
3.11.2.2 Recalculation of CS and PS From Traffic Loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
3.11.2.3 Testing the Blocked Cost Using Traffic Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
3.11.3 Interference. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
3.11.3.1 Using Interferences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
3.11.3.2 Cumulative Density Function of C/I Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
3.11.3.3 Precise Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
3.11.3.4 Precise Interference Distribution Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
3.11.3.4.1 Direct Availability of Precise Interference Distribution to the AFP . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
3.11.3.4.2 Efficient Calculation and Storage of Interference Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
3.11.3.4.3 Robustness of the IM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
3.11.3.5 Traffic Load and Interference Information Discrimination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
4 UMTS HSPA Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211

4.1 General Prediction Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
4.1.1 Calculation Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
4.1.2 Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
4.1.2.1 Profile Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
4.1.2.2 Reception Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
4.1.3 Coverage Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
4.1.3.1.1 All Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
4.1.3.2.1 Plot Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
4.1.3.2.2 Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
4.2 Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214

4.2.1 Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
4.2.2 Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
4.2.3 Ec/I0 Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
4.2.4 DL Eb/Nt Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
4.2.5 UL Eb/Nt Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
4.3 Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225

4.3.1 Generating a Realistic User Distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
4.3.1.1 Simulations Based on User Profile Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
10
4.3.1.1.1 Circuit Switched Service (i) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226

4.3.1.1.2 Packet Switched Service (j). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
4.3.1.2 Simulations Based on Sector Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
4.3.1.2.1 Throughputs in Uplink and Downlink . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
4.3.1.2.2 Total Number of Users (All Activity Statuses) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
4.3.1.2.3 Number of Users per Activity Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
4.3.2 Power Control Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
4.3.2.1 Algorithm Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
4.3.2.2 R99 Part of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
4.3.2.3 HSDPA Part of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
4.3.2.3.1 HSDPA Power Allocation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
4.3.2.3.2 Number of HS-SCCH Channels and Maximum Number of HSDPA Bearer Users . . . . . . . . . . . . . . . . . 238
4.3.2.3.3 HSDPA Bearer Allocation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
4.3.2.3.4 Fast Link Adaptation Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
4.3.2.3.5 MIMO Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
4.3.2.3.6 Scheduling Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
4.3.2.3.7 Dual-Cell HSDPA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
4.3.2.4 HSUPA Part of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
4.3.2.4.1 Admission Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
4.3.2.4.2 HSUPA Bearer Allocation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
4.3.2.4.3 Noise Rise Scheduling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
4.3.2.4.4 Radio Resource Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
4.3.2.5 Convergence Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
4.3.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
4.3.3.1 R99 Related Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
4.3.3.2 HSPA Related Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
4.3.3.2.1 Statistics Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
4.3.3.2.2 Mobiles Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
4.3.3.2.3 Cells Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
4.3.3.2.4 Sites Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
4.3.4 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
4.3.4.1 Admission Control in the R99 Part . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
4.3.4.2 Resources Management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
4.3.4.2.1 OVSF Codes Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
4.3.4.2.2 Channel Elements Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
4.3.4.2.3 Iub Backhaul Throughput . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
4.3.4.3 Downlink Load Factor Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
4.3.4.3.1 Downlink Load Factor per Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
4.3.4.3.2 Downlink Load Factor per Mobile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
4.3.4.4 Uplink Load Factor Due to One User . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
4.3.4.5 Inter-carrier Power Sharing Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
4.3.4.6 Best Serving Cell Determination in Monte Carlo Simulations - Old Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
4.4 UMTS HSPA Prediction Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284

4.4.1 Best Serving Cell and Active Set Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
4.4.2 Point Analysis - AS Analysis Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
4.4.2.1 Bar Graph and Pilot Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
4.4.2.2 Downlink R99 Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
4.4.2.3 Uplink R99 Sub-Menu. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
4.4.2.4 HSDPA Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
4.4.2.5 HSUPA Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
4.4.3 Coverage Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
4.4.3.1 Pilot Quality Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
4.4.3.1.1 Prediction Study Inputs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
4.4.3.1.2 Study Display Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
4.4.3.2 Downlink Service Area Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
4.4.3.3 Uplink Service Area Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
4.4.3.4 Downlink Total Noise Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
4.4.3.4.1 Study Inputs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
4.4.3.4.2 Display Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
4.4.3.5 HSDPA Prediction Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
11
4.4.3.6 HSUPA Prediction Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309

4.4.3.6.1 Prediction Study Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
4.4.3.6.2 Calculation Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
4.4.3.6.3 Display Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
4.5 Automatic Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312

4.5.1 Neighbour Allocation for All Transmitters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
4.5.2 Neighbour Allocation for a Group of Transmitters or One Transmitter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
4.5.3 Importance Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
4.5.3.1 Importance of Intra-carrier Neighbours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
4.5.3.2 Importance of Inter-carrier Neighbours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 318
4.5.4 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
4.5.4.1 Calculation of the Inter-Transmitter Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
4.6 Primary Scrambling Code Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319

4.6.1 Automatic Allocation Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
4.6.1.1 Options and Constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
4.6.1.2 Allocation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
4.6.1.2.1 Single Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
4.6.1.2.2 Multi-Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
4.6.1.3 Priority Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
4.6.1.3.1 Cell Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
4.6.1.3.2 Transmitter Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
4.6.1.3.3 Site Priority. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
4.6.2 Allocation Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
4.6.2.1 Allocation Strategies and Use a Maximum of Codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
4.6.2.1.1 Strategy: Clustered . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
4.6.2.1.2 Strategy: Distributed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
4.6.2.1.3 Strategy: ‘One Cluster per Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
4.6.2.1.4 Strategy: ‘Distributed per Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
4.6.2.2 Allocate Carriers Identically . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
4.7 Automatic GSM-UMTS Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329

4.7.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
4.7.2.1 Algorithm Based on Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
4.7.2.2 Algorithm Based on Coverage Overlapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
4.7.2.3 Appendices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333
4.7.2.3.1 Delete Existing Neighbours Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333
5 CDMA2000 Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337

5.1 General Prediction Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337
5.1.1 Calculation Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337
5.1.2 Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
5.1.2.1 Profile Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
5.1.2.2 Reception Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
5.1.3 Coverage Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
5.1.3.1.1 All Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
5.1.3.2.1 Plot Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
5.1.3.2.2 Display Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
5.2 Definitions and Formulas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340

5.2.1 Parameters Used for CDMA2000 1xRTT Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
5.2.1.1 Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
5.2.1.2 Ec/I0 Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
5.2.1.3 DL Eb/Nt Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
5.2.1.4 UL Eb/Nt Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
5.2.1.5 Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
5.2.2 Parameters Used for CDMA2000 1xEV-DO Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
5.2.2.1 Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
5.2.2.2 Ec/I0 and Ec/Nt Calculations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354
5.2.2.3 UL Eb/Nt Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
5.2.2.4 Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
12
5.3 Active Set Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
5.4 Simulations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358

5.4.1 Generating a Realistic User Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359
5.4.1.1 Number of Users, User Activity Status and User Throughput. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359
5.4.1.1.1 Simulations Based on User Profile Traffic Maps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359
5.4.1.1.2 Simulations Based on Sector Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
5.4.1.2 Transition Flags for 1xEV-DO Rev.0 User Throughputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
5.4.1.3 User Geographical Position . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
5.4.2 Network Regulation Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
5.4.2.1 CDMA2000 1xRTT Power Control Simulation Algorithm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
5.4.2.1.1 Algorithm Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
5.4.2.1.2 Presentation of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
5.4.2.1.3 Convergence Criterion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376
5.4.2.2 CDMA2000 1xEV-DO Power/Data Rate Control Simulation Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
5.4.2.2.1 Algorithm Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
5.4.2.2.2 Presentation of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
5.4.2.2.3 Convergence Criterion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
5.4.3 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
5.4.3.1 Admission Control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
5.4.3.2 Resources Management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
5.4.3.2.1 Walsh Code Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
5.4.3.2.2 Channel Element Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
5.4.3.3 Downlink Load Factor Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388
5.4.3.3.1 Downlink Load Factor per Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388
5.4.3.3.2 Downlink Load Factor per Mobile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
5.4.3.4 Best Server Determination in Monte Carlo Simulations - Old Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
5.4.3.5 Radio Bearer Allocation Algorithm for Multi-carrier EVDO Rev.B - Old Method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392
5.5 CDMA2000 Prediction Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392

5.5.1 Point Analysis: The AS Analysis Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392
5.5.1.1 Bar Graph and Pilot Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392
5.5.1.2 Downlink Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
5.5.1.2.1 CDMA2000 1xRTT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
5.5.1.2.2 CDMA2000 1xEV-DO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
5.5.1.3 Uplink Sub-Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
5.5.1.3.1 CDMA2000 1xRTT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
5.5.1.3.2 CDMA2000 1xEV-DO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406
5.5.2 Coverage Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
5.5.2.1 Pilot Quality Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
5.5.2.2 Downlink Service Area Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
5.5.2.2.1 CDMA2000 1xRTT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
5.5.2.2.2 CDMA2000 1xEV-DO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
5.5.2.3 Uplink Service Area Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417
5.5.2.3.1 CDMA2000 1xRTT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417
5.5.2.3.2 CDMA2000 1xEV-DO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
5.5.2.4 Downlink Total Noise Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422
5.5.2.4.1 Analysis on the Best Carrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422
5.5.2.4.2 Analysis on a Specific Carrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423
5.6 Automatic Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423

5.6.1 Neighbour Allocation for all Transmitters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424
5.6.2 Neighbour Allocation for a Group of Transmitters or One Transmitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
5.6.3 Importance Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
5.6.3.1 Importance of Intra-carrier Neighbours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
5.6.3.2 Importance of Inter-carrier Neighbours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429
5.6.4 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
5.6.4.1 Calculation of the Inter-Transmitter Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
5.7 PN Offset Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430

5.7.1 Automatic Allocation Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
5.7.1.1 Options and Constraints. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
5.7.1.2 Allocation Process. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
5.7.1.2.1 Single Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
5.7.1.2.3 Difference between Adjacent and Distributed PN-Clusters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
5.7.1.3 Priority Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
5.7.1.3.1 Cell Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
13
5.7.1.3.2 Transmitter Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435

5.7.1.3.3 Site Priority. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
5.7.2 Allocation Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
5.7.2.1 Strategy: PN Offset per Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
5.7.2.2 Strategy: Adjacent PN-Clusters Per Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
5.7.2.3 Strategy: ‘Distributed PN-Clusters Per Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
5.8 Automatic GSM-CDMA Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438

5.8.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
5.8.2.1 Algorithm Based on Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
5.8.2.2 Algorithm Based on Coverage Overlapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
5.8.2.3 Delete Existing Neighbours Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441
6 LTE Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445

6.1 Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
6.2 Calculation Quick Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450

6.2.1 Downlink Transmission Powers Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450
6.2.2 Co- and Adjacent Channel Overlaps Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
6.2.3 Signal Level Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
6.2.4 Noise Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
6.2.5 Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456
6.2.6 C/N Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460
6.2.7 C/(I+N) Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460
6.2.8 Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 462
6.2.9 Noise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 462
6.2.10 Interference Calculation (UL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 462
6.2.11 Noise Rise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463
6.2.12 C/N Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463
6.2.13 C/(I+N) Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463
6.2.14 Calculation of Downlink Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464
6.2.15 Calculation of Uplink Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
6.2.16 Calculation of Downlink UE Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
6.2.17 Calculation of Uplink UE Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
6.2.18 Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-user Throughput Calculation . . . . . . . 466
6.2.19 Scheduling and Radio Resource Management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
6.2.20 User Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
6.3 Available Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470

6.3.1 Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470
6.3.1.1 Profile View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470
6.3.1.2 Reception View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470
6.3.1.3 Interference View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470
6.3.1.4 Details View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470
6.3.2 Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471
6.3.2.1 Downlink Signal Level Coverage Predictions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471
6.3.2.2 Effective Signal Analysis Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 472
6.3.2.3 C/(I+N)-based Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473
6.3.2.4 Cell Identifier Collision Zones Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
6.3.3 Calculations on Subscriber Lists. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 476
6.3.4 Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 476
6.3.4.1 User Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 476
6.3.4.1.1 Simulations Based on User Profile Traffic Maps and Subscriber Lists . . . . . . . . . . . . . . . . . . . . . . . . . 476
6.3.4.1.2 Simulations Based on Sector Traffic Maps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
6.3.4.2 Simulation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 479
6.4 Calculation Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 485

6.4.1 Downlink Transmission Power Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 485
6.4.2.1 Conversion From Channel Numbers to Start and End Frequencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494
6.4.2.2 Co-Channel Overlap Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495
6.4.2.3 Adjacent Channel Overlap Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 496
6.4.2.4 Total Overlap Ratio Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 496
6.4.3 Subframe Pattern Collision Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497
6.4.3.1 Subframe Pattern Normalisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497
6.4.3.2 Determination of Effective Subframe Patterns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498
14
6.4.3.3 Calculation of Subframe Collision Probabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499

6.4.4 Signal Level and Signal Quality Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501
6.4.4.1 Signal Level Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501
6.4.4.2 Noise Calculation (DL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505
6.4.4.3 Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505
6.4.4.4 C/N Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516
6.4.4.5 C/(I+N) and Bearer Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 518
6.4.4.6 Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 523
6.4.4.7 Noise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525
6.4.4.8 Interference Calculation (UL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525
6.4.4.8.1 Interfering Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526
6.4.4.8.2 Noise Rise Calculation (UL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528
6.4.4.9 C/N Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 529
6.4.4.10 C/(I+N) and Bearer Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 532
6.4.5 Best Server Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535
6.4.6 Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538
6.4.6.1 Calculation of Total Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538
6.4.6.1.1 Calculation of Downlink Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538
6.4.6.1.2 Calculation of Uplink Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543
6.4.6.2 Calculation UE Capacities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545
6.4.6.2.1 Calculation of Downlink UE Capacity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545
6.4.6.2.2 Calculation of Uplink UE Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 546
6.4.6.3 Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-user Throughput Calculation . . . . . . 547
6.4.7 Scheduling and Radio Resource Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 552
6.4.7.1 Scheduling and Radio Resource Allocation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 552
6.4.7.2 User Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561
6.5 Automatic Planning Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563

6.5.1 Automatic Neighbour Planning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563
6.5.2 Automatic Inter-technology Neighbour Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567
6.5.3 Automatic Frequency Planning Using the AFP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 570
6.5.3.1 Constraint and Relationship Weights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 570
6.5.3.2 Cost Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 571
6.5.3.3 AFP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 571
6.5.4 Automatic Physical Cell ID Planning Using the AFP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 572
6.5.4.3 AFP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 576
6.5.5 Automatic PRACH RSI Planning Using the AFP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 576
6.5.5.3 AFP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578
6.5.6 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578
6.5.6.1 Interference Matrix Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578
6.5.6.2 Distance Importance Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
7 3GPP Multi-RAT Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .583

7.1 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583
7.2 Multi-RAT Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583

7.2.1 User Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583
7.2.2 Simulation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585
7.3 Multi-RAT Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585
8 3GPP2 Multi-RAT Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .589

8.1 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
8.2 Multi-RAT Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589

8.2.1 User Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
8.2.2 Simulation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 590
8.3 Multi-RAT Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
9 TD-SCDMA Networks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .595
15
9.1 Definitions and Formulas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595

9.1.1 Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
9.1.2 P-CCPCH Eb/Nt and C/I Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 600
9.1.3 DwPCH C/I Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 600
9.1.4 DL TCH Eb/Nt and C/I Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 601
9.1.5 UL TCH Eb/Nt and C/I Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 601
9.1.6 Interference Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602
9.1.7 HSDPA Dynamic Power Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602
9.2 Signal Level Based Calculations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602

9.2.1 Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602
9.2.1.1 Profile Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 603
9.2.1.2 Reception Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 603
9.2.2 RSCP Based Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 603
9.2.2.1 Calculation Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 603
9.2.2.2 P-CCPCH RSCP Coverage Prediction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604
9.2.2.2.1 Coverage Condition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604
9.2.2.2.2 Coverage Display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604
9.2.2.3 Best Server P-CCPCH Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605
9.2.2.4 P-CCPCH Pollution Analysis Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605
9.2.2.5 DwPCH RSCP Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605
9.2.2.6 UpPCH RSCP Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606
9.2.2.7 Baton Handover Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 607
9.2.2.8 Scrambling Code Interference Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 607
9.3 Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608

9.3.1 Generating a Realistic User Distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608
9.3.1.1 Simulations Based on User Profile Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608
9.3.1.1.1 Circuit Switched Service (i) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609
9.3.1.1.2 Packet Switched Service (j) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609
9.3.1.2 Simulations Based on Sector Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 612
9.3.1.2.1 Throughputs in Uplink and Downlink. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 612
9.3.1.2.2 Total Number of Users (All Activity Statuses) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 613
9.3.1.2.3 Number of Users per Activity Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 613
9.3.2 Power Control Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 613
9.3.2.1 Algorithm Initialisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 614
9.3.2.2 R99 Part of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 614
9.3.2.2.1 Determination of Mi’s Best Server (SBS(Mi)) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 614
9.3.2.2.2 Dynamic Channel Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615
9.3.2.2.3 Uplink Power Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 617
9.3.2.2.4 Downlink Power Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 619
9.3.2.2.5 Uplink Signals Update . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621
9.3.2.2.6 Downlink Signals Update. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621
9.3.2.2.7 Control of Radio Resource Limits (Downlink Traffic Power and Uplink Load) . . . . . . . . . . . . . . . . . . 621
9.3.2.3 HSDPA Part of the Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 622
9.3.2.3.1 HSDPA Power Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 622
9.3.2.3.2 Connection Status and Number of HSDPA Users . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624
9.3.2.3.3 HSDPA Admission Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624
9.3.2.3.4 HSDPA Dynamic Channel Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625
9.3.2.3.5 Ressource Unit Saturation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625
9.3.2.4 Convergence Criteria. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625
9.4 TD-SCDMA Prediction Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626

9.4.1 P-CCPCH Reception Analysis (Eb/Nt) or (C/I) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626
9.4.2 DwPCH Reception Analysis (C/I) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 628
9.4.3 Downlink TCH RSCP Coverage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 629
9.4.4 Uplink TCH RSCP Coverage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 630
9.4.5 Downlink Total Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 631
9.4.6 Downlink Service Area Analysis (Eb/Nt) or (C/I). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 631
9.4.7 Uplink Service Area Analysis (Eb/Nt) or (C/I) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 633
9.4.8 Effective Service Area Analysis (Eb/Nt) or (C/I) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 635
9.4.9 Cell to Cell Interference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 636
16
9.4.10 UpPCH Interference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 637

9.4.11 HSDPA Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 637
9.5 Smart Antenna Modelling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 638

9.5.1 Modelling in Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 639
9.5.1.1 Grid of Beams Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 639
9.5.1.2 Adaptive Beam Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 640
9.5.1.3 Statistical Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 641
9.5.1.4 Beamforming Smart Antenna Models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 641
9.5.1.5 3rd Party Smart Antenna Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642
9.5.2 Construction of the Geographic Distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642
9.5.3 Modelling in Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 643
9.5.4 HSDPA Quality and Throughput Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644
9.5.4.1 Fast Link Adaptation Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644
9.5.4.1.1 CQI Based on P-CCPCH Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644
9.5.4.1.2 CQI Based on HS-PDSCH Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 648
9.5.4.2 Coverage Prediction Display Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 649
9.5.4.2.1 Colour per CQI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 649
9.5.4.2.2 Colour per Peak Throughput . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650
9.5.4.2.3 Colour per HS-PDSCH Ec/Nt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650
9.6 N-Frequency Mode and Carrier Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650

9.6.1 Automatic Carrier Allocation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650
9.7 Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651

9.7.1 Neighbour Allocation for All Transmitters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651
9.7.2 Neighbour Allocation for a Group of Transmitters or One Transmitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655
9.7.3 Importance Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655
9.7.4 Appendix: Calculation of the Inter-Transmitter Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656
9.8 Scrambling Code Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656

9.8.1.1 Allocation Constraints and Options. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 657
9.8.1.2 Allocation Strategies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658
9.8.1.3 Allocation Process. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658
9.8.1.3.1 Single Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658
9.8.1.4 Priority Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 660
9.8.1.4.1 Cell Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 660
9.8.1.4.2 Transmitter Priority. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 662
9.8.1.4.3 Site Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 663
9.8.2 Scrambling Code Allocation Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 663
9.8.2.1 Single Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 663
9.8.2.1.1 Strategy: Clustered . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664
9.8.2.1.2 Strategy: Distributed per Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664
9.8.2.1.3 Strategy: One SYNC_DL Code per Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
9.8.2.1.4 Strategy: Distributed per Site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
9.8.2.2 Multi Carrier Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
9.9 Automatic GSM/TD-SCDMA Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 666

9.9.1.1 Algorithm Based on Distance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 667
9.9.1.2 Algorithm Based on Coverage Overlapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 667
9.9.1.3 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669
9.9.1.3.1 Delete Existing Neighbours Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669
10 WiMAX BWA Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .673

10.1 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 673
10.2 Calculation Quick Reference. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 678

10.2.1 Co- and Adjacent Channel Overlaps Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 678
10.2.2 Preamble Signal Level Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679
10.2.3 Preamble Noise Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679
10.2.4 Preamble Interference Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679
10.2.5 Preamble C/N Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679
10.2.6 Preamble C/(I+N) Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 680
10.2.7 Traffic and Pilot Signal Level Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 680
17
10.2.8 Traffic and Pilot Noise Calculation (DL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 680

10.2.9 Traffic and Pilot Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 681
10.2.10 Traffic and Pilot C/N Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 682
10.2.11 Traffic and Pilot C/(I+N) Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 682
10.2.12 Traffic Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 683
10.2.13 Traffic Noise Calculation (UL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 683
10.2.14 Traffic Interference Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 683
10.2.15 Traffic C/N Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684
10.2.16 Traffic C/(I+N) Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684
10.2.17 Calculation of Total Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684
10.2.18 Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-user Throughput Calculation . . . . . . . 685
10.3 Available Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 690

10.3.1 Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 690
10.3.1.1 Profile View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 690
10.3.1.2 Reception View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 690
10.3.1.3 Interference View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 690
10.3.1.4 Details View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 690
10.3.2 Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 691
10.3.2.1 Preamble Signal Level Coverage Predictions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 691
10.3.2.2 Effective Signal Analysis Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 692
10.3.2.3 C/(I+N)-based Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 693
10.3.2.4 Cell Identifier Collision Zones Coverage Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 695
10.3.3 Calculations on Subscriber Lists. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 695
10.3.4 Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 696
10.3.4.1 User Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 696
10.3.4.1.2 Simulations Based on Sector Traffic Maps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 698
10.3.4.2 Simulation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 699
10.4 Calculation Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 702

10.4.1.1 Conversion From Channel Numbers to Start and End Frequencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 703
10.4.1.2 Co-Channel Overlap Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 704
10.4.1.3 Adjacent Channel Overlap Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 705
10.4.1.4 FDD – TDD Overlap Ratio Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 705
10.4.2 Preamble Signal Level and Quality Calculations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707
10.4.2.1 Preamble Signal Level Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707
10.4.2.2 Preamble Noise Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708
10.4.2.3 Preamble Interference Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 710
10.4.2.4 Preamble C/N Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 712
10.4.2.5 Preamble C/(I+N) Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 712
10.4.3 Best Server Determination. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 713
10.4.4 Service Area Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 714
10.4.5 Permutation Zone Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 714
10.4.6 Traffic and Pilot Signal Level and Quality Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715
10.4.6.1 Traffic and Pilot Signal Level Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715
10.4.6.2 Traffic and Pilot Noise Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 717
10.4.6.3 Traffic and Pilot Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 718
10.4.6.3.1 Traffic and Pilot Interference Signal Levels Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 718
10.4.6.3.2 Effective Traffic and Pilot Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 722
10.4.6.4 Traffic and Pilot C/N Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 726
10.4.6.5 Traffic and Pilot C/(I+N) and Bearer Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 727
10.4.6.6 Traffic Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 729
10.4.6.7 Traffic Noise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 730
10.4.6.8 Traffic Interference Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 731
10.4.6.8.1 Traffic Interference Signal Levels Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 731
10.4.6.8.2 Noise Rise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 733
10.4.6.9 Traffic C/N Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 734
10.4.6.10 Traffic C/(I+N) and Bearer Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 737
10.4.7 Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 740
10.4.7.1 Calculation of Total Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 740
10.4.7.1.1 Calculation of Sampling Frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 740
10.4.7.1.2 Calculation of Symbol Duration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 741
18
10.4.7.1.3 Calculation of Total Cell Resources - TDD Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 741

10.4.7.1.4 Calculation of Total Cell Resources - FDD Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 742
10.4.7.2 Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-User Throughput Calculation . . . . . . 743
10.4.8.1 Scheduling and Radio Resource Allocation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748
10.4.8.2 User Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 757
10.5 Automatic Planning Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 758

10.5.1 Automatic Neighbour Planning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 759
10.5.2 Automatic Inter-technology Neighbour Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 763
10.5.3 Automatic Frequency Planning Using the AFP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 765
10.5.3.3 AFP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767
10.5.4 Automatic Preamble Index Planning Using the AFP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767
10.5.4.3 AFP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771
10.5.5 Automatic Zone PermBase Planning Using the AFP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771
10.5.5.3 AFP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774
10.5.6 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774
10.5.6.1 Interference Matrix Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774
10.5.6.2 Distance Importance Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774
11 Wi-Fi Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .779

11.1 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 779
11.2 Calculation Quick Reference. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 782

11.2.2 Signal Level Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783
11.2.3 Noise Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783
11.2.4 Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783
11.2.5 C/N Calculation (DL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 784
11.2.6 C/(I+N) Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 784
11.2.7 Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 784
11.2.8 Noise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 784
11.2.9 Interference Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 784
11.2.10 C/N Calculation (UL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 785
11.2.11 C/(I+N) Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 785
11.2.12 Calculation of Total Cell Resources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 785
11.2.13 Channel Throughput, Cell Capacity, and Per-user Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 786
11.3 Available Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789

11.3.1 Point Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789
11.3.1.1 Profile View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789
11.3.1.2 Reception View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789
11.3.1.3 Interference View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789
11.3.2 Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789
11.3.2.1 Signal Level Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789
11.3.2.2 Effective Signal Analysis Coverage Predictions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 791
11.3.2.3 C/(I+N)-based Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 791
11.3.3 Calculations on Subscriber Lists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 793
11.3.4 Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 793
11.3.4.1 User Distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 793
11.3.4.1.2 Simulations Based on Sector Traffic Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 795
11.3.4.2 Simulation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 797
11.4 Calculation Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 799

11.4.1.1 Conversion From Channel Numbers to Start and End Frequencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 800
11.4.1.2 Co-Channel Overlap Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 800
11.4.1.3 Adjacent Channel Overlap Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 801
19

11.4.2 Signal Level and Quality Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 803
11.4.2.1 Signal Level Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 803
11.4.2.2 Noise Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 804
11.4.2.3 Interference Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 804
11.4.2.4 C/N Calculation (DL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 806
11.4.2.5 C/(I+N) and Bearer Calculation (DL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 807
11.4.2.6 Signal Level Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 809
11.4.2.7 Noise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 810
11.4.2.8 Interference Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 810
11.4.2.8.1 Interference Signal Levels Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 811
11.4.2.8.2 Noise Rise Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 811
11.4.2.9 C/N Calculation (UL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 812
11.4.2.10 C/(I+N) and Bearer Calculation (UL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 813
11.4.3 Best Server Determination. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 815
11.4.4 Service Area Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 815
11.4.5 Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 816
11.4.5.1 Calculation of Total Cell Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 816
11.4.5.2 Channel Throughput, Cell Capacity, and Per-user Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 817
11.4.6.1 Scheduling and Radio Resource Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 820
11.4.6.2 User Throughput Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 823
11.5 Automatic Planning Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 824

11.5.1 Automatic Neighbour Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 825
11.5.2 Automatic Inter-technology Neighbour Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 829
11.5.3 Automatic Frequency Planning Using the AFP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 831
11.5.3.1 Constraint and Relationship Weights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 832
11.5.3.2 Cost Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 832
11.5.3.3 AFP Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 833
11.5.4 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 833
11.5.4.1 Interference Matrix Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 833
11.5.4.2 Distance Importance Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 834
12 ACP Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 839

12.1 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 839
12.1.1 Quality Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 839
12.1.1.1 Definition and Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 839
12.1.1.2 Progressive Thresholds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 840
12.1.1.3 Target Filtering. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 841
12.1.2 Quality Indicators in the ACP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 841
12.1.2.1 GSM Quality Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 841
12.1.2.2 UMTS Quality Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 841
12.1.2.3 CDMA2000 Quality Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 841
12.1.2.4 LTE Quality Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 842
12.1.2.5 WiMAX Quality Indicators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 842
12.1.2.6 Quality Indicator Parameters and Reference Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 842
12.1.2.7 Advanced Objective Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 843
12.1.2.8 Cost Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 843
12.1.3 Atoll and ACP Prediction Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 843
12.2 Quality Predictions and the Antenna Masking Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 844

12.2.1 Optimised Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 844
12.2.2 Antenna Masking Modes for Non-Native Propagation Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 844
12.2.2.1 Basic Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 844
12.2.2.2 Improved Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 845
12.2.2.3 Antenna Correction Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 845
12.2.2.4 Full Path Loss Method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 845
12.2.3 CrossWave Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 846
12.2.4 Antenna Masking and Repeaters, Remote Antennas, and Secondary Antennas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 846
12.3 Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 846

12.3.1 Configuring an Optimisation Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 846
12.3.1.1 Antenna Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 846
12.3.1.2 Additional Electrical Tilt (AEDT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 847
12.3.1.3 Relative Electrical Tilt Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 847
20
12.4 Multi-RAT and Co-planning Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 848

12.4.1 Multi-RAT and Co-planning Modes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 848
12.4.2 Technology Layer Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 849
12.5 Optimisation Methodology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 849

12.5.1 Search Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 849
12.5.2 Tuning Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 850
12.5.3 Sorting Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 851
12.5.4 Global Score Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 851
12.5.4.1 Search Algorithm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 851
12.5.4.2 Tuning Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 852
12.5.5 Weighting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 852
12.5.6 Controlling the Optimisation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 852
12.5.7 Implementation Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 853
12.5.8 Memory Usage and Optimisation Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 854
12.5.9 Internal Data Management and Performance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 854
12.5.9.1 Memory Usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 854
12.5.9.2 Disk Space Usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 855
12.6 Load Balancing Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 855

12.6.1 Principle Used in ACP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 855
12.6.2 Optimisation Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 855
12.6.2.1 Traffic Capture for Load Balancing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 855
12.6.2.2 Cell Capacity Load Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 856
12.6.2.3 Load Balancing Score Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 856
12.6.2.4 Load Quality Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 857
12.6.2.5 Captured Traffic Ratio. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 857
12.6.2.6 Introduction of Load Balancing as a Quality Indicator. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 858
12.6.3 Quality Figures Used for Graphs and Statistics Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 858
12.6.3.1 Load Balance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 859
12.6.3.2 Average Load. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 859
12.6.4 Optimisation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 859
12.6.4.1 Load Balancing Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 859
12.6.4.2 Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 860
12.6.5 Impact on the Global Score Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 861
12.7 EMF Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 861

12.7.1 Concepts of ACP EMF Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 862
12.7.1.1 Propagation Classes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 862
12.7.1.2 Terrain Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 862
12.7.1.3 Distribution of Evaluation Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 862
12.7.1.4 The Contribution of Transmitter Power to EMF Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 863
12.7.1.5 Worst-case Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 863
12.7.2 General Workflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 863
12.7.3 EMF Exposure Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 863
12.8 Shadowing Margin and Indoor Coverage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865
12.9 Multi-Storey Optimisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865

12.9.1 Path Loss Calculation and Data Caching. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865
12.9.2 Pixel Weighting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865
12.9.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865
12.9.4 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 866
12.9.5 Concepts of ACP EMF Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 866
12.9.5.1 Propagation Classes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 866
12.9.5.2 Terrain Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 867
12.9.5.3 Distribution of Evaluation Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 867
12.9.5.4 The Contribution of Transmitter Power to EMF Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 867
12.9.5.5 Worst-case Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 867
12.9.6 General Workflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 867
12.10 ACP Software Data Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 868
21
22
Chapter 1
Antennas and
Equipment
This chapter covers the following topics:
• "Antenna Attenuation" on page 25
• "Antenna Pattern Smoothing" on page 27
• "Power Received From Secondary Antennas" on page 29
• "Transmitter Radio Equipment" on page 30
• "Repeaters and Remote Antennas" on page 32
• "Beamforming Smart Antenna Models" on page 43
• "Grid-of-Beams Smart Antenna Model" on page 51
• "Adaptive Beam Smart Antenna Model" on page 52
• "Statistical Smart Antenna Gain Model" on page 53
Atoll 3.3.0 Technical Reference Guidefor Radio Networks
© Forsk 2015
24
AT330_TRR_E1 Chapter 1: Antennas and Equipment
1 Antennas and Equipment

1.1 Antenna Attenuation
To determine the transmitter antenna attenuation, Atoll calculates the accurate azimuth and tilt angles and performs 3D
interpolation of the horizontal and vertical patterns.
1.1.1 Calculation of Azimuth and Tilt Angles

From the direction of the transmitter antenna and the receiver position relative to the transmitter, Atoll determines the
receiver position relative to the direction of the transmitter antenna (i.e. the direction of the transmitter-receiver path in the
transmitter antenna coordinate system).
aTx and eTx are respectively the transmitter (Tx) antenna azimuth and tilt in the coordinate system S 0  x y z  .
aRx and eRx are respectively the azimuth and tilt of the receiver (Rx) in the coordinate system S 0  x y z  .
d is the distance between the transmitter (Tx) and the receiver (Rx).
Figure 1.1: Azimuth and Tilt Computation
In the coordinate system S 0  x y z  , the receiver coordinates are:
x Rx cos  e Rx   sin  a Rx   d
y Rx = cos  e Rx   cos  a Rx   d (1)
z Rx – sin  e Rx   d
Let az and el respectively be the azimuth and tilt of the receiver in the transmitter antenna coordinate system S Tx  x'' y'' z''  .
These angles describe the direction of the transmitter-receiver path in the transmitter antenna coordinate system. Therefore,
the receiver coordinates in S Tx  x'' y'' z''  are:
x'' Rx cos  el   sin  az   d

y'' Rx = cos  el   cos  az   d (2)
z'' Rx – sin  el   d
According to the figure above, we have the following relations:
x' cos  a Tx  – sin  a Tx  0 x

y' = sin  a Tx  cos  a Tx  0  y (3)
z' 0 0 1 z
and
25
Chapter 1: Antennas and Equipment ©Forsk 2015
x'' 1 0 0 x'
y'' = 0 cos  e Tx  – sin  e Tx   y' (4)
z'' 0 sin  e Tx  cos  e Tx  z'
Therefore, the relation between the system S 0  x y z  and the transmitter antenna system S Tx  x'' y'' z''  is:
x'' 1 0 0 cos  a Tx  – sin  a Tx  0 x

y'' = 0 cos  e Tx  – sin  e Tx   sin  a Tx  cos  a Tx  0  y (5)
z'' 0 sin  e Tx  cos  e Tx  0 0 1 z
We get,
x'' cos  a Tx  – sin  a Tx  0 x

y'' = cos  e Tx   sin  a Tx  cos  e Tx   cos  a Tx  – sin  e Tx   y (6)
z'' sin  e Tx   sin  a Tx  sin  e Tx   cos  a Tx  cos  e Tx  z
Then, substituting the receiver coordinates in the system S0 from Eq. (1) and the receiver coordinates in the system STx from
Eq. (2) in Eq. (6) leads to a system where two solutions are possible:
1st solution: If a Rx = a Tx , then az = 0 and el = eRx – e Tx
2nd solution: If a Rx  a Tx , then
1
az = atan ----------------------------------------------------------------------------------------
cos  e Tx  sin  e Tx   tan  e Rx 
----------------------------------- + ----------------------------------------------
tan  a Rx – a Tx  sin  a Rx – a Tx 
and
 – sin  e Tx  cos  e Tx   tan  e Rx  

el = atan sin  az    ----------------------------------
- + ----------------------------------------------
-
 tan  a Rx – a Tx  sin  a Rx – a Tx  
If sin  az   sin  a Rx – a Tx   0 , then az = az + 180
1.1.2 Antenna Pattern 3D Interpolation

The direction of transmitter-receiver path in the transmitter antenna coordinate system is given by angle values az and el.
Atoll considers these values to determine transmitter antenna attenuations in horizontal and vertical patterns.
It reads the following:
• H(az) the attenuation in the horizontal pattern for the calculated azimuth angle az
H(a0) the attenuation in the horizontal pattern for the electrical azimuth angle a0
• V(el) the attenuation V(el) in the vertical pattern for the calculated tilt angle el
Then it calculates the antenna total attenuation, L antTx  az el  :
180 – az – a 0 az – a 0
L antTx  az el  = H  az  – ----------------------------------
-   H  a 0  – V  el   + -------------------
-   H  180 + a 0  – V  180 – el   if |el| ≠ 90°
180 180
Else: L antTx  az el  = V(el)
Atoll assumes that the horizontal and vertical patterns are cross-sections of a 3D pattern. In other words, the description of
the antenna pattern must satisfy the following:
H(0)=V(0) and H()=V()
If the electrical tilt is e0, the horizontal pattern is a conical section with an elevation of e0 degrees off the horizontal plane.
If the electrical azimuth is a0, the vertical pattern is a plane section with a rotation a0 degrees from the vertical plane.
In this case, the description of the antenna pattern must satisfy the following conditions:
H(a0)=V(e0) and H(180+a0)=V(180-e0)
26
If the constraints listed above are satisfied, this implies that:

• Interpolated horizontal and vertical patterns respectively fit in with the entered horizontal and vertical patterns, even
in case of electrical tilt, and
• The contribution of both front and back parts of the vertical pattern are taken into account.
Otherwise, only the second point is guaranteed.
• This interpolation is performed in dBs.

• Angle values in formulas are stated in degrees.
• This interpolation is not used with 3D antenna patterns.
1.1.3 Additional Electrical Downtilt Modelling

The additional electrical downtilt, AEDT, also referred to as remote electrical downtilt or REDT, introduces a conical
transformation of the 3D antenna pattern in the vertical axis. In order to take it into account, the vertical pattern is
transformed as follows:
V  x  = V  x – AEDT  when x  [– 90,90]
V  x  = V  x + AEDT  when x  [90,270]

Where, the angle values are in degrees.
The vertical pattern transformation is represented below. The left picture shows the initial vertical pattern when there is no
electrical downtilt and the right one shows the vertical pattern transformation due to an electrical downtilt of 10°.
Then, Atoll proceeds as explained in the previous section. It determines the antenna attenuation in the transformed vertical
pattern for the calculated tilt angle (V(el)) and applies the 3D interpolation formula in order to calculate the antenna total
attenuation, L antTx  az el  .
Figure 1.2: Vertical Pattern Transformation due to Electrical Downtilt
1.2 Antenna Pattern Smoothing

Empirical propagation models, like the Standard Propagation Model (SPM), require antenna pattern smoothing in the vertical
plane to simulate the effects of reflections and diffractions. Signal level predictions can be improved by smoothing the high-
attenuation points of the vertical pattern. You can smooth vertical as well as horizontal antenna patterns in Atoll.
The antenna pattern smoothing algorithm in Atoll first determines the peaks and nulls in the pattern within the smoothing
angle (ASmoothing) defined by the user. Peaks (P) are the lowest attenuation angles and nulls (N) are the highest attenuation
angles in the pattern. Then, it determines the nulls to be smoothed (NSmoothing) and their corresponding angles according to
the defined Peak-to-Null Deviation (DPeak-to-Null). DPeak-to-Null is the minimum difference of attenuation in dBs between two
peaks and a null between them. Finally, Atoll smoothes the pattern between 0 and the smoothing angle (ASmoothing) by
applying the smoothing factor (FSmoothing) defined by the user.
Let’s take an example of an antenna pattern to be smoothed, as shown in Figure 1.3 on page 28. Let DPeak-to-Null be 10 dB,
ASmoothing = 90 degrees, and FSmoothing = 0.5.
27
Figure 1.3: Vertical Antenna Pattern
Atoll first determines the peaks and nulls in the part of the pattern to be smoothed by verifying the slopes of the pattern curve
at each angle.
Figure 1.4: Peaks and Nulls in the Antenna Pattern
Peaks (P) and Nulls (N)
Angle (°) Attenuation (dB)

1 0.1
15 33.5
21 13.2
30 37.6
38 16.9
49 32.2
67 15.6
Then, Atoll verifies whether the difference of attenuation at a given angle is DPeak-to-Null less than the before and after it. This
comparison determines the nulls to be smoothed (NSmoothing).
Nulls to be smoothed (NSmoothing)
Angle (°) Attenuation (dB)

15 33.5
30 37.6
49 32.2
28
Once the nulls are known, Atoll applies the smoothing algorithm to all the attenuation values at all the angles between the
first peak, the null, and the last peak.
Smoothing Algorithm
For all nulls n  N Smoothing surrounded by two peaks P1 and P2 at angles  1 and  2 ,
  A 2 – A  1 
-   i –  1  
A i Smoothed = A i – F Smoothing A i –  A  +  ----------------------
1  2 – 1 
 
Where,
i is the angle in degrees from  1 to  2 incremented by 1 degree,
AAngle is the attenuation at any given angle which can be i,  1 or  2 , and
FSmoothing is the smoothing factor defined by the user.
1.3 Power Received From Secondary Antennas

When secondary antennas are installed on a transmitter, the signal level received from it is calculated as follows:
 
   G ant – m Tx G ant – i 
 P Tx   1 –
   X i  ----------------------
 L Tx P Tx  X i  -------------------- Tx
L Tx

i
 ------------------------------------------------------------------ + -
----------------------------------------
 L ant – m  az m el m  L ant – i  az i el i  
 Tx i Tx 
P rec = -------------------------------------------------------------------------------------------------------------------------------- (not in dB1)
L model
Where,
PTx is the transmitter power (Ppilot in UMTS HSPA and CDMA2000, PP-CCPCH in TD-SCDMA, PPreamble in WiMAX, and PDLRS in
LTE),
i is the secondary antenna index,
xi is the percentage of power dedicated to the secondary antenna, i,
G ant – m is the gain of the main antenna installed on the transmitter,

Tx
LTx are transmitter losses (LTx=Ltotal-DL),
G ant – i is the gain of the secondary antenna, i, installed on the transmitter,

Tx
Lmodel is the path loss calculated by the propagation model,
L ant – m  az m el m  is the attenuation due to main antenna pattern,

Tx
L ant – i  az i el i  is the attenuation due to pattern of the secondary antenna, i.

Tx
The definition of angles, az and el, depends on the used calculation method.
• Method 1 (must be indicated in an Atoll.ini file):
• azm: the difference between the receiver antenna azimuth and azimuth of the transmitter main antenna,
• elm: the difference between the receiver antenna tilt and tilt of the transmitter main antenna,
• azi : the difference between the receiver antenna azimuth and azimuth of the transmitter secondary antenna, i,
• eli : the difference between the receiver antenna tilt and tilt of the transmitter secondary antenna, i,
• Method 2 (default):
• azm : the receiver azimuth in the coordinate system of the transmitter main antenna,
• elm : the receiver tilt in the coordinate system of the transmitter main antenna,
• azi : the receiver azimuth in the coordinate system of the transmitter secondary antenna, i,
• eli : the receiver tilt in the coordinate system of the transmitter secondary antenna, i.
1. Formula cannot be directly calculated from components stated in dB and must be converted in linear values.
29
1.4 Transmitter Radio Equipment

Radio equipment such as TMA, feeder and BTS, are taken into account to evaluate:
• Total UL and DL losses of transmitter ( L total – UL L total – DL ) and transmitter noise figure  NF Tx  in UMTS HSPA,
CDMA2000 1xRTT 1xEV-DO, TD-SCDMA, WiMAX, and LTE documents,
• Transmitter total losses  L Total  in GSM GPRS EGPRS documents.
In Atoll, the transmitter-equipment pair is modelled a single entity. The entry to the BTS is considered the reference point
which is the location of the transmission/reception parameters.
Figure 1.5: Reference Point - Location of the Transmission/Reception parameters
• According to the book “Radio network planning and optimisation for UMTS” by Laiho
J., Wacker A., Novosad T., the noise figure corresponds to the loss in case of passive
components. Therefore, feeder noise figure is equal to the cable uplink losses.
UL
NF Feeder = L Feeder
• Loss and gain inputs specified in ATL documents must be positive values.
1.4.1 GSM Documents

Atoll calculates DL total losses as follows:
DL DL DL DL
L Total – DL = L TMA + L Feeder + L Misc + L BTS – Conf
Where,
DL
• L TMA is the TMA transmission loss.
DL DL DL DL DL DL
• L Feeder is the feeder transmission loss ( L Feeder = L Feeder  I Feeder + L Connector , where L Feeder , I Feeder and L Connector
are respectively the feeder loss per metre, the transmission feeder length in metre and the connector transmission
loss).
DL
• L Misc are the miscellaneous transmission losses.
DL
• L BTS – Conf are the losses due to BTS configuration (BTS property).
1.4.2 UMTS, CDMA2000, TD-SCDMA, WiMAX, and LTE Documents

As the reference point is the BTS entry, the transmitter noise figure corresponds to the BTS noise figure. Therefore, we have
NF TX = NF BTS . Where NF BTS is the BTS noise figure.
Uplink Total Losses
Atoll calculates total UL losses as follows:

UL UL UL UL
L Total – UL = L Misc + L Feeder + L BTS – Conf + NR Repeaters – G Ant – div – G TMA
Where,
UL
• L Misc are the miscellaneous reception losses (Transmitter property).
30
UL UL UL UL UL
• L Feeder are the feeder reception losses ( L Feeder = L Feeder  I Feeder + L Connector , where L Feeder , I Feeder and
UL
L Connector are respectively the feeder loss per metre (Feeder property), the reception feeder length in metre
(Transmitter property) and the connector reception losses.
UL
UL
• G Ant – div is the antenna diversity gain (Transmitter property). This gain does not exist in WiMAX and LTE documents.
• NR Repeaters is the noise rise at transmitter due to repeaters. This parameter is taken into account only if the
transmitter has active repeater(s). The noise rise at transmitter due to repeaters is calculated as follows:
 1 
NR Repeaters = 10  Log  1 +
  NIM
------------------
Rp 
r r
For each active repeater ( k ), Atoll calculates a noise injection margin ( NIM Rp ). This is the difference between the
k
donor transmitter noise figure ( NF TX ) and the repeater noise figure received at the donor.
Rp k TX – Rp k
NIM Rp = NF TX –  NF Rp + G amp – L 
r  k 
Where,
• NF Rp is the repeater noise figure,

k
Rp k
• G amp is the repeater amplification gain (repeater property),
TX – R p k
• L are the losses between the donor transmitter and the repeater (repeater property).
• For each active repeater ( k ), Atoll converts the noise injection margin ( NIM Rp ) to Watt. Then, it uses the values
k
to calculate the noise rise at the donor transmitter due to active repeaters ( NR Repeaters ).
• G TMA is the gain due to TMA, which is calculated as follows:
WithoutTMA WithTMA
G TMA = NF Composite – NF Composite
WithTMA WithoutTMA
Where NF Composite and NF Composite are the composite noise figures with and without TMA respectively.
Friis' equation is used to calculate the composite noise figure when there is a TMA.
 NF NF Feeder
--------------------------
NF BTS
-----------------

 -------------------
TMA
10 10 
– –
= 10  Log  10 + ---------------------------------- + -----------------------------------------------
WithTMA 10 10 1 10 1
NF Composite
 UL UL UL 
 G TMA
----------------
G TMA
---------------- ----------------------
G Feeder
 10 10 10 
10 10  10
WithoutTMA
And, NF Composite = NF BTS + NF Feeder
Where,
• NF Feeder is the feeder noise figure.

• NF TMA is the TMA noise figure.
• NF BTS is the BTS noise figure.
UL
• G TMA is the TMA reception gain.
UL UL UL
• G Feeder is the feeder UL gain G Feeder = – L Feeder .
UL UL UL UL UL
• L Feeder is the feeder reception loss ( L Feeder = L Feeder  I Feeder + L Connector , where L Feeder , I Feeder and
UL
L Connector are respectively the feeder loss per metre, the reception feeder length in metre and the connector
reception loss).
Downlink Total Losses
Atoll calculates total DL losses as follows.
31
DL DL DL DL
L Total – DL = L TMA + L Feeder + L Misc + L BTS – Conf
Where,
DL
• L TMA is the TMA transmission loss.
DL DL DL DL DL DL
• L Feeder is the feeder transmission loss ( L Feeder = L Feeder  I Feeder + L Connector , where L Feeder , I Feeder and L Connector
are respectively the feeder loss per metre, the transmission feeder length in metre and the connector transmission
losses).
DL
• L Misc are the miscellaneous transmission losses.
DL
1.5 Repeaters and Remote Antennas

A repeater receives, amplifies, and re-transmits the radiated or conducted RF carrier both in downlink and uplink. It has a
donor side and a server side. The donor side receives the signal from a donor (transmitter, repeater, or remote antenna), and
the server side amplifies and re-transmits the received signal. Repeaters increase the coverage area of their donors by re-
transmitting all the frequencies (TRXs in GSM, carriers in UMTS, CDMA2000 and TD-SCDMA, and channels in WiMAX and LTE
documents).
Donors and repeaters may be linked through:
• Air: User-defined or calculated propagation losses
• Microwave Links: User-defined link losses
• Optical Fibre Links: User-defined link losses
Remote antennas are antennas located far from the transmitters, at locations that would normally require long runs of feeder
cable. A remote antenna is connected to the base station with an optic fibre. Remote antennas allow you to ensure radio
coverage in an area without a new base station. In Atoll, remote antennas should be connected to base stations that do not
have any antennas. A remote antenna, as opposed to a repeater, does not have any equipment and therefore generates
neither amplification gain nor noise.
In UMTS, CDMA2000, TD-SCDMA, WiMAX, and LTE documents, Atoll calculates the signal level received from a repeater or a
remote antenna by determining the total downlink and uplink gains (described in "UMTS, CDMA2000, TD-SCDMA, WiMAX,
and LTE Documents" on page 32). In GSM documents, the received signal level from a repeater or a remote antenna is
calculated by determining the EIRP transmitted by the repeater or remote antenna (described in "GSM Documents" on
page 39).
The following sections describe how received signal levels, and the related downlink and uplink gains and EIRP, are calculated
from a repeater or remote antenna R with a donor D.
1.5.1 UMTS, CDMA2000, TD-SCDMA, WiMAX, and LTE Documents

1.5.1.1 Signal Level Calculation
The received signal level (dBm) on a carrier ic from a donor D at a pixel/mobile Mi via a repeater or remote antenna R (see
Figure 1.6 on page 34) is calculated as follows:
R – Mi Mi Mi Mi Mi
R D R R
C DL  ic  = P DL  ic  + G Total – L Path – M Shadowing – L Indoor + G –L – L Ant – L Body – L Misc – DL
If a pixel/mobile Mi receives signals from the donor D and its repeater R, the total signal
D R
strength is the sum of the two signals: C DL  ic  + C DL  ic 
The received signal level (dBm) from a pixel/mobile Mi at a donor D via a repeater or remote antenna R (see Figure 1.6 on
page 34) is calculated as follows:
Mi Mi R – Mi Mi Mi Mi Mi
R R
C UL = P UL + G Total – L Path – M Shadowing – L Indoor + G –L – L Ant – L Body – L Misc – UL
Here:
D
• P DL  ic  is the downlink transmission power of a donor D on carrier ic.
Mi
• P UL is the uplink transmission power of a pixel/mobile Mi.
32
R
• G Total is the total gain, user-defined or calculated as explained in "Total Gain Calculation" on page 34.
R–M
i
• L Path is the path loss (dB) calculated as follows:
R–M
i R
L Path = L Model + L Ant , with:
• L Model is the path loss calculated using a propagation model.

R
• L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the repeater or remote
antenna R.
• M Shadowing is the shadowing margin.
• L Indoor is the indoor loss.
Mi
• G is the terminal antenna gain for the pixel/mobile Mi.
Mi
• L is the terminal loss for the pixel/mobile Mi.
Mi
• L Ant is the terminal antenna attenuation (from antenna patterns) calculated for the pixel/mobile Mi (available in
WiMAX and LTE only).
Mi
For calculating the useful signal level from the best serving cell, L Ant is determined in the
direction (H,V) = (0,0) from the antenna patterns of the antenna used by Mi. For
Mi
calculating the interfering signal level from any interferer, L Ant is determined in the
direction of the interfering cell from the antenna patterns of the antenna used by Mi,
while the antenna is pointed towards Mi’s best serving cell.
Mi
• L Body is the body loss defined for the service used by the pixel/mobile Mi.
Mi Mi Mi Mi
L ,G , L Ant , and L Body are not used in all the calculations. For more information, see
the technology-specific chapters.
R
• L Misc – DL is the miscellaneous transmission losses defined for the repeater or remote antenna R.
R
• L Misc – UL is the miscellaneous reception losses defined for the repeater or remote antenna R.
33
Figure 1.6: UMTS, CDMA2000, TD-SCDMA, WiMAX, and LTE: Signal Level Calculation
1.5.1.2 Total Gain Calculation

The total gain is calculated from the donor transmitter reference point ( ) to the repeater or remote antenna reference point
( ) as follows:
Over-the-Air Repeaters
D D D–R R R R R R
G Total = – L Total – DL + G Ant – L Model + G Donor – Ant – LDonor + G Amp – LCov + G Cov – Ant
R
RX – Feeder TX – Feeder
Figure 1.7: Downlink Total Gain: Over-the-Air Repeaters
34
Here:
D
• L Total – DL are the total downlink losses of the donor D.
D
• G Ant is the gain of the antenna used at the donor D.
D–R
• L Model is the path loss between the donor D and the repeater or remote antenna R. This can be user-defined or
calculated using the selected propagation model. If you do not select a propagation model, the propagation losses
between the donor and the repeater or remote antenna are calculated using the ITU 526-5 propagation model.
R
• G Donor – Ant is the gain of the donor-side antenna used at the repeater or remote antenna R.
R
• LDonor are the donor-side reception feeder losses for the repeater or remote antenna R.
RX – Feeder
R
• G Amp is the amplifier gain of the repeater R. For remote antennas, this is 0.
R
• LCov are the coverage-side transmission feeder losses for the repeater or remote antenna R.
TX – Feeder
R
• G Cov – Ant is the gain of the coverage-side antenna used at the repeater or remote antenna R.
Secondary antennas are fully supported in the evaluation of the repeater gains.
Microwave Link Repeaters
D–R R R R
G Total = – L MW + G Amp – LCov + G Cov – Ant
R
TX – Feeder
Figure 1.8: Downlink Total Gain: Microwave Link Repeaters
Here:
D–R
• L MW are the user-defined microwave link losses between the donor D and the repeater or remote antenna R.
R
R
TX – Feeder
R
Optical Fibre Link Repeaters and Remote Antennas
D–R R R R
G Total = – L Fibre + G Amp – LCov + G Cov – Ant
R
TX – Feeder
35
Figure 1.9: Downlink Total Gain: Optical Fibre Link Repeaters or Remote Antennas
Here:
D–R
• L Fibre are the user-defined optical fibre link losses between the donor D and the repeater or remote antenna R.
R
R
TX – Feeder
R
Repeater Downlink Power Limitation
Atoll verifies that the downlink power after amplification is consistent with the repeater equipment limitation.
D R R R R
P DL  ic  + G Total  P Max + G Cov – Ant – LCov
TX – Feeder
Here:
D
• P DL  ic  is the downlink transmission power of a donor D on carrier ic. When the donor has more than one cell, Atoll
considers the highest power.
R
• G Total is the total downlink gain, user-defined or calculated as explained in "Total Gain Calculation" on page 34.
R
• P Max is the maximum downlink power allowed by the equipment.
R
TX – Feeder
R
1.5.1.3 Repeater Noise Figure

You can define and assign a repeater equipment to each repeater. In addition to the allowed ranges of gains and powers
allowed to each repeater, these equipment contain a noise figure which is applied to the repeater they are assigned to. This
noise figure has an impact on the donor total reception losses. For information, see "Transmitter Radio Equipment" on
page 30.
1.5.1.4 Appendix: Carrier Power and Interference Calculation

This section explains how Atoll calculates the received carrier power and interference when a transmitter has a connected
repeater.
A mobile receiver receives signal from the donor transmitter as well as its repeater. Similarly, the signal from the mobile is
received at the donor transmitter as well as its repeater. In practice, when a mobile receiver is in the vicinity of the donor
transmitter, the signal to/from the repeater would be very weak due to high pathloss between the repeater and the mobile
receiver. Similarly, when the mobile receiver is located in the vicinity of the repeater, the signal to/from the donor transmitter
would be very weak due to the same reason.
Atoll does not differentiate between the mobile receiver being in the transmitter coverage area or being in its repeater
coverage area. Atoll adds the signals received from the donor transmitter and its repeater to generate a combined pathloss
matrix that is associated with the donor transmitter and includes the effect of its repeater.
36
Calculation of Total Path Loss
The total pathloss, L Total , is calculated by computing a downlink budget. If we take the case of a CDMA project, without
considering any shadowing margin or indoor loss, the power received from the donor transmitter, Txd on a carrier ic, at the
mobile receiver can be stated as (for a link over the air):
D D
D  P Pilot  ic   G Ant 
P Rec  ic  = -------------------------------------------
-
D D – Mi
 L Total –DL  L Path 
Where,
D
P Rec  ic  is the carrier power received at the receiver from the donor transmitter on a carrier ic (in W)
D
P Pilot  ic  is the pilot power of the donor transmitter on the carrier ic (in W)
D
G Ant is the donor transmitter antenna gain.
D
L Total – DL is the transmission feeder loss of the donor transmitter.
D – Mi
L Path is the path loss between the donor transmitter and the mobile receiver.
Similarly, the power received at the mobile receiver from the repeater R is:
D R
R  P Pilot  ic   G Total 
P Rec  ic  = --------------------------------------------
-
R – Mi
L Path
Where,
R
P Rec  ic  is the carrier power received at the mobile receiver from the repeater on a carrier ic (in W)
D
P Pilot  ic  is the pilot power of the donor transmitter on the carrier ic (in W)
R
G Total is the total gain of repeater linked to a donor transmitter with an air link.
R – Mi
L Path is the path loss between the repeater and the mobile receiver
So, the total carrier power received at the mobile receiver is:
D R
D–R R R D  G Ant G Total
P Rec  ic  = P Rec  ic  + P Rec  ic  = P Pilot  ic    --------------------------------------------- -
- + -------------
– Mi R – Mi
  L DTotal – DL  L DPath  L Path 
Since,
D D
P Pilot  ic   G Ant
L Total = ------------------------------------------------
D D–R
L Total – DL  P Rec  ic 
Therefore,
D D
P Pilot  ic   G Ant
L Total = -----------------------------------------------------------------------------------------------------------------------------
-
D R
D D  G Ant G Total
L Total – DL  P Pilot  ic    ---------------------------------------------- + --------------
– Mi R – Mi
Hence,
D
G Ant
L Total = ---------------------------------------------------------------------------------------------------
D R
D  G Ant G Total
L Total – DL   --------------------------------------------- - + ------------- -
– Mi R – Mi
This total path loss depends on the location of the mobile receiver in realistic network scenarios. As a mobile in the donor
transmitter/repeater coverage area is likely to be far from the repeater/donor transmitter coverage area, the respective
pathloss value will be very large. This implies that we can study the two cases separately without influencing the results much.
• Case 1: Receiver in Donor Transmitter Coverage Area
37
R
R – Mi G Total
L Path is likely to be very high, so the term -------------
R – Mi
- can be ignored. This implies that:
L Path
D – Mi
L Total = L Path
Considering this total pathloss value, the total received power in the uplink and in the downlink can be stated as:
D D D D
D  P Pilot  ic   G Ant   P Pilot  ic   G Ant 
P Rec – DL  ic  = -------------------------------------------
- = ---------------------------------------------
-
D D D – Mi
 L Total – DL  L Total   L Total – DL  L Path 
Mi D Mi D
D  P Output  ic   G Ant   P Output  ic   G Ant 
P Rec – UL  ic  = ----------------------------------------------
- = ----------------------------------------------
-
D D D – Mi
 L Total – UL  L Total   L Total – UL  L Path 
Where,
Mi
P Output  ic  is the transmitted power from the mobile terminal on the carrier ic (in W)
D
L Total – UL is the reception feeder loss of the transmitter
• Case 2: Receiver in Repeater Coverage Area

D
D – Mi G Ant
L Path is likely to be very high, so the term ---------------------------------------------
D D – Mi
- can be ignored. This implies that:
 L Total – DL  L Path 
D D
G Ant G Ant
L Total = ---------------------------------------------------
- = -----------------------------------------
R R R
 D  G Total  G Total  L Total – DL
 L Total – DL   ------------- -  -----------------------------------------
– Mi R – Mi
  L RPath  L Path
D D D R
D  P Pilot  ic   G Ant   P Pilot  ic   G Total 
P Rec – DL  ic  = -------------------------------------------
- = --------------------------------------------
-
D R – Mi
 L Total – DL  L Total   L Path 
Mi D Mi R D
D  P Output  ic   G Ant   P Output  ic   G Total  L Total – DL
-  ----------------------
- = -------------------------------------------------
P Rec – UL = ----------------------------------------------
D R – Mi D
 L Total – UL  L Total   L Path  L Total – UL
Where,
Mi
P Output  ic  is the transmitted power from the mobile terminal (in W)
D
L Total – UL is the reception feeder loss of the transmitter
Calculation of Eb/Nt Uplink
In the uplink, the quality level at the transmitter on a traffic channel is:
E b
 ---- C W
- = ---  -----
 N t UL I R
Where,
C is the carrier power received from the mobile terminal (in W)
I is the total interference (in W)
W is the spreading bandwidth (Hz)
R is the effective service throughput in the uplink (bits/s)
(W/R is the service processing gain in the uplink)
C and I are both evaluated at the same reference point, which is the entry of BTS using the following formulas.
Mi D
P Output  G Ant
C = P Total – UL = ---------------------------------------
D
L Total – UL  L Total
I = I Total + N 0
38
Where,
I Total is the sum of the signals received from mobile terminals inside the same cell and those outside (in W)
N 0 is the transmitter equipment thermal noise (in W)
Therefore, for each mobile terminal Mi,

Mi D
 P Output  G Ant 
I Total =   ---------------------------------------
 L DTotal – UL  L Mi Total

Mi
And,
D
N 0 = NF  K  T  W
Where,
D
NF is the noise figure of the transmitter equipment at the reference point, i.e. the entry of the BTS
K is Boltzman constant
T is the ambient temperature (in K)
Hence
BTS
N 0 = NF KTW
1.5.2 GSM Documents

1.5.2.1 Signal Level Calculation
The received signal level (dBm) on a TRX type tt from a donor D at a pixel/mobile Mi via a repeater or remote antenna R (see
Figure 1.10 on page 40) is calculated as follows:
R – Mi Mi Mi
R R R
C DL  tt  = EIRP DL  tt  – P  tt  – L Path – M Shadowing – L Indoor + G –L – L Misc – DL
If a pixel/mobile Mi receives signals from the donor D and its repeater R, the total signal
D R
strength is the sum of the two signals: C DL  tt  + C DL  tt 
Here:
R
• EIRP DL  tt  is the effective isotropic radiated power of the repeater or remote antenna R on the TRX type tt. It can be
user-defined or calculated as explained in "EIRP Calculation" on page 40.
• the downlink transmission power of a donor D on carrier ic.
• P  tt  is the power offset defined for the TRX type tt.
R – Mi
• L Path is the path loss (dB) calculated as follows:
R – Mi
R
L Path = L Model + L Ant , with:
• L Model is the path loss calculated using a propagation model.

R
• L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the repeater or remote
antenna R.
• M Shadowing is the shadowing margin.
• L Indoor is the indoor loss.
Mi
• G is the terminal antenna gain for the pixel/mobile Mi.
Mi
• L is the terminal loss for the pixel/mobile Mi.
R
• L Misc – DL is the miscellaneous transmission losses defined for the repeater or remote antenna R.
39
s
Figure 1.10: GSM: Signal Level Calculation
1.5.2.2 EIRP Calculation

D
The EIRP of a repeater or remote antenna R is calculated at the repeater or remote antenna reference point ( ) w. r. t. P DL
at the donor reference point ( ) as follows:
Over-the-Air Repeaters
R D D D D–R R R R R R
EIRP DL  tt  = P DL – L Total – DL + G Ant – L Model + G Donor – Ant – LDonor + G Amp – LCov + G Cov – Ant
RX – Feeder TX – Feeder
Figure 1.11: EIRP: Over-the-Air Repeaters
Here:
D
• P DL is the downlink transmission power of the donor D.
D
• L Total – DL are the total downlink losses of the donor D.
40
D
• G Ant is the gain of the antenna used at the donor D.
D–R
• L Model is the path loss between the donor D and the repeater or remote antenna R. This can be user-defined or
calculated using the selected propagation model. If you do not select a propagation model, the propagation losses
between the donor and the repeater or remote antenna are calculated using the ITU 526-5 propagation model.
R
• G Donor – Ant is the gain of the donor-side antenna used at the repeater or remote antenna R.
R
• LDonor are the donor-side reception feeder losses for the repeater or remote antenna R.
RX – Feeder
R
R
TX – Feeder
R
Secondary antennas are fully supported in the evaluation of the repeater gains.
Microwave Link Repeaters
D D–R R R R
EIRP DL  tt  = P DL – L MW + G Amp – LCov + G Cov – Ant
R
TX – Feeder
Figure 1.12: Downlink Total Gain: Microwave Link Repeaters
Here:
D
D–R
• L MW are the user-defined microwave link losses between the donor D and the repeater or remote antenna R.
R
R
TX – Feeder
R
Optical Fibre Link Repeaters and Remote Antennas
D D–R R R R
EIRP DL  tt  = P DL – L Fibre + G Amp – LCov + G Cov – Ant
R
TX – Feeder
41
Figure 1.13: Downlink Total Gain: Optical Fibre Link Repeaters or Remote Antennas
Here:
D
D–R
• L Fibre are the user-defined optical fibre link losses between the donor D and the repeater or remote antenna R.
R
R
TX – Feeder
R
Repeater Downlink Power Limitation
Atoll verifies that the EIRP after amplification is consistent with the repeater equipment limitation.
R R R R
EIRP DL  tt   P Max + G Cov – Ant – LCov
TX – Feeder
Here:
R
• EIRP DL  tt  is the effective isotropic radiated power of the repeater R on the TRX type tt.
R
• P Max is the maximum downlink power allowed by the equipment.
R
TX – Feeder
R
1.5.3 Donor-side Parameter Calculations

1.5.3.1 Azimuth
This is the angle at which the donor antenna is situated with respect to the North at the repeater or remote antenna. This
angle is measured clock-wise as shown in the figure below. It is the absolute horizontal angle at which the donor-side antenna
of the repeater should be pointed in order to be aligned with the donor antenna.
Figure 1.14: Angle from North (Azimuth)
1.5.3.2 Mechanical Downtilt

This is the tilt angle for the repeater’s donor-side antenna, which ensures that it points towards the donor antenna in the
vertical plane. As a general rule, downtilt angles are considered positive and uptilt angles negative.
42
Figure 1.15: Positive/Negative Mechanical Downtilt
Since this parameter depends on the difference of heights/altitudes between the donor transmitter and the repeater, it can
be automatically calculated in the repeater’s Donor side properties. If the height/altitude of the antenna is modified, the
corresponding tilt angle can be found out and applied using the Calculate button.
Example
Figure 1.16: Tilt Angle Computation
The tilt angle repeater’s donor-side antenna in the above figure would be:
R D
R  H Donor – Ant – H Ant
T Donor – Ant = atan  -------------------------------------------
D–R
-
 D 
As obvious, this angle will be negative for uptilts and positive for downtilts of the antenna.
Here:
R
• H Donor – Ant is the height of the donor-side antenna of the repeater or remote antenna R.
D
• H Ant is the height of the antenna of the donor D.
D–R
• D is the distance between the antenna of the donor D and the antenna of the repeater or remote antenna R.
1.6 Beamforming Smart Antenna Models

Adaptive antenna systems use more than one antenna elements, along with smart signal processing, to locate and track
various types of signals, to dynamically minimize interference, and maximize useful signal reception. The signal processor
dynamically applies weights to each element of the adaptive antenna system to create array patterns in real-time.
Beamforming smart antennas dynamically create antenna patterns with a main beam pointed in the direction of the user
being served, i.e., the useful signal. Adaptive algorithms can also be used in order to minimize the interference received by
the cells. These algorithms are based on optimization methods such as the minimum mean square error method.
The following beamforming smart antenna models are available in Atoll. These smart antenna models support linear adaptive
array systems, such as the one shown in Figure 1.17 on page 44.
• Optimum Beamformer: The Optimum Beamformer smart antenna model performs dynamic beamforming in
downlink as explained in "Downlink Beamforming" on page 46, and beamforming and interference cancellation in
uplink using the minimum mean square error algorithm as explained in "Uplink Beamforming and Interference
Cancellation (MMSE)" on page 49. Smart antenna results are later on used in coverage prediction calculations.
• Conventional Beamformer: The Conventional Beamformer smart antenna model performs dynamic beamforming in
downlink and uplink as explained in "Downlink Beamforming" on page 46 and "Uplink Beamforming" on page 48,
respectively. Smart antenna results are later on used in coverage prediction calculations.
43
Figure 1.17: Linear Adaptive Antenna Array
In the following explanations, we assume:

• There are a total of E SA elements in the adaptive antenna system.
•  is the angle of arrival for the useful signal.
•  is the angle at which we want to calculate the smart antenna gain.
• d is the distance between two adjacent antenna elements.
1.6.1 Definitions and Formulas

The tables in the following subsections list the parameters and formulas used in beamforming smart antenna models.
1.6.1.1 Definitions
Name Value Unit Description
E SA Smart antenna model parameter None Number of smart antenna elements
 Calculation parameter Degrees Angle of arrival for the useful signal

Angle at which the smart antenna
 Calculation parameter Degrees
effect is calculated
 Distance between two adjacent
d --- , where  is the wavelength of the signal m
2 antenna elements
1.6.1.2 Downlink Beamforming

gn    Smart antenna model parameter None Gain of a single element
2 2 2 T
j  ------  d  sin  j  ------  2d  sin  j  ------   E SA – 1 d  sin 
   Steering vector for the direction of 
S 1 e e  ... e None
2
– j  ------  nd  sin 

e
wn None Complex smart antenna weight
with d = 
– j    n  sin 
e ---
2
H Array correlation matrix for a given

R S  S None
user direction 
Smart antenna gain in any direction
G SA    H H H
g n     S   R   S  = g n     S   S   S   S  = g n     E SA
2
None

44

J
Downlink array correlation matrix for
Rk  pj  Rj None
iteration k
j=1
K
Average downlink array correlation
--1- 
R Avg
K  Rk None matrix over a simulation (K
iterations)
k=1
1.6.1.3 Uplink Beamforming

S Vector of ESA complex weights for

w -----------
- None
E SA the conventional beamformer
 pj  Sj  Sj
2 H
RN Rn + RI = n I+ None Total noise correlation matrix
j=1
2
Rn n  I None Thermal noise correlation matrix
H
RI None Interference correlation matrix
j=1
H
PN w  RN  w W Total uplink noise power
H H Total power received from the

P p   w  S   S   w = p   E SA W
served user
P p   E SA
CINR UL ------ = -------------------------- None
C/(I+N) in the uplink
PN H
w  RN  w (WiMAX)
P p   E SA
SA
Q UL ------ = -------------------------- None
Signal quality in the uplink
PN H
w  RN  w (TD-SCDMA)
Uplink smart antenna beamforming

G SA E SA None gain in the direction of the served
user
K
--1- 
RN
Avg K  RN k W Average noise correlation matrix
k=1
H 2
I UL    w  RN  w – n W Uplink interference
Avg
2
I UL    +  n Angular distribution of uplink noise
NR UL    --------------------------- None
n
2 rise
1.6.1.4 Uplink Beamforming and Interference Cancellation (MMSE)

–1 Vector of ESA complex weights for

ŵ   RN  S None
the optimum beamformer
E SA
 -----------------------------
- None MMSE optimization constant
H –1
S  RN  S
45

J
2 H
RN Rn + RI = n I+ None Total noise correlation matrix
j=1
2
Rn n  I None Thermal noise correlation matrix
H
RI None Interference correlation matrix
j=1
2 H –1 Total uplink noise power (optimum

P̂ N   S  RN  S W
beamformer)
2 H –1 2 Total power received from the
P̂  p     S  RN  S  W
served user (optimum beamformer)
P P̂ H –1 C/(I+N) in the uplink
CINR UL ------ = ------ = p   S   R N  S  None
PN P̂ N (WiMAX)
P P̂ H –1 Signal quality in the uplink

SA
Q UL ------ = ------ = p   S   R N  S  None
PN P̂ N (TD-SCDMA)
Uplink smart antenna beamforming

H
G SA S   I  S  = E SA None gain in the direction of the served
user
K
Average inverse noise correlation
--1- 
–1
 RN
–1
RN W
Avg K k matrix
k=1
E SA 2
I UL    - – n
------------------------------------ W Uplink interference
H –1
S  RN  S
Avg
2
I UL    +  n Angular distribution of uplink noise
NR UL    --------------------------- None
n
2 rise
1.6.2 Downlink Beamforming
Figure 1.18: Downlink Beamforming
Beamforming dynamically creates a beam towards the served user. The smart antenna processor applies complex weights,
w n , to each antenna element in order to form a beam towards the served user. The magnitude of these complex weights is
46
set to 1. The beamforming is performed using only the phase of the complex weights. The steering vector, S  , representing
the complex weights for forming a beam towards the served user, i.e., at the angle of arrival  is given by:
2 2 2 T
j  ------  d  sin  j  ------  2d  sin  j  ------   E – 1 d  sin 
   SA
S  = 1 e e  ... e
Where T represents the transpose of a matrix.
Therefore, the complex weight at any nth antenna element can be given by:
2
– j  ------  nd  sin 

wn = e
In Atoll, d = 
– j    n  sin 
--- , therefore, w n = e .
2
The smart antenna gain in any direction  can be given by:

H
G SA    = g n     S   R   S 
Where H represents the Hilbert transform, which is the complex conjugate transpose of a matrix, g n    is the gain of the nth
antenna element in the direction  , and R  is the array correlation matrix for a given user direction  , given by:
H
R = S  S
For the direction of the served user, i.e.,  , the smart antenna gain is calculated as follows:
H H H 2
G SA    = g n     S   R   S  = g n     S   S   S   S  = g n     E SA
The smart antenna gain includes the gain of the beamforming as well as the gain of power combination.
The smart antenna gain in dB will be G SA    = 10  Log  G SA     .
The smart antenna is able to form the beam only in the horizontal plane, therefore, the vertical pattern is assumed to remain
the same.
Power Combining Gain
Cell transmission power is fed to each antenna element of the smart antenna system. Since each element transmits the same
input power, this results in a gain due to power combination, i.e., the powers fed to each antenna element are combined for
transmission.
Additional Processing in Monte Carlo Simulations
During Monte Carlo simulations, Atoll calculates the smart antenna gains (array correlation matrix R  ) for each served mobile
in a cell’s coverage area in each iteration. The sum of these array correlation matrices for all the users served in one iteration
k is calculated as follows:
J
Rk =  pj  Rj
j=1
Where R k for any cell is the downlink array correlation matrix for iteration k, J is the number of served mobiles during the
iteration, p j is the EIRP transmitted towards the mobile j, and R j is the array correlation matrix for the mobile j.
Atoll calculates a moving average of the array correlation matrices calculated in each iteration. At the end of a simulation with
K iterations, the average downlink array correlation matrix for any cell is given by:
K
1
R Avg = --- 
K  Rk
k=1
47
1.6.3 Uplink Beamforming
Figure 1.19: Uplink Beamforming
Let w represent the vector of ESA complex weights for the beamformer. w is given by:
S
w = -----------
-
E SA
Where S  is the steering vector in the direction of the served user,  .
The total noise received in the uplink, i.e., interference and thermal noise, is stored in a total noise correlation matrix, R N .
The total noise correlation matrix is the sum of the thermal noise correlation matrix R n , and the interference correlation
matrix R I , given by:
2 H
RN = Rn + RI = n I+
j=1
2 H
Where R n =  n  I and R I =
j=1
2
 n is the thermal noise power. I is the identity matrix. p j is the power received by one element of the smart antenna from
the jth interfering mobile. S j is the steering vector in the direction of the jth interfering mobile,  . J is the total number of
interfering mobiles.
The total noise power, including thermal noise and interference from all uplink interferers, received by a cell is given by:
H
PN = w  RN  w
And, the total power received from the served user is given by:
H H
P  = p   w  S   S   w = p   E SA
Where p  is the power received by one element of the smart antenna from the served user.
In TD-SCDMA, the uplink signal quality is calculated by:

P p   E SA
Q UL = ------ = --------------------------
SA
PN H
w  RN  w
In WiMAX, the C/(I+N) in the uplink is then calculated by:

P p   E SA
CINR UL = ------ = --------------------------
PN H
w  RN  w
48
From the above equation, we can determine the uplink smart antenna beamforming gain in the direction of the served user,
which equals the number of smart antenna elements, i.e., G SA = E SA .
The noise correlation matrix R N for each iteration k includes the effect of the matrix calculated for the previous iteration. The
result is the angular distribution of the uplink load (TD-SCDMA) or the uplink noise rise (WiMAX), which is calculated from the
noise correlation matrix obtained at the end of the last iteration of a Monte Carlo simulation. This angular distribution of the
uplink load (TD-SCDMA) or the uplink noise rise (WiMAX) can be stored in the Cells table. The average of the noise correlation
matrices is calculated as follows:
K
1
RN
Avg
= --- 
K  RN k
k=1
Where R N is the average of the noise correlation matrices of all the iterations from k = 1 to K, and R N is the noise
Avg k
correlation matrix of the kth iteration.

The interference can be isolated from the thermal noise and can be calculated for any direction using the formula.
H 2
I UL    = w  R N  w – n
Avg
Where I UL    is the interfering signal in the direction  , ESA is the number of smart antenna elements, S  is the steering
2
vector in the direction  , and  n is the thermal noise power.
In TD-SCDMA, the uplink load is calculated from the average noise correlation matrix. In WiMAX, the angular distribution of
the uplink noise rise is given by:
2
I UL    +  n
NRUL    = ---------------------------
2
n
1.6.4 Uplink Beamforming and Interference Cancellation (MMSE)

The optimum beamformer uses the Minimum Mean Square Error algorithm in the uplink in order to cancel interference. The
Minimum Mean Square Error algorithm optimizes the useful signal as well as maximizes the signal quality.
A simple null steering beamformer can cancel the interference from the most interfering E SA – 1 interfering mobiles. The
optimum beamforming method used in Atoll overcomes this limitation. It calculates the optimum smart antenna weights
using the knowledge of directions and power levels of interference. These weights do not try to fully cancel E SA – 1
interference signals, but rather try to reduce the overall received interference as much as possible.
Figure 1.20: Uplink Adaptive Algorithm
Let ŵ represent the vector of ESA complex weights for the beamformer. ŵ is given by:
49
–1
ŵ =    R N  S 
Where S  is the steering vector in the direction of the served user,  .   , which is a constant value for a given useful signal
that optimizes the beamformer weights. It is given by the equation:
E SA
  = -----------------------------
-
H –1
S  RN  S
–1
R N is the inverse of the total noise correlation matrix. The total noise correlation matrix is the sum of the thermal noise
correlation matrix R n , and the interference correlation matrix R I , given by:
2 H
RN = Rn + RI = n I+
j=1
2 H
Where R n = n  I and R I =
j=1
2
 n is the thermal noise power. I is the identity matrix. p j is the power received by one element of the smart antenna from
the jth interfering mobile. S j is the steering vector in the direction of the jth interfering mobile,  . J is the total number of
interfering mobiles.
The total noise power, including thermal noise and interference from all uplink interferers, received by a cell is given by:
2 H –1
P̂ N =    S   R N  S 
And, the total power received from the served user is given by:
2 H –1 2
P̂  = p       S   R N  S  
Where p  is the power received by one element of the smart antenna from the served user.
In TD-SCDMA, the uplink signal quality is calculated by:
P̂ H –1
Q UL = ------ = p   S   R N  S 
SA
P̂ N
In WiMAX, the C/(I+N) in the uplink is then calculated by:
P̂ H –1
CINR UL = ------ = p   S   R N  S 
P̂ N
From the above equation, we can determine the uplink smart antenna beamforming gain in the direction of the served user.
RSCP TCH – UL (TD-SCDMA) or C UL (WiMAX) can be calculated from the above equation by considering the interference and
–1
noise to be null, i.e., R N = I . This gives:
H
In TD-SCDMA, RSCP TCH – UL = p   S   I  S  = p   E SA
H
In WiMAX, C UL = p   S   I  S  = p   E SA
From the above equation, the uplink smart antenna beamforming gain equals the number of smart antenna elements, i.e.,
G SA = E SA .
–1
The inverse noise correlation matrix R N for each iteration k includes the effect of the matrix calculated for the previous
iteration. Hence, Atoll is able to calculate an average of the smart antenna interference-cancellation effect. The result is the
angular distribution of the uplink load (TD-SCDMA) or the uplink noise rise (WiMAX), which is calculated from the inverse of
the noise correlation matrix obtained at the end of the last iteration of a Monte Carlo simulation. This angular distribution of
the uplink load (TD-SCDMA) or the uplink noise rise (WiMAX) can be stored in the Cells table. The average of the inverse noise
correlation matrices is calculated as follows:
50
K
1
 RN
–1 –1
RN = --- 
Avg K k
k=1
–1 –1
Where R N is the average of the inverse noise correlation matrices of all the iterations from k = 1 to K, and R N is the
Avg k
inverse noise correlation matrix of the kth iteration.

The interference can be isolated from the thermal noise and can be calculated for any direction using the formula.
E SA 2
I UL    = ------------------------------------
- – n
H –1
S  RN  S
Avg
Where I UL    is the interfering signal in the direction  , ESA is the number of smart antenna elements, S  is the steering
2
vector in the direction  , and  n is the thermal noise power.
In TD-SCDMA, the uplink load is calculated from the average inverse noise correlation matrix. In WiMAX, the angular
distribution of the uplink noise rise is given by:
2
I UL    +  n
NRUL    = ---------------------------
2
n
1.7 Grid-of-Beams Smart Antenna Model

A grid-of-beams smart antenna, called GOB, consists of more than one directional antenna pattern (beam) in different
directions. Each beam of a GOB has a different azimuth so that the GOB as a whole covers an entire sector. During the
simulations, Atoll determines the most suitable beam from the GOB for each user served by the smart antenna. The most
suitable beam (best beam) is the one which provides the highest gain towards the served user:
BeamBest = Beam
H V
Max  G Beam – L Beam – L Beam
H V
Where G Beam , L Beam , and L Beam are the gains, horizontal, and vertical attenuations of the beams of the GOB. In words, the
best beam is the one among all the beams of a GOB that has the highest difference between gain, and horizontal and vertical
SA SA SA SA
attenuations. The gains and losses of the GOB ( G DL , G UL , L DL , and L UL ) are determined from the selected best beam.
The following example shows how Atoll calculates the GOB gains and losses.
Example:
Let us assume a GOB with 5 beams that have the same vertical patterns, and whose horizontal patterns are pointed towards
different directions as shown in the figure below:
51
Figure 1.21: Grid Of Beams Modelling
Let us assume that all the beams and the main antenna have the same 18 dBi gain, and the vertical attenuation at the user
location is 15 dB, which is also the same for all the beams because we assume that the vertical patterns are the same.
If the user is located at  = 70 azimuth, as shown in the figure below, Atoll determines the best beam, which has the
highest gain towards  , as follows:
Horizontal Vertical H V Total Gain

Beam Gain (dBi) G Beam – L Beam – L Beam
Attenuation (dB) Attenuation (dB) (dB)
0° 18 60 15 18 - 60 - 15 -57
30° 18 60 15 18 - 60 - 15 -57
60° 18 2.21 15 18 - 2.21- 15 0.79
-30° 18 60 15 18 - 60 - 15 -57
-60° 18 60 15 18 - 60 - 15 -57
Transmitter
Centre of the pixel where the served user is

located
Angle between the user and the transmitter

azimuth
Figure 1.22: GOB Modelling - Determination of the Best Beam
In our example, the total gain of the beam at 60° is the highest. Therefore this beam is selected as the best beam.
If this beam has been selected in the downlink,
SA SA H V
G DL = 18 dB and L DL = L Beam + L Beam = 17.21 dB
If this beam has been selected in the uplink,

SA SA H V
G UL = 18 dB and L UL = L Beam + L Beam = 17.21 dB
52
1.8 Adaptive Beam Smart Antenna Model

An adaptive beam smart antenna is capable of steering a given antenna pattern towards the direction of the served signal. In
Atoll, this is modelled using a single antenna pattern, called a beam because of its highly directional shape. During the
simulations, this adaptive beam is oriented in the direction of each served user in order to model the effect of the smart
antenna.
SA SA SA
The adaptive beam gains ( G DL and G UL ) are the antenna gains defined for the beam, and the adaptive beam losses ( L DL and
SA H V
L UL ) are the horizontal and vertical pattern attenuations L Beam + L Beam towards the user direction.
The following example shows how Atoll calculates the adaptive beam gains and losses.
Example:
Let us assume an adaptive beam smart antenna selected for a transmitter along with a main antenna. Let us assume that the
adaptive beam and the main antenna have the same 18 dBi gain, and the vertical attenuation at the user location is 15 dB.
If the user is located at  = 60 azimuth, as shown in the figure below:
Transmitter

located

azimuth
Figure 1.23: Adaptive Beam Modelling - Determination of the Best Beam
If the adaptive beam smart antenna is selected in the downlink, the gain and losses of the adaptive beam at  are:
SA SA H V
G DL = 18 dB and L DL = L Beam + L Beam = 15 dB
If the adaptive beam smart antenna is selected in the uplink, the gain and losses of the adaptive beam at  are:
SA SA H V
G UL = 18 dB and L UL = L Beam + L Beam = 15 dB
H
In fact, as the ideal beam steering algorithm steers the beam towards the served user, L Beam = 0 . These values are used in
interference calculation to determine the downlink interfering signal due to transmission towards the served user, as well as
for calculating the uplink interfering signals received at transmitter when decoding signal received from the served user.
1.9 Statistical Smart Antenna Gain Model

A statistical modelling approach is also available in Atoll which can be used to model the effect of smart antennas through C/
I gains. You can create smart antenna equipment in Atoll based on the statistical approach by providing C/I gains and their
cumulative probabilities for different spreading angles,  Spread .
You can assign a spreading angle to each clutter class in your document. Atoll reads the clutter class in which the served user
is located to determine the spreading angle. Different clutter types have different spreading effects on the propagation of
radio waves. Urban and dense urban clutter types introduce more multipath and spread the signal at a wider angle than an
open or rural clutte type.
Once you have assigned the spreading angles to clutter classes, you can enter the C/I gains and their cumulative probabilities
for each spreading angle, in the smart antenna equipment based on the statistical model. For each smart antenna equipment
SA
based on statistical modelling, you can set a probability threshold, TProb .
To find the smart antenna gain, Atoll determines the clutter class of the served user, it reads the spreading angle from the
clutter class properties, it reads the probability threshold from the smart antenna properties, and reads the smart antenna C/
SA
I gain defined for the Probability = 1 – TProb corresponding to the spreading angle.
53
The following example shows how Atoll calculates the statistical C/I gains and losses.
Example:
Let us assume that the served user is located at a an urban clutter class with  Spread = 10 . The smart antenna equipment
SA SA
has TProb = 80 % . Atoll will read the smart antenna C/I gain G for Prob = 20 % . If a gain for the exact probability
value of 20% is not defined, Atoll linearly interpolates the gain value from the two surrounding values.
SA SA SA
If G Prob = 19%
= 4.6298 dB and G Prob = 20.4%
= 4.7196 dB , then G Prob = 20%
= 4.6941 dB
The smart antenna gains are the same for uplink and downlink. Their are no losses for this type of smart antenna equipment.
Negative values of C/I gains are considered as losses.
54
Chapter 2
Radio Propagation

• "Path Loss Calculation Prerequisites" on page 57
• "List of Default Propagation Models" on page 62
• "Okumura-Hata and Cost-Hata Propagation Models" on
page 63
• "ITU 529-3 Propagation Model" on page 64
• "Standard Propagation Model (SPM)" on page 65
• "WLL Propagation Model" on page 75
• "ITU-R P.526-5 Propagation Model" on page 76
• "Erceg-Greenstein (SUI) Propagation Model" on page 78
• "Sakagami Extended Propagation Model" on page 84
• "Free Space Loss" on page 86
• "Diffraction" on page 86
• "Shadow Fading Model" on page 90
• "Path Loss Matrices" on page 103
• "File Formats" on page 107
© Forsk 2015
56
AT330_TRR_E1 Chapter 2: Radio Propagation
2 Radio Propagation
Path loss calculations are carried out between a transmitter and a receiver using propagation models and other calculations
related to radio wave propagation such as diffraction and shadow fading. Propagation models are mathematical
representations of the average loss in signal strength over distance. Diffraction loss and shadow fading margins are added to
this average loss in order to get more precise path loss values.
Path loss matrices are calculated for each transmitter and their results used in other calculations (coverage predictions, Monte
Carlo simulations, point analysis, etc.). The method of calculation may differ depending on the analysis being performed:
Receiver Profile
Analysis type Calculation Result
position extraction
Centre of each bin
Based on path loss One value for the bin’s
Coverage predictions inside the calculation
matrices Radiala surface area
area
Different values inside
Point analysis (Profile) Anywhere Real-time Systematic
a calculation bin
Anywhere inside the Based on path loss One value for the bin’s
Point analysis (other)
calculation areas matrices Radiala surface area
Monte Carlo Based on path loss One value at the
simulations
Mobile coordinates
matrices Radiala mobile location
One value at the
Subscriber lists Subscriber coordinates Real-time Radiala subscriber location
a. With the Standard Propagation Model, you can choose between radial or systematic.
This chapter describes the various propagation models available in Atoll, and other radio wave propagation phenomena such
as diffraction and shadow fading.
2.1 Path Loss Calculation Prerequisites

2.1.1 Ground Altitude Determination
Atoll determines reception and transmission site altitude from Digital Terrain Model map. The method used to evaluate site
altitude is based on a bilinear interpolation.
DTM files provide altitude value z (in metre) on evenly spaced points. Abscissa and ordinate axes are respectively oriented in
right and downwards directions. Space between points is defined by pixel size P (in metre). Pixel size must be the same in both
directions. The first point given in the file corresponds to the centre of the top-left pixel of the map (northwest point geo-
referenced by Atoll).
Figure 2.1: Digital Terrain Model
Four points (hence, four altitude values) are necessary to describe a “bin”; these points are bin vertices. Therefore, a DTM file
that contains N x N bins requires N2 points (altitude values).
57
Chapter 2: Radio Propagation ©Forsk 2015
Figure 2.2: Schematic view of a DTM file
• In Atoll, DEM (Digital Elevation Model) is the same as Digital Terrain Model (DTM). In
literature, DEM and DTM do not always have the same meaning. By definition, DEM
refers to the altitude above sea level including ground and clutter, while DTM refers
to the ground altitude above sea level alone.
Let us suppose a site S located inside a bin. Atoll knows the altitudes of four bin vertices, S’1, S’’1, S’2 and S’’2, from the DTM
file (centre of each DTM pixel).
Figure 2.3: Ground Altitude Determination - 1
1. Atoll draws a vertical line through S. This line respectively intersects (S’1,S’’1) and (S’2, S’’2) lines at S1 and S2.
2. Atoll determines the S1 and S2 altitudes using a linear interpolation method.
3. Atoll performs a second linear interpolation to evaluate the S altitude.
58
2.1.2 Clutter Determination

Some propagation models need clutter class and clutter height as information at receiver or along a transmitter-receiver
profile.
2.1.2.1 Clutter Classes

Atoll uses clutter classes file to determine the clutter class. The clutter classes map is a grid representing the ground with each
bin assigned a clutter class code corresponding to its clutter type. It is possible to specify an average height for each clutter
class in Atoll.
Clutter class files provide a clutter code per bin. Bin size is defined by pixel size P (in metre). Pixel size must be the same in
both directions. Abscissa and ordinate axes are respectively oriented in right and downwards directions. The first point given
in the file corresponds to the centre of the top-left pixel of the map (northwest point geo-referenced by Atoll.
Figure 2.7: Clutter Classes
Atoll supports a maximum of 255 clutter classes (8 bits/pixel). A clutter classes file file that contains N x N bins requires N2
code values.
2.1.2.2 Clutter Heights

To evaluate the clutter height, Atoll uses clutter heights file if available in the .atl document; clutter height of a site is the
height of the nearest point in the file.
Example: Let us suppose a site S. In the clutter heights file, Atoll reads clutter heights of four points around the site, S’1, S’’1,
S’2 and S’’2. Here, the nearest point to S is S”2; therefore Atoll takes the S”2 clutter height as clutter height of S.
Figure 2.8: Clutter Height
If you do not have any clutter height file, Atoll takes clutter height information in clutter classes file. In this case, clutter height
is an average height related to a clutter class.
59
2.1.3 Geographic Profile Extraction

Geographic profile extraction is needed in order to calculate diffraction losses. Profiles can be based on DTM only or on DTM
and clutter both, depending on the selected propagation model.
Method 1: Radial Extraction
Atoll draws radials from the site (where transmitter is located) to each calculation bin located along the transmitter
calculation area border. In other words, Atoll determines a geographic profile between site and each bin centre.
Figure 2.9: Radial calculation method
Transmitter location
Radials (Atoll extracts a geographic profile for each radial)
Centres of bins located on the calculation border
Receiver location
Figure 2.10: Site-bin centre profile
Depending on the calculation being carried out, the receiver may be located at the centre of a calculation bin (coverage
predictions) or anywhere within a calculation bin. Atoll uses the profile nearest to the receiver for calculations (the receiver
is assumed to be located on the profile).
60
Method 2: Systematic Extraction
Atoll extracts a precise geographic profile between the site and the receiver.
Figure 2.11: Radial calculation method
Transmitter location
Geographic profile
Receiver location
2.1.4 Resolution of the Extracted Profiles

Geographic profile resolution depends on resolution of geographic data used by the propagation model (DTM and/or clutter).
The selected profile resolution does not depend on the geographic layer order.
• If the propagation model uses both DTM and clutter heights along the profile, the profile resolution will be the highest
of the two.
Example 1 (Using the Standard Propagation Model)
A DTM map with a 40 m resolution and a clutter heights map with a 20 m resolution are available. The profile
resolution will be 20 m. It means that Atoll will extract geographic information, ground altitude and clutter height,
every 20 m.
To get ground altitude every 20 m, Atoll uses the bilinear interpolation method described in "Ground Altitude
Determination" on page 57. Clutter heights are read from the clutter heights map. Atoll takes the clutter height of the
nearest point every 20 m.
Example 2 (Using the Standard Propagation Model)
A DTM map with a 40 m resolution and a clutter classes map with a 20 m resolution are available. No clutter height
file has been imported in the document. The profile resolution will be 20 m. It means that Atoll will extract geographic
information, ground altitude and clutter height, every 20 m.
Determination" on page 57. Atoll uses the clutter classes map to determine clutter height. Every 20 m, it determines
clutter class and takes associated average height.
• If the propagation model uses only DTM along the profile, the profile resolution will be the highest resolution among
the DTM files.
Example (Using the Cost-Hata Propagation Model)
61
DTM maps with 40 m and 25 m resolutions and a clutter map with a 20 m resolution are available. The profile
resolution will be 25 m. It means that Atoll will extract geographic information, only the ground altitude, every 25 m.
The profile resolution does not depend on the geographic layer order in the Geo tab of the Explorer window.
However, the geographic layer order has influence on the usage of the data. For example, when DTM 1 is on the top
of DTM 2, Atoll will use DTM 1 for extracting the profile where DTM a is available and it will use DTM 2 elsewhere.
Determination" on page 57.
Geo Tab of the Explorer Window

> DTM
> DTM 1 (25m)
> DTM 2 (40m)
> Clutter
> Clutter (20m)
2.2 List of Default Propagation Models

Propagation models available in Atoll are listed in the table below along with their main characteristics.
Propagation ITU 370-7

ITU 1546 ITU 526-5 WLL
model (Vienna 93)
Frequency band 100-400 MHz 30-3000 MHz 30-10000 MHz 30-10000 MHz
Physical Free space loss Free space loss Free space loss
Free space loss + corrections
phenomena Corrected standard loss Diffraction loss Diffraction loss
Diffraction Deygout (3 obstacles)
calculation - - Deygout corrected Deygout (3 obstacles)
method (3 obstacles)
DTM
Profile based on - - DTM
Clutter
Profile extraction
- - Radial Radial
mode
Cell size Macro cell Macro cell Macro cell -
Street
Receiver location Rooftop Rooftop Street
Rooftop
Receiver Fixed Mobile Fixed Fixed
d > 10 km 1 < d < 1000 km Fixed receivers
Fixed receivers
Use Low frequencies Land and maritime mobile, WLL, Microwave links,
WLL
Broadcast broadcast WiMAX
Propagation Standard Propagation COST-Hata

Erceg-Greenstein (SUI) ITU 529-3
model Model Okumura-Hata
Frequency band 150-3500 MHz 1900-6000 MHz 300-1500 MHz 150-2000 MHz
L(d, f, HTx, HRx) L(d, f, HRx) L(d, f, HRx)
Physical L(d, HTxeff, HRxeff, Diff loss,
(per environment) (per environment) (per environment)
phenomena clutter)
Diffraction loss Diffraction loss Diffraction loss
Deygout (3 obstacles)
Epstein-Peterson
Diffraction
(3 obstacles) Deygout Deygout Deygout
calculation
Deygout corrected (1 obstacle) (1 obstacle) (1 obstacle)
method
(3 obstacles)
Millington (1 obstacle)
62
Propagation Standard Propagation COST-Hata

Erceg-Greenstein (SUI) ITU 529-3
model Model Okumura-Hata
DTM
Profile based on DTM DTM DTM
Clutter
Profile extraction Radial
Radial Radial Radial
mode Systematic
Macro cell Macro cell Macro cell Macro cell
Cell size
Mini cell Mini cell Mini cell Mini cell
Street
Receiver location Street Street Street
Rooftop
Receiver Mobile and Fixed Fixed Mobile Mobile
1 < d < 20 km Urban and suburban areas
1 < d < 100 km
Use GSM, UMTS, CDMA2000, 100 m < d < 8 km GSM, UMTS, CDMA2000, LTE
GSM, CDMA2000, LTE
WiMAX, LTE Fixed WiMAX
2.3 Okumura-Hata and Cost-Hata Propagation Models

2.3.1 Hata Path Loss Formula
Hata formula empirically describes the path loss as a function of frequency, receiver-transmitter distance and antenna heights
for an urban environment. This formula is valid for flat, urban environments and 1.5 metre mobile antenna height.
Path loss (Lu) is calculated (in dB) as follows:
Lu = A 1 + A 2 log  f  + A 3 log  h Tx  +  B 1 + B 2 log  h Tx  + B 3 h Tx  log d
f is the frequency (MHz).

hTx is the transmitter antenna height above ground (m) (Hb notation is also used in Atoll).
d is the distance between the transmitter and the receiver (km).
The parameters A1, A2, A3, B1, B2, and B3 can be user-defined. Default values are proposed in the table below:
Okumura-Hata Cost-Hata
Parameters
f 1500 MHz f > 1500 MHz
A1 69.55 49.30
A2 26.16 33.90
A3 -13.82 -13.82
B1 44.90 44.90
B2 -6.55 -6.55
B3 0 0
2.3.2 Corrections to the Hata Path Loss Formula

As described above, the Hata formula is valid for urban environment and a receiver antenna height of 1.5m. For other
environments and mobile antenna heights, corrective formulas must be applied.
• For urban areas: L model1 = Lu – a  h Rx 
f 2
• For suburban areas: L model1 = Lu – a  h Rx  – 2  log  ------  – 5.4
28
2
• For quasi-open rural areas: L model1 = Lu – a  h Rx  – 4.78  log  f   + 18.33 log  f  – 35.94
2
• For open rural areas: L model1 = Lu – a  h Rx  – 4.78  log  f   + 18.33 log  f  – 40.94
a(hRx) is a correction for a receiver antenna height different from 1.5m.
• For rural/small cities: a  h Rx  =  1.1 log  f  – 0.7 h Rx –  1.56 log  f  – 0.8 

2
• For large cities: a  h Rx  = 3.2  log  11.75h Rx   – 4.97
63
When receiver antenna height equals 1.5m, a(hRx) is close to 0 dB regardless of frequency.
2.3.3 Calculations in Atoll

Hata models take into account topo map (DTM) between transmitter and receiver and morpho map (clutter) at the receiver.
1st step: For each calculation bin, Atoll determines the clutter bin on which the receiver is located. This clutter bin corresponds
to a clutter class. Then, it uses the Hata formula assigned to this clutter class to evaluate L model1 .
2nd step: This step depends on whether the ‘Add diffraction loss’ option is checked.
• If the ‘Add diffraction loss’ option is unchecked, Atoll stops calculations.
L model = L model1
• If the ‘Add diffraction loss’ option is selected, Atoll proceeds as follows:

a. It extracts a geographic profile between the transmitter and the receiver based on the radial calculation mode.
b. It determines the largest obstacle along the profile in accordance with the Deygout method and evaluates losses
due to diffraction L model2 .
L model = L model1 + L model2
2.4 ITU 529-3 Propagation Model

2.4.1 ITU 529-3 Path Loss Formula
The ITU 529.3 model is a Hata-based model. For this reason, its formula empirically describes the path loss as a function of
frequency, receiver-transmitter distance and antenna heights for a urban environment. This formula is valid for flat, urban
environments and 1.5 metre mobile antenna height.
The standard ITU 529-3 formula, for a receiver located on a urban environment, is given by:
b
E = 69.82 – 6.16 log f + 13.82 log h Tx –  44.9 – 6.55 log h Tx   log d 
where:
E is the field strength for 1 kW ERP
f is the frequency (MHz).
h Tx is the transmitter antenna height above ground (m) (Hb notation is also used in Atoll)
h Rx is the receiver antenna height above ground (m)
d is the distance between the transmitter and the receiver (km)

b is the distance correction
The domain of validity of such is formula is:
• Frequency range: 300-1500 MHz
• Base Station height: 30-200 m
• Mobile height: 1-10 m
• Distance range: 1-100 km
Since Atoll needs the path loss (Lu) formula, a conversion has to be made. One can find the following conversion formula:
Lu = 139.37 + 20 log f – E
which gives the following path loss formula for the ITU 529-3 model:
b
Lu = 69.55 + 26.16 log f – 13.82 log h Tx +  44.9 – 6.55 log hTx   log d 
2.4.2 Corrections to the ITU 529-3 Path Loss Formula

Environment Correction
As described above, the Hata formula is valid for urban environment. For other environments and mobile antenna heights,
corrective formulas must be applied.
64
L model1 = Lu – a  h Rx  for large city and urban environments
f 2
L model1 = Lu – a  h Rx  – 2  log  ------  – 5.4 for suburban area
  28 
2
L model1 = Lu – a  h Rx  – 4.78  log f  + 18.33 log f – 40.94 for rural area
Area Size Correction
In the formulas above, a  h Rx  is the environment correction and is defined according to the area size.
• For rural/small cities: a  h Rx  =  1.1 log  f  – 0.7 h Rx –  1.56 log  f  – 0.8 

2
• For large cities: a  h Rx  = 3.2  log  11.75h Rx   – 4.97
Distance Correction
The distance correction refers to the term b above.

• d<20 km: b = 1
d 0.8 h Tx
d>20 km: b = 1 +  0.14 + 1.87  10 f + 1.07  10 h' Tx    log ------
–4 –3
• with h' Tx = -----------------------------------------
 20 –6 2
1 + 7  10 h Tx

Hata-based models take into account topo map (DTM) between transmitter and receiver and morpho map (clutter) at the
receiver.
1st step: For each calculation bin, Atoll determines the clutter bin on which the receiver is located. This clutter bin corresponds
to a clutter class. Then, it uses the ITU 529-3 formula assigned to this clutter class to evaluate L model1 .
2nd step: This step depends on whether the ‘Add diffraction loss’ option is checked.
• If the ‘Add diffraction loss’ option is unchecked, Atoll stops calculations.
L model = L model1

a. It extracts a geographic profile between the transmitter and the receiver based on the radial calculation mode.
due to diffraction  L model2  .
L model = L model1 + L model2
2.5 Standard Propagation Model (SPM)

2.5.1 SPM Path Loss Formula
SPM is based on the following formula:
L model = K 1 + K 2 log  d  + K 3 log  H Txeff  + K 4  DiffractionLoss + K 5 log  d   log  H Txeff  +

K 6  H Rxeff  + K 7 log  H Rxeff  + K clutter f  clutter 
with,
K1: constant offset (dB).
K2: multiplying factor for log(d).
d: distance between the receiver and the transmitter (m).
K3: multiplying factor for log(HTxeff).
HTxeff: effective height of the transmitter antenna (m).
K4: multiplying factor for diffraction calculation. K4 has to be a positive number.
65
Diffraction loss: loss due to diffraction over an obstructed path (dB).

K5: multiplying factor for log  d   log  H Txeff 
K6: multiplying factor for H Rxeff
K7: multiplying factor for log  H Rxeff  .
H Rxeff : effective mobile antenna height (m).
Kclutter: multiplying factor for f(clutter).

f(clutter): average of weighted losses due to clutter.

2.5.2.1 Visibility and Distance Between Transmitter and Receiver
For each calculation bin, Atoll determines:
• The distance between the transmitter and the receiver.
If the distance Tx-Rx is less than the maximum user-defined distance (break distance), the receiver is considered to be near
the transmitter. Atoll will use the set of values marked “Near transmitter”.
If the distance Tx-Rx is greater than the maximum distance, receiver is considered far from transmitter. Atoll will use the set
of values “Far from transmitter”.
• Whether the receiver is in the transmitter line of sight or not.
If the receiver is in the transmitter line of sight, Atoll will take into account the set of values (K1,K2)LOS. The LOS is defined by
no obstruction along the direct ray between the transmitter and the receiver.
If the receiver is not in the transmitter line of sight, Atoll will use the set of values (K1,K2)NLOS.
2.5.2.2 Effective Transmitter Antenna Height

Effective transmitter antenna height (HTxeff) may be calculated with six different methods.
Height Above Ground
The transmitter antenna height is above the ground (HTx in m).

HTxeff = HTx
Height Above Average Profile
The transmitter antenna height is determined relative to an average ground height calculated along the profile between a
transmitter and a receiver. The profile length depends on distance min and distance max values and is limited by the
transmitter and receiver locations. Distance min and Distance max are minimum and maximum distances from the transmitter
respectively.
H Txeff = H Tx +  H 0Tx – H 0 
where,
H 0Tx is the ground height (ground elevation) above sea level at transmitter (m).
H 0 is the average ground height above sea level along the profile (m).
If the profile is not located between the transmitter and the receiver, HTxeff equals HTx
only.
Slope at Receiver Between 0 and Minimum Distance
The transmitter antenna height is calculated using the ground slope at receiver.
H Txeff =  H Tx + H 0Tx  – H 0Rx + K  d
where,
66
H 0Rx is the ground height (ground elevation) above sea level at receiver (m).
K is the ground slope calculated over a user-defined distance (Distance min). In this case, Distance min is a distance from
receiver.
• If H Txeff  20m then, Atoll uses 20m in calculations.

• If H Txeff  200m then, Atoll takes 200m.
Spot Ht
If H 0Tx  H 0Rx then, H Txeff = H Tx +  H 0Tx – H 0Rx 
If H 0Tx  H 0Rx then, H Txeff = H Tx
Absolute Spot Ht
H Txeff = H Tx + H 0Tx – H 0Rx
Distance min and distance max are set to 3000 and 15000 m according to ITU
recommendations (low frequency broadcast f < 500 Mhz) and to 0 and 15000 m
according Okumura recommendations (high frequency mobile telephony).
These values are only used in the two last methods and have different meanings according to the method.
Enhanced Slope at Receiver
Atoll offers a new method called “Enhanced slope at receiver” to evaluate the effective transmitter antenna height.
Figure 2.12: Enhanced Slope at Receiver
Let x-axis and y-axis respectively represent positions and heights. We assume that x-axis is oriented from the transmitter
(origin) towards the receiver.
This calculation is achieved in several steps:
1. Atoll determines line of sight between transmitter and receiver.
The LOS line equation is:
  H 0Tx + H Tx  –  H 0Rx + H Rx  
Los  i  =  H 0Tx + H Tx  – -----------------------------------------------------------------------
- Res  i 
d
where,
H Rx is the receiver antenna height above the ground (m).
i is the point index.

Res is the profile resolution (distance between two points).
2. Atoll extracts the transmitter-receiver terrain profile.
67
3. Hills and mountains are already taken into account in diffraction calculations. Therefore, in order for them not to
unfavourably influence the regression line calculation, Atoll filters the terrain profile.
Atoll calculates two filtered terrain profiles; one established from the transmitter and another from the receiver. It
determines filtered height of every profile point. Profile points are evenly spaced on the basis of profile resolution. To
determine filtered terrain height at a point, Atoll evaluates ground slope between two points and compares it with a
threshold set to 0.05; where three cases are possible.
Some notations defined hereafter are used in next part.
H filt is the filtered height.
H orig is the original height. Original terrain height is determined from extracted ground profile.
• Filter starting from transmitter

Let us assume that H filt – Tx  Tx  = H orig  Tx 
For each point, we have three different cases:
H orig  i  – H orig  i – 1 
i. If H orig  i   H orig  i – 1  and ---------------------------------------------------
-  0.05 ,
Res
Then, H filt – Tx  i  = H filt – Tx  i – 1  +  H orig  i  – H orig  i – 1  
H orig  i  – H orig  i – 1 
-  0.05
ii. If H orig  i   H orig  i – 1  and ---------------------------------------------------
Res
Then, H filt – Tx  i  = H filt – Tx  i – 1 
iii. If H orig  i   H orig  i – 1 
Then, H filt – Tx  i  = H filt – Tx  i – 1 
If H filt  i   H orig  i  additionally
Then, H filt – Tx  i  = H orig  i 
• Filter starting from receiver

Let us assume that H filt  Rx  = H orig  Rx 
For each point, we have three different cases:
H orig  i  – H orig  i + 1 
i. -  0.05 ,
If H orig  i   H orig  i + 1  and ---------------------------------------------------
Res
Then, H filt – Rx  i  = H filt – Rx  i + 1  +  H orig  i  – H orig  i + 1  
H orig  i  – H orig  i + 1 
ii. If H orig  i   H orig  i + 1  and ----------------------------------------------------  0.05
Res
Then, H filt – Rx  i  = H filt – Rx  i + 1 
iii. 3rd case: If H orig  i   H orig  i + 1 
Then, H filt – Rx  i  = H filt – Rx  i + 1 
If H filt  i   H orig  i  additionally
Then, H filt – Rx  i  = H orig  i 
Then, for every point of profile, Atoll compares the two filtered heights and chooses the higher one.
H filt  i  = max  H filt – Tx  i  H filt – Rx  i  
4. Atoll determines the influence area, R. It corresponds to the distance from receiver at which the original terrain profile
plus 30 metres intersects the LOS line for the first time (when beginning from transmitter).
The influence area must satisfy additional conditions:
• R  3000m
68
• R  0.01  d
• R must contain at least three bins.
• When several influence areas are possible, Atoll chooses the highest one.
• If d < 3000m, R = d.
5. Atoll performs a linear regression on the filtered profile within R in order to determine a regression line.
The regression line equation is:
y = ax + b
  d  i  – dm   Hfilt  i  – Hm 
a = ---------------------------------------------------------------------
i and b = H m – ad m

2
 d  i  – dm 
i
where,
1
H m = ---
n  Hfilt  i 
i
i is the point index. Only points within R are taken into account.
dm = d – R
---
2
d(i) is the distance between i and the transmitter (m).

Then, Atoll extends the regression line to the transmitter location. Therefore, its equation is:
regr  i  = a   i  Res  + b
6. Then, Atoll calculates effective transmitter antenna height, H Txeff (m).
H 0Tx + H Tx – b
H Txeff = ---------------------------------
-
2
1+a
If HTxeff is less than 20m, Atoll recalculates it with a new influence area, which begins at transmitter.
• In case H Txeff  1000m , 1000m will be used in calculations.

• If H Txeff is still less than 20m, an additional correction is taken into account (7th step).
7. If H Txeff is still less than 20m (even negative), Atoll evaluates path loss using H Txeff = 20m and applies a correction
factor.
Therefore, if H Txeff  20m ,
L model = L model   H Txeff = 20m  d f  + K lowant
d - –  0.3   H 20   1 –  H Txeff – 20  
where, K lowant = -------
5 Txeff – 20   – --------------------------------------------------------------------------
10  9.63 + ----------- d -   6.93 + ----------- d -
 1000  1000
2.5.2.3 Effective Receiver Antenna Height

H Rxeff =  H Rx + H 0Rx  – H 0Tx
where,
H Rx is the receiver antenna height above the ground (m).
H 0Rx is the ground height (ground elevation) above sea level at the receiver (m).
69
H 0Tx is the ground height (ground elevation) above sea level at the transmitter (m).
The calculation of effective antenna heights ( H Rxeff and H Txeff ) is based on extracted
DTM profiles. They are not properly performed if you have not imported heights (DTM
file) beforehand.
2.5.2.4 Correction for Hilly Regions in Case of LOS

An optional corrective term enables Atoll to correct path loss for hilly regions when the transmitter and the receiver are in
Line-of-sight.
Therefore, if the receiver is in the transmitter line of sight and the Hilly terrain correction option is active, we have:
L model = K 1 LOS + K 2 LOS log  d  + K 3 log  H Txeff  + K 5 log  H Txeff  log  d  + K 6  H Rx + K clutter f  clutter  + K hill LOS
When the transmitter and the receiver are not in line of sight, the path loss formula is:
L model = K 1 NLOS + K 2 NLOS log  d  + K 3 log  H Txeff  + K 4  Diffraction + K 5 log  H Txeff  log  d  + K 6  H Rx + K clutter f  clutter 
K hill LOS is determined in three steps. Influence area, R, and regression line are supposed available.
1st step: For every profile point within influence area, Atoll calculates height deviation between the original terrain profile and
regression line. Then, it sorts points according to the deviation and draws two lines (parallel to the regression line), one which
is exceeded by 10% of the profile points and the other one by 90%.
2nd step: Atoll evaluates the terrain roughness, h; it is the distance between the two lines.
3rd step: Atoll calculates K hill LOS .
We have K hill LOS = K h + K hf
If 0  h  20m , K h = 0
2
Else K h = 7.73  log  h   – 15.29 log  h  + 6.746
If 0  h  10m , K hf = – 2  0.1924   H 0Rx + H Rx – regr  i Rx  
2 H 0Rx + H Rx – regr  i Rx 
Else K hf = – 2   – 1.616  log  h   + 14.75 log  h  – 11.21   -----------------------------------------------------
-
h
iRx is the point index at receiver.
2.5.2.5 Diffraction
Four methods are available to calculate diffraction loss over the transmitter-receiver profile.
Along the transmitter-receiver profile, you may consider:
• Either ground altitude and clutter height (Consider heights in diffraction option),
In this case, Atoll uses clutter height information from clutter heights file if available in the .atl document. Otherwise,
it considers average clutter height specified for each clutter class in the clutter classes file description.
• Or only ground altitude.
2.5.2.6 Losses due to Clutter

n
Atoll calculates f(clutter) over a maximum distance from receiver: f  clutter  =  Li wi

i=1
where,
L: loss due to clutter defined in the Clutter tab by the user (in dB).
w: weight determined through the weighting function.
n: number of points taken into account over the profile. Points are evenly spaced depending on the profile resolution.
Four weighting functions are available:
70
• Uniform weighting function: w i = --1-

n
di
• Triangular weighting function: w i = ------------
n
-
 dj
j=1
• d i = D – d' i , where d’i is the distance between the receiver and the ith point and D is the maximum distance defined.
d
log  ----i + 1
D 
• Logarithmic weighting function: w i = ------------------------------------
n
d
log  ----j + 1
 D
j=1
di
----
D
• e –1
Exponential weighting function: w i = ------------------------
n dj
----
e
D
–1
j=1
The chart below shows the weight variation with the distance for each weighting function.
Figure 2.13: Losses due to Clutter
2.5.2.7 Recommendations
Beware that the clutter influence may be taken into account in two terms, Diffraction loss and f(clutter) at the same time. To
avoid this, we advise:
1. Not to consider clutter heights to evaluate diffraction loss over the transmitter-receiver profile if you specify losses
per clutter class.
This approach is recommended if the clutter height information is statistical (clutter roughly defined, no altitude).
Or
2. Not to define any loss per clutter class if you take clutter heights into account in the diffraction loss.
In this case, f(clutter)=0. Losses due to clutter are only taken into account in the computed Diffraction loss term.
This approach is recommended if the clutter height information is either semi-deterministic (clutter roughly defined,
altitude defined with an average height per clutter class) or deterministic (clutter sharply defined, altitude defined
with an average height per clutter class or - even better - via a clutter height file).
In case of semi-deterministic clutter information, specify receiver clearance (m) per clutter class. Both ground altitude and
clutter height are considered along the whole transmitter-receiver profile except over a specific distance around the receiver
(clearance), where Atoll proceeds as if there was only the DTM map. The clearance information is used to model streets.
71
Figure 2.14: Tx-Rx profile
In the above figure, the ground altitude and clutter height (in this case, average height specified for each clutter class in the
clutter classes map description) are taken into account along the profile.
Clearance definition is not necessary in case of deterministic clutter height information. Clutter height information is accurate
enough to be used directly without additional information such as clearance. Two cases can be considered:
1. If the receiver is in the street (clutter height lower than receiver height), Atoll calculates the path loss by considering
potentially some diffraction loss at reception.
2. If the receiver is supposed to be inside a building (clutter height higher than receiver height), Atoll does not consider
any difraction (and clearance) from the building but takes into account the indoor loss as an additional penetration
loss.
• To consider indoor losses in building only when using a deterministic clutter map
(clutter height map), the 'Indoor Coverage' box must not be checked in predictions
unless this loss will be counted twice inside buildings (on the entire reception clutter
class and not only inside the building).
• Even with no clearance, the clutter height (extracted either from clutter class or
clutter height folders) is never considered at the last profile point.
2.5.3 Automatic Calibration

The goal of this tool is to calibrate parameters and methods of the SPM formula in a simple and reproducible way. Calibration
is based on imported CW measurement data. It is the process of limiting the difference between predicted and measured
values. For a complete description of the calibration procedure (including the very important prerequisite filtering work on
the CW measurement points), please refer to the User Manual and the SPM Calibration Guide.
The following SPM formula parameters can be estimated:
• K1, K2, K3, K4, K5, K6 and K7
• Losses per clutter class (Kclutter must be user-defined)
• Effective antenna height method
• Diffraction method
Automatic model calibration provides a mathematical solution. The relevance of this mathematical solution with a physical
and realistic solution must be determined before committing these results.
You must keep in mind that the model calibration and its result (standard deviation and root mean square) strongly depend
on the CW measurement samples you use. A calibrated model must restore the behaviour of CW measurements depending
on their configuration on a large scale, and not just totally coincide with a few number of CW measurements. The calibrated
model has to give correct results for every new CW measurement point in the same geographical zone, without having been
calibrated on these new CW measurements.
2.5.3.1 General Algorithm

Propagation model calibration is a special case of the more general Least-Square problems, i.e. given a real m x n matrix A,
and a real m-vector b, find a real n-vector x0 that minimises the Euclidean length of Ax - b.
Here,
m is the number of measurement points,
n is the number of parameters to calibrate,
72
A is the values of parameter associated variables (log(d), log(heff), etc.) at each measurement point, and
b is the vector of measurement values.
The vector x0 is the set of parameters found at the end of the calibration.
The theoretical mathematical solution of this problem was found by Gauss (around 1830). Further enhancements to the
original method were proposed in the 60's in order to solve the numerical instability problem.
In 1974, Lawson & Hanson [2] proposed a theoretical solution of the least-square problem with general linear inequality
constraints on the vector x0. Atoll implementation is based on this method, which is explained in detail in [1].
References:
[1] Björck A. “Numerical Methods for Least Square Problems”, SIAM, 1996.
[2] Lawson C.L., Hanson R.J. “Solving Least Squares Problems”, SIAM, 1974.
2.5.3.2 Sample Values for SPM Path Loss Formula Parameters

The following tables list some sample orders of magnitudes for the different parameters composing the Standard Propagation
Model formula.
Minimum Typical Maximum

K1 Variable Variable Variable
K2 20 44.9 70
K3 -20 5.83 20
K4 0 0.5 0.8
K5 -10 -6.55 0
K6 -1 0 0
K7 -10 0 0
It is recommended to set K6 to 0, and use K7 instead of K6. K6 is a multiplicative coefficient to a value in dB, which means that
slight variations in K6 have considerable impact on the path loss.
K1 depends on the frequency and the technology. Here are some sample values:
Project type Frequency (MHz) K1

GSM 900 935 12.5
GSM 1800 1805 22
GSM 1900 1930 23
UMTS 2045a 23.8
1xRTT 1900 23
2300 25.6
2500 26.8
WiMAX 2700 27.9
3300 30.9
3500 31.7
a. 2045 MHz = (2140 + 1950)/2. It is the average of the downlink

and uplink centre frequencies of the band.
The above K1 values for WiMAX are extrapolated estimates for different frequency ranges. It is highly recommended to
calibrate the SPM using measurement data collected on the field for WiMAX networks before using the SPM for predictions.
All K paramaters can be defined by the automatic calibration wizard. Since Kclutter is a constant, its value is strongly dependant
on the values given to the losses per clutter classes. From experience, typical losses (in dB) per clutter class are:
Dense urban From 4 to 5

Woodland From 2 to 3
Urban 0
73
Suburban From -5 to -3
Industrial From -5 to -3
Open in urban From -6 to -4
Open From -12 to -10
Water From -14 to -12
These values have to be entered only when considering statistical clutter class maps only.
The Standard Propagation Model is derived from the Hata formulae, valid for urban
environments. The above values are normalized for urban clutter types (0 dB for urban
clutter class). Positive values correspond to more dense clutter classes and negative
values to less dense clutter classes.
2.5.4 Unmasked Path Loss Calculation

You can use the SPM to calculate unmasked path losses. Unmasked path losses are calculated by not taking into account the
transmitter antenna patterns, i.e., the attenuation due to the transmitter antenna pattern is not included. Such path losses
are useful when using path loss matrices calculated by Atoll with automatic optimisation tools.
The instance of the SPM available by default, under the Propagation Models folder in the Modules tab, has the following
characteristics:
• Signature: {D5701837-B081-11D4-931D-00C04FA05664}
• Type: Atoll.StdPropagModel.1
You can access these parameters in the Propagation Models table by double-clicking the Propagation Models folder in the
Modules tab.
To make the SPM calculate path losses excluding the antenna pattern attenuation, you have to change the type of the SPM to:
• Type: Atoll.StdPropagModelUnmasked.1
However, changing the type only does not invalidate the already calculated path loss matrices, because the signature of the
propagation model is still the same. If you want Atoll to recognize that the SPM has changed, and to invalidate the path loss
matrices calculated with this model, you have to change the signature of the model as well. The default signature for the SPM
that calculates unmasked path loss matrices is:
• Signature: {EEE060E5-255C-4C1F-B36C-A80D3D972583}
The above signature is a default signature. Atoll automatically creates different signatures for different instances of the same
propagation model. Therefore, it is possible to create different instances of the SPM, with different parameter settings, and
create unmasked versions of these instances.
You can change the signature and type of the original instance of the SPM, but it is recommended to make a copy of the SPM
in order not to lose the original SPM parameters. So, you will be able to keep different versions of the SPM, those that
calculate path losses with antenna pattern attenuation, and others that calculate path losses without it.
The usual process flow of an ACP working on an Atoll document through the API would be to:
1. Backup the storage directory of path loss matrices.
2. Set a different storage directory for calculating and storing unmasked path loss matrices.
3. Select the SPM used, backup it’s signature, and change its signature and type as shown above.
4. Perform optimisation using the path loss matrices calculated by the unmasked version of the SPM.
5. Restore the type and the signature of the SPM.
6. Reset the path loss storage directory to the original one.
74
• It is not possible to calibrate the unmasked version of the SPM using measurement
data.
• Using the SPM, you can also calculate the angles of incidence by creating a new
instance of the SPM with the following characteristics:
Type: Atoll.StdPropagModelIncidence.1
Signature: {659F0B9E-2810-4e59-9F0D-DA9E78E1E64B}
• The "masked" version of the algorithm has not been changed. It still takes into
account Atoll.ini options. However, the "unmasked" version does not take Atoll.ini
options into account.
• It’s highly recommended to use one method (Atoll.ini options) or the other one (new
identifier & signature) but not to combine both.
2.6 WLL Propagation Model

2.6.1 WLL Path Loss Formula
L model = L FS + F Diff  L Diff
Where L FS is the free space loss calculated using the formula entered in the model properties, L Diff is the diffraction loss
calculated using the 3-obstacle Deygout method, and F Diff is the diffraction multiplying factor defined in the model
properties.

Free Space Loss
For free space loss calculation, see "Free Space Loss" on page 86.
Diffraction
Atoll calculates diffraction loss along the transmitter-receiver profile built from DTM and clutter maps. Therefore, losses due
to clutter are taken into account in diffraction losses. Atoll takes clutter height information from the clutter heights file if
available in the .atl document. Otherwise, it considers average clutter height specified for each clutter class in the clutter
classes file description.
The Deygout construction (considering 3 obstacles) is used. This method is described under "Diffraction" on page 86. The final
diffraction losses are determined by multiplying the diffraction losses calculated using the Deygout method by the Diffraction
multiplying factor defined in the model properties.
• Receiver Clearance
Define receiver clearance (m) per clutter class when clutter height information is either statistical or semi-
deterministic. Both ground altitude and clutter height are considered along the whole profile except over a specific
distance around the receiver (clearance), where Atoll proceeds as if there was only the DTM map (see SPM part). Atoll
uses the clearance information to model streets.
If the clutter is deterministic, do not define any receiver clearance (m) per clutter class. In this case, clutter height
information is accurate enough to be used directly without additional information such as clearance (Atoll can locate
streets).
• Receiver Height
Entering receiver height per clutter class enables Atoll to consider the fact that receivers are fixed and located on the
roofs.
• Visibility
If the option ‘Line of sight only’ is not selected, Atoll computes Lmodel on each calculation bin using the formula defined
above. When selecting the option ‘Line of sight only’, Atoll checks for each calculation bin if the Diffraction loss (as
defined in the Diffraction loss: Deygout part) calculated along profile equals 0.
• In this case, receiver is considered in ‘line of sight’ and Atoll computes Lmodel on each calculation bin using the
formula defined above.
• Otherwise, Atoll considers that Lmodel tends to infinity.
75
2.7 ITU-R P.526-5 Propagation Model

L model = L FS + L Diff
Where L FS is the free space loss calculated using the formula entered in the model properties and L Diff is the diffraction loss
calculated using the 3-obstacle Deygout method.

Free Space Loss
Diffraction
Atoll calculates diffraction loss along the transmitter-receiver profile is built from the DTM map. The Deygout construction
(considering 3 obstacles), with or without correction, is used. These methods are described under "Diffraction" on page 86.

If d<1 km, L model = L FS
If d>1000 km, L model = 1000
If 1<d<1000 km, L model = max  L FS CorrectedS tan dardLoss 
d is the distance between the transmitter and the receiver (km).

Free Space Loss
Corrected Standard Loss
This formula is given for a 60 dBm (1kW) transmitter power.

CorrectedS tan dardLoss = 60 – C n – A H – A cl – 108.75 + 31.54 – 20 log f
Rxeff
where,
Cn is the field strength received in dBV/m,
AH is a correction factor for effective receiver antenna height (dB),

Rxeff
Acl is the correction for terrain clearance angle (dB),

f is the frequency in MHz.
• Cn Calculation
The Cn value is determined from charts Cn=f(d, HTxeff).
In the following part, let us assume that Cn=En(d,HTxeff) (where En(d,HTxeff) is the field received in dBV/m) is read from
charts for a distance, d (in km), and an effective transmitter antenna height, HTxeff (in m).
First of all, Atoll evaluates the effective transmitter antenna height, H Txeff , as follows:
If 0  d  3km , H Txeff = H 0Tx + H Tx – H 0Rx
76
If 3  d  15km , H Txeff = H 0Tx + H Tx – H 0  3 ;d 
If 15  d , H Txeff = H 0Tx + H Tx – H 0  3 ;15 
where,
H Tx is the transmitter antenna height above the ground (m).
H 0Tx is the ground height (ground elevation) above sea level at the transmitter (m).
H 0  3 ;d  is the average ground height (m) above sea level for the profile between a point 3 km from transmitter and
the receiver (located at d km from transmitter).
H 0  3 ;15  is the average ground height (m) above sea level for the profile between a point 3 km and another 15 km
from transmitter.
Then, depending on d and HTxeff, Atoll determines Cn using bilinear interpolation as follows.
If 37.5 HTxeff 1200, Cn= En(d,HTxeff)
Otherwise, Atoll considers d horizon = 4.1  H Txeff (d is stated in km)
Therefore,
If HTxeff < 37.5
If d  d horizon , we have C n = E n  d + 25 – d horizon 37.5 
Else Cn=En(d, 37.5) – En(dhorizon, 37.5) + En(25, 37.5)
If HTxeff > 1200
If d  d horizon , we have C n = E n  d + 142 – d horizon 1200 
Else Cn=En(d, 1200) – En(dhorizon, 1200) + En(142, 1200)
• AHRxeff Calculation
H Rx
= ---  20  log  -------
c -
AH
Rxeff 6 10 
where,
HRx is the user-defined receiver height,
c is the height gain factor.
c values are provided in the recommendation 370-7; for example, c=4 in a rural case.
• Acl Calculation
2
If f  300 MHz, A cl = 8.1 –  6.9 + 20 log     – 0.1  + 1  +   – 0.1   
2
Otherwise, A cl = 14.9 –  6.9 + 20 log     – 0.1  + 1  +   – 0.1   
f
With  = –   4000  ---------
300
where,
 is the clearance angle (in radians) determined according to the recommendation 370-7 (figure 19),
f is the frequency stated in MHz.
77
2.9 Erceg-Greenstein (SUI) Propagation Model

Erceg-Greenstein propagation model is a statistical path loss model derived from experimental data collected at 1.9 GHz in 95
macrocells. The model is for suburban areas, and it distinguishes between different terrain categories called the Stanford
University Interim Terrain Models. This propagation model is well suited for distances and base station antenna heights that
are not well-covered by other models. The path loss model applies to base antenna heights from 10 to 80 m, base-to-terminal
distances from 0.1 to 8 km, and three distinct terrain categories.
The basic path loss equation of the Erceg-Greenstein propagation model is:
PL = A + 10  a  H BS   Log 10  -----
d
 d 0
4d 0
Where A = 20  Log 10  ------------ . This is a fixed quantity which depends upon the frequency of operation. d is the distance

between the base station antenna and the receiver terminal and d0 is a fixed reference distance (100 m). a(HBS) is the
correction factor for base station antenna heights, HBS:
c-
a  H BS  = a – b  H BS + -------
H BS
Where 10 m  HBS  80 m , and a, b, and c are correction coefficients which depend on the SUI terrain type.
The Erceg-Greenstein propagation model is further developed through the correction factors introduced by the Stanford
University Interim model. The standards proposed by the IEEE working group 802.16 include channel models developed by
Stanford University. The basic path loss equation with correction factors is presented below:
PL = A + 10  a  H BS   Log 10  ----- + a  f  – a  H R 
d
d0
Where a(f) is the correction factor for the operating frequency, a  f  = 6  Log 10  ------------ , with f being the operating
f
2000
HR
frequency in MHz. a(HR) is the correction factor for the receiver antenna height, a  H R  = X  Log 10  ------ , where d depends
 2
on the terrain type.
• a(HR) = 0 for HR = 2 m.
• References:
• [1] V. Erceg et. al, “An empirically based path loss model for wireless channels in
suburban environments,” IEEE J. Select Areas Commun., vol. 17, no. 7, July 1999,
pp. 1205-1211.
• [2] Abhayawardhana, V.S.; Wassell, I.J.; Crosby, D.; Sellars, M.P.; Brown, M.G.;
"Comparison of empirical propagation path loss models for fixed wireless access
systems," Vehicular Technology Conference, 2005. IEEE 61st Volume 1, 30 May-1
June 2005 Page(s):73 - 77 Vol. 1
2.9.1 SUI Terrain Types

The SUI models are divided into three types of terrains2, namely A, B and C.
• Type A is associated with maximum path loss and is appropriate for hilly terrain with moderate to heavy tree
densities.
• Type B is characterised with either mostly flat terrains with moderate to heavy tree densities or hilly terrains with
light tree densities.
• Type C is associated with minimum path loss and applies to flat terrain with light tree densities.
The constants used for a, b, and c are given in the table below.
Model Parameter Terrain A Terrain B Terrain C

a 4.6 4.0 3.6
b (m-1) 0.0075 0.0065 0.005
2. The word ‘terrain’ is used in the original definition of the model rather than ‘environment’. Hence it is used
interchangeably with ‘environment’ in this description.
78
Model Parameter Terrain A Terrain B Terrain C

c (m) 12.6 17.1 20
X 10.8 10.8 20
2.9.2 Erceg-Greenstein (SUI) Path Loss Formula

The Erceg-Greenstein (SUI) propagation model formula can be simplified from the following equation:
4d
PL = 20  Log 10  -----------0- + 10  a  H BS   Log 10  ----- + a  f  – a  H R 
d
(1)
    d 0
to the equation below:

PL = – 7.366 + 26  Log 10  f  + 10  a  H BS    1 + Log 10  d   – a  H R  (2)
Where,
• f is the operating frequency in MHz
• d is the distance from the transmitter to the received in m in equation (1) and in km in equation (2)
• HBS is the transmitter height in m
• HR is the receiver height in m
The above equation is divided into two parts in Atoll:
PL = Lu – a  H R 
Where,
Lu = – 7.366 + 26  Log 10  f  + 10  a  H BS    1 + Log 10  d  
The above path loss formulas are valid for d > d0, i.e. d > 100 m. For d < 100 m, the path loss has been restricted to the free
space path loss with correction factors for operating frequency and receiver height:
4d 4d
PL = 20  Log 10  ------------------ + a  f  – a  H R  instead of PL = 20  Log 10  ------------------
 
Where a(f) and a(Hr) have the same definition as given above. Simplifying the above equation, we get,
PL = 12.634 + 26  Log 10  f  + 20  Log 10  d  – a  H R  , or Lu = 12.634 + 26  Log 10  f  + 20  Log 10  d 
The above equation is not user-modifiable in Atoll except for the coefficient of Log 10  f  , i.e. 26. Atoll uses the same
coefficient as the one you enter for Log 10  f  in Atoll for the case d > d0.
You can get the same resulting equation by setting a(hBS) = 2.

The Erceg-Greenstein (SUI) propagation model takes DTM into account between the transmitter and the receiver, and it can
also take clutter into account at the receiver location.
1st step: For each pixel in the calculation radius, Atoll determines the clutter bin on which the receiver is located. This clutter
bin corresponds to a clutter class. Atoll uses the Erceg-Greenstein (SUI) path loss formula assigned to this clutter class to
evaluate path loss.
2nd step: This step depends on whether the ‘Add diffraction loss’ option is selected or not.
• If the ‘Add diffraction loss’ option is not selected, 1st step gives the final path loss result.
a. It extracts a geographic profile between the transmitter and the receiver using the radial calculation method.
due to diffraction L Diffraction . For more information on the Deygout method, see "3 Knife-edge Deygout Method"
on page 87.
The final path loss is the sum of the path loss determined in 1st step and L Diffraction .
79
Shadow fading is computed in Atoll independent of the propagation model. For more information on the shadow fading
calculation, see "Shadow Fading Model" on page 90.

This propagation model is based on the P.1546-2 recommendations of the ITU-R. These recommendations extend the P.370-
7 recommendations, and are suited for operating frequencies from 30 to 3000 MHz. The path loss is calculated by this
propagation model with the help of graphs available in the recommendations. The graphs provided in the recommendations
represent field (or signal) strength, given in db  V  m  , as a function of distance for:
• Nominal frequencies, f n : 100, 600, and 1000 MHz
The graphs provided for 100 MHz are applicable to frequencies from 30 to 300 MHz, those for 600 MHz are applicable
to frequencies from 300 to 1000 MHz, and the graphs for 1000 MHz are applicable to frequencies from 1000 to
3000 MHz. The method for interpolation is described in the recommendations (Annex 5, § 6).
• Transmitter antenna heights, h 1 : 10, 20, 37.5, 75, 150, 300, 600, and 1200 m
For any values of h 1 from 10 to 3000 m, an interpolation or extrapolation from the appropriate two curves is used, as
described in the recommendations (Annex 5, § 4.1). For h 1 below 10 m, the extrapolation to be applied is given in
Annex 5, § 4.2. It is possible for the value of h 1 to be negative, in which case the method is given in Annex 5, § 4.3.
• Time variability, t : 1, 10, and 50 %

The propagation curves represent the field strength values exceeded for 1, 10 and 50 % of time.
• Receiver antenna height, h 2 : 10 m
For land paths, the graphs represent field strength values for a receiver antenna height above ground, equal to the
representative height of the clutter around the receiver. The minimum value of the representative height of clutter is
10 m. For sea paths, the graphs represent field strength values for a receiver antenna height of 10 m.
For other values of receiver antenna height, a correction is applied according to the environment of the receiver. The
method for calculating this correction is given in Annex 5, § 9.
These recommendations are not valid for transmitter-receiver distances less than 1 km or greater than 1000 km. Therefore in
Atoll, the path loss between a transmitter and a receiver over less than 1 km is the same as the path loss over 1 km. Similarly,
the path loss between a transmitter and a receiver over more than 1000 km is the same as the path loss over 1000 km.
Moreover, these recommendations are not valid for transmitter antenna heights less than the average clutter height
surrounding the transmitter.
• The cold sea graphs are used for calculations over warm and cold sea both.
• The mixture of land and sea paths is not supported by Atoll.

The input to the propagation model are the transmission frequency, transmitter and receiver heights, the distance between
the transmitter and the receiver, the precentage of time the field strength values are exceeded, the type of environment (i.e.,
land or sea), and the clutter at the receiver location.
In the following calculations, f is the transmission frequency, d is the transmitter-receiver distance, and t is the percentage
of time for which the path loss has to be calculated.
The following calculations are performed in Atoll to calculate the path loss using this propagation model.
2.10.1.1 Step 1: Determination of Graphs to be Used

First of all, the upper and lower nominal frequencies are determined for any given transmission frequency. The upper and
lower nominal frequencies are the nominal frequencies (100, 600, and 2000 MHz) between which the transmission frequency
is located, i.e., f n1  f  f n2 .
Once f n1 and f n1 are known, along with the information about the percentage of time t and the type of path (land or sea),
the sets of graphs which will be used for the calculation are also known.
80
2.10.1.2 Step 2: Calculation of Maximum Field Strength

A field strength must not exceed a maximum value, E Max , which is given by:
E Max = E FS = 106.9 – 20  Log  d  for land paths, and
E Max = E FS + E SE = 106.9 – 20  Log  d  + 2.38  1 – exp  – d  8.94    Log  50  t  for sea paths.
Where E FS is the free space field strength for 1 kW ERP, E SE is an enhancement for sea graphs.
2.10.1.3 Step 3: Determination of Transmitter Antenna Height

The transmitter antenna height to be used in the calculation depends on the type and length of the path.
• Land paths
h 1 = h eff
• Sea paths
h 1 = Max  1 h a 
Here, all antenna heights (i.e., h 1 , h eff , and h a ) are in expressed in m. h a is the antenna height above ground and h eff is the
effective height of the transmitter antenna, which is its height over the average level of the ground between distances of
0.2  d and d km from the transmitter in the direction of the receiver.
2.10.1.4 Step 4: Interpolation/Extrapolation of Field Strength

The interpolations are performed in series in the same order as described below. The first interpolation/extrapolation is
performed over the field strength values, E , from the graphs for transmitter antenna height to determine E h1 . The second
interpolation/extrapolation is performed over the interpolated/extrapolated values of E h1 to determine E d . And, the thrid
and final interpolation/extrapolation is performed over the interpolated/extrapolated values of E d to determine E f .
Step 4.1: Interpolation/Extrapolation of Field Strength for Transmitter Antenna Height
If the value of h 1 coincides with one of the eight heights for which the field strength graphs are provided, namely 10, 20, 37.5,
75, 150, 300, 600, and 1200 m, the required field strength is obtained directly from the corresponding graph. Otherwise:
• If 10 m  h 1  3000 m
The field strength is interpolated or extrapolated from field strengths obtained from two curves using the following
equation:
Log  h 1  h Low 
E h1 = E Low +  E Up – E Low   -------------------------------------
-
Log  h Up  h Low 
Where h Low = 600 m if h 1  1200 m , otherwise h Low is the nearest nominal effective height below h 1 ,
h Up = 1200 m if h 1  1200 m , otherwise h Up is the nearest nominal effective height above h 1 , E Low is the field
strength value for h Low at the required distance, and E Up is the field strength value for h Up at the required distance.
• If 0 m  h 1  10 m
• For land path if the transmitter-receiver distance is less than the smooth-Earth horizon distance
d H  h 1  = 4.1  h 1 , i.e., if d  4.1  h 1 ,
E h1 = E 10  d H  10   + E 10  d  – E 10  d H  h 1   , or
E h1 = E 10  12.9 km  + E 10  d  – E 10  d H  h 1   because d H  10  = 12.9 km
• For land path if the transmitter-receiver distance is greater than or equal to the smooth-Earth horizon distance
d H  h 1  = 4.1  h 1 , i.e., if d  4.1  h 1 ,
E h1 = E 10  d H  10  + d – d H  h 1   , or E h1 = E 10  12.9 km + d – d H  h 1   because d H  10  = 12.9 km
Where E x  y  is the field strength value read for the transmitter-receiver distance of y from the graph available
for the transmitter antenna height of x.
81
If in the above equation, d H  10  + d – d H  h 1   1000 km even though d  1000 km , the field strength is
determined from linear extrapolation for Log (distance) of the graph given by:
Log  d  D Low 
E h1 = E Low +  E Up – E Low   --------------------------------------
-
Log  D Up  D Low 
Where D Low is penultimate tabulation distance (km), D Up is the final tabulation distance (km), E Low is the field
strength value for D Low , and E Up is the field strength value for D Up .
• For sea path, h 1 should not be less than 1 m. This calculation requires the distance at which the path has 0.6 of
the first Fresnel zone just unobstructed by the sea surface. This distance is given by:
D h1 = D 0.6  f h 1  h 2 = 10 m   (km)
Df  Dh
Where D 0.6 = Max  0.001 ----------------- (km) with D f = 0.0000389  f  h 1  h 2 (frequency-dependent term), and
 D f + D h
D h = 4.1   h 1 + h 2  (asymptotic term defined by the horizon distance).
If d  D h1 the 0.6 Fresnel clearance distance for the sea path where the transmitter antenna height is 20 m is also
calculated as:
D 20 = D 0.6  f  h 1 = 20 m   h 2 = 10 m   (km)
Once D h1 and D 20 are known, the field strength for the required distance is given by:
 E Max for d  D h1

 Log  d  D h1 
E h1 =  E D +  E D – E D   ----------------------------------
- for D h1  d  D 20
 h1 20 h1 Log  D 20  D h1 

 E'   1 – F S  + E''  F S for d  D 20
Where E Max is the maximum field strength at the required distance as calculated in "Step 2: Calculation of
Maximum Field Strength" on page 81, E D is E Max for d = D h1 ,
h1
Log  h1  10  Log  h1  10 
ED = E 10  D 20  +  E 20  D 20  – E 10  D 20    ------------------------------- , E' = E 10  d  +  E 20  d  – E 10  d    ------------------------------- , and
20 Log  20  10  Log  20  10 
E'' is the field strength calculated as described for land paths. E 10  y  and E 20  y  are field strengths interpolated
for distance y and h 1 = 10 m and 20 m , respectively, and F S =  d – D 20   d .
• If h 1  0 m
A correction is applied to the field strength, E h1 , calculated in the above description in order to take into account the
diffraction and tropospheric scattering. This correction is the maximum of the diffraction correction,, and
tropospheric scattering correction, .
C h1 = Max  C h1d C h1t 
2
Where C h1d = 6.03 – J    with J    =  6.9 + 20  Log    – 0.1  + 1 +  – 0.1   and  = K    eff2 ,
–h1 
 eff2 = arc tan  -----------
- , and K  is 1.35 for 100 MHz, 3.31 for 600 MHz, 6.00 for 2000 MHz.
9000
e  180  d- a = 6370 km
C h1t = 30  Log  ----------------------
- with  e = -------------------- , (radius of the Earth), and k = 4  3 is the
  e +  eff2 ak
effective Earth radius factor for mean refractivity conditions.
Step 4.2: Interpolation/Extrapolation of Field Strength for Transmitter-Receiver Distance
In the field strength graphs in the recommendations, the field strength is plotted against distance from 1 km to 1000 km. The
distance values for which field strengths are tabulated are 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 25,
30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 110, 120, 130, 140, 150, 160, 170, 180, 190, 200, 225, 250, 275, 300,
325, 350, 375, 400, 425, 450, 475, 500, 525, 550, 575, 600, 625, 650, 675, 700, 725, 750, 775, 800, 825, 850, 875, 900, 925,
950, 975, 1000. If the transmitter-receiver distance is a value from this list, then interpolation of field strength is not required
and the field strength can be directly read from the graphs.
82
If the transmitter-receiver distance does not coincide with the list of distances for which the field strengths are accurately
available from the graphs, the field strength are linearly interpolated or extrapolated for the logarithm of the distance using
the following equation:
Log  d  d Low 
E d = E Low +  E Up – E Low   -------------------------------------
-
Log  d Up  d Low 
Where d Low is the lower value of the nearest tabulated distance to d , d Up is the higher value of the nearest tabulated
distance to d , E Low is the field strength value for d Low , and E Up is the field strength value for d Up .
Step 4.3: Interpolation/Extrapolation of Field Strength for Transmission Frequency
The field strength at the transmission frequency is interpolated from the graphs available for the upper and lower nominal
frequencies as follows:
Log  f  f Low 
E f = E Low +  E Up – E Low   -----------------------------------
Log  f Up  f Low 
Where f Low is the lower nominal frequency (100 MHz if f < 600 MHz, 600 MHz otherwise), f Up is the higher nominal
frequency (600 MHz if f < 600 MHz, 2000 MHz otherwise), E Low is the field strength value for f Low , and E Up is the field
strength value for f Up .
In the case of transmission frequencies below 100 MHz or above 2000 MHz, the field strength values are extrapolated from
the two nearer nominal frequency values. The above equation is used for all land paths and sea paths.
2.10.1.5 Step 5: Calculation of Correction Factors

Step 5.1: Correction for Receiver Antenna Height
The receiver antenna height correction depends on the type of path and clutter in which the receiver is located. The field
strength values given by the graphs for land paths are for a reference receiver antenna at a height, R (m), representative of
the height of the clutter surrounding the receiver, subject to a minimum height value of 10 m. Examples of reference heights
are 20 m for an urban area, 30 m for a dense urban area, and 10 m for a suburban area. For sea paths the notional value of R
is 10 m.
For land paths, the elevation angle of the arriving ray is taken into account by calculating a modified representative clutter
 1000  d  R – 15  h 1 
height R' , given by R' = Max  1 ----------------------------------------------------------- .
 1000  d – 15 
Note that for h 1  6.5  d + R , R'  R .
The different correction factors are calculated as follows:

• For land path in urban and suburban zones
 6.03 – J    for h 2  R'


C Receiver =  h
  3.2 + 6.2  Log  f    Log  ----2- for h 2  R'
  R' 
R' – h 2
With J    =  6.9 + 20  Log    – 0.1  + 1 +  – 0.1   and  = 0.0108  f   R' – h 2   arc tan  --------------- .
2
 27 
If R'  10 m , C Receiver is reduced by  3.2 + 6.2  Log  f    Log  ------ .

10
 R' 
• For land path other zones

h
C Receiver =  3.2 + 6.2  Log  f    Log  -----2-
 10
• For sea path

d 10 and d h2 are determined as distances at which at which the path has 0.6 of the first Fresnel zone just unobstructed
by the sea surface with h 2 = 10 m and variable h 2 , respectively. These distances are given by
83
D f  D h
d 10 = D 0.6  f h 1  h 2 = 10 m   and d h2 = D 0.6  f h 1 h 2  (km), respectively. Here D 0.6 = Max  0.001 ----------------
- as
 D f + D h
explained earlier.
h2
• If h 2  10 m , C Receiver =  3.2 + 6.2  Log  f    Log  ------
10
h2
• If h 2  10 m and d  d 10 , C Receiver =  3.2 + 6.2  Log  f    Log  ------
10
• If h 2  10 m and d  d 10 and d  d h2 , C Receiver = 0
h2 Log  d  d h2 
• If h 2  10 m and d  d 10 and d  d h2 , C Receiver =  3.2 + 6.2  Log  f    Log  ------   ----------------------------------
 10  Log  d 10  d h2 
Step 5.2: Correction for Short Urban/Suburban Paths
This correction is only applied when the path loss is to be calculated over land paths, over a transmitter-receiver distance less
than 15 km, in urban and suburban zones. This correction takes into account the presence of buildings in these zones. The
buildings are assumed to be of uniform height.
The correction represents a reduction in the field strength due to building clutter. It is added to the field strength and is given
by:
C Building = – 3.3  Log  f    1 – 0.85  Log  d    1 – 0.46  Log  1 + h a – R  
Where h a is the antenna height above the ground, and R is the clutter height of the clutter class where the receiver is located.
This correction is only applied when d  15 km and h 1 – R  150 m .
Step 5.3: Correction for Receiver Clearance Angle
This correction is only applied when the path loss is to be calculated over land paths, and over a transmitter-receiver distance
less than 16 km. This correction gives more precise field strength prediction over small reception areas. The correction is
added to the field strength and is given by:
C Clearance = J  '  – J   
2
Where J    =  6.9 + 20  Log    – 0.1  + 1 +  – 0.1   , ' = 0.036  f , and  = 0.065   Clearance  f
 Clearance is the clearance angle in degrees determined from:
•  : The elevation angle of the line from the receiver which just clears all terrain obstructions in the direction of the
transmitter over a distance of up to 16 km but not going beyond the transmitter.
h 1S – h 2S
•  Ref : The reference angle,  Ref = arc tan  --------------------
- .
 1000  d
Where h 1S and h 2S are the heights of the transmitter and the receiver above sea level, respectively.
2.10.1.6 Step 6: Calculation of Path Loss

First, the final field strength is calculated from the interpolated/extrapolated field strength, E f , by applying the corrections
calculated earlier. The calculated field strength is given by:
E Calc = E f + C Receiver + C Building + C Clearance
The resulting field strength is given by E = Min  E Calc E Max  , from which the path loss (basic transmission loss, L B ) is
calculated as follows:
L B = 139 – E + 20  Log  f 
2.11 Sakagami Extended Propagation Model

The Sakagami extended propagation model is based on the simplification of the extended Sakagami-Kuboi propagation
model. The Sakagami extended propagation model is valid for frequencies above 3 GHz. Therefore, it is only available in
WiMAX documents by default.
84
The Sakagami-Kuboi propagation model requires detailed information about the environment, such as widths of the streets
where the receiver is located, the angles formed by the street axes and the directions of the incident waves, heights of the
buildings close to the receiver, etc. The path loss formula for the Sakagami-Kuboi propagation model is [1]:
H 2
L Model = 100 – 7.1  Log  W  + 0.023   + 1.4  Log  h s  + 6.1  Log  H 1  – 24.37 – 3.7   --------  Log  h b  +
h b0
13   Log  f  – 3.23 
 43.2 – 3.1  Log  h b    Log  d  + 20  Log  f  + e
Where,
• W is the width (in meters) of the streets where the receiver is located
•  is the angle (in degrees) formed by the street axes and the direction of the incident wave
• hs is the height (in meters) of the buildings close to the receiver
• H1 is the average height (in meters) of the buildings close to the receiver
• hb is the height (in meters) of the transmitter antenna with respect to the observer
• hb0 is the height (in meters) of the transmitter antenna with respect to the ground level
• H is the average height (in meters) of the buildings close to the base station
• d is the separation (in kilometres) between the transmitter and the receiver
• f is the frequency (in MHz)
The Sakagami-Kuboi propagation model is valid for:
5m <W< 50 m
0° < < 90°
5m < hs < 80 m
5m < H1 < 50 m
20 m < hb < 100 m
0.5 km <d< 10 km
450 MHz <f< 2200 MHz
h b0  H
Studies [2] have shown that the Sakagami-Kuboi propagation model can be extended to frequencies higher than 3 GHz, which
also allows a simplification in terms of the input required by the model.
The path loss formula for the extended Sakagami-Kuboi propagation model is:
L Model = 54 + 40  Log  d  – 30  Log  h b  + 21  Log  f  + a
Where a is a corrective factor with three components:

H hm 
a = a  H 0  + a  W  + a  h m  = 11  Log  -----0- – 7.1  Log  ------ – 5  Log  ------
W -
20 20 1.5
• W is the width (in meters) of the streets where the receiver is located
• H0 (= hs = H1) is the height (in meters) of the buildings close to the receiver
• hb (= hb0) is the height (in meters) of the transmitter antenna with respect to the ground
• hm is the height (in meters) of the receiver antenna
• H is the average height (in meters) of the buildings close to the base station
• d is the separation (in metres) between the transmitter and the receiver
• f is the frequency (in GHz)
The extended Sakagami-Kuboi propagation model is valid for:
5m <W< 50 m
10 m < H0 < 30 m
10 m < hb < 100 m
0.1 km <d< 3 km
0.8 GHz <f< 8 GHz
1.5 m < hm < 5m
85
Studies also show that above 3 GHz, the path loss predicted by the extended model is almost independant of the input
parameters such as street widths and angles. Therefore, the extended Sakagami-Kuboi propagation model can be simplified
to the extended Sakagami propagation model:
L Model = 54 + 40  Log  d  – 30  Log  h b  + 21  Log  f  – 5  Log  h m 
The extended Sakagami propagation model is valid for:
10 m < hb < 100 m

0.1 km <d< 3 km
3 GHz <f< 8 GHz
1.5 m < hm < 5m
The path loss calculation formula of the Sakagami extended propagation model resembles the formula of the Standard
Propagation Model. In Atoll, this model is in fact a copy of the Standard Propagation Model with the following values assigned
to the K coefficients:
65.4
K1
(calculated for 3.5 GHz)
K2 40
K3 -30
K4 0
K5 0
K6 0
K7 -5
For more information on the Standard Propagation Model, see "Standard Propagation Model (SPM)" on page 65.
References:
• [1] Manuel F. Catedra, Jesus Perez-Arriaga, "Cell Planning for Wireless
Communications," Artech House Publishers, 1999.
• [2] Koshiro Kitao, Shinichi Ichitsubo, "Path Loss Prediction Formula for Urban and
Suburban Areas for 4G Systems," IEEE, 2006.
2.12 Free Space Loss

The calculation of free space loss is based on ITU 525 recommendations.
L FS = 32.4 + 20 log  f  + 20 log  d 
where,
f is the frequency in MHz,
d is the Tx-Rx distance in km,
Free space loss is stated in dB.
2.13 Diffraction
The calculation of diffraction is based on ITU 526-5 recommendations. General method for one or more obstacles (knife-edge
diffraction) is used to evaluate diffraction losses (Diffraction loss in dB). Four construction modes are implemented in Atoll.
All of them are based on this same physical principle presented hereafter, but differ in the way they consider one or several
obstacles. Calculations take the earth curvature into account through the effective Earth radius concept (K factor=1.333).
2.13.1 Knife-edge Diffraction

The procedure checks whether a knife-edge obstructs the first Fresnel zone constructed between the transmitter and the
receiver. The diffraction loss, J(), depends on the obstruction parameter (), which corresponds to the ratio of the
obstruction height (h) and the radius of the Fresnel zone (R).
86
Figure 2.15: Knife-Edge Diffraction
c0  n  d1  d2
R = --------------------------------
f   d1 + d2 
where,
n is the Fresnel zone index,
c0 is the speed of light (2.99792 x108 ms-1),

f is the frequency in Hz
d1 is the distance from the transmitter to obstacle in m,
d2 is the distance from obstacle to receiver in m.
We have:  = h
---
r
where,
R
r = -------
2
h is the obstruction height (height from the obstacle top to the Tx-Rx axis).
Hence,
2
For 1 knife-edge method, if   – 0.7 , J    = 6.9 + 20  log    – 0.1  + 1 +   – 0.1  
Else, J    = 0
In case of multiple-knife edge method, the minimum  required to estimate diffraction

loss is -0.78.
2.13.2 3 Knife-edge Deygout Method

The Deygout construction, limited to a maximum of three edges, is applied to the entire profile from transmitter to receiver.
This method is used to evaluate path loss incurred by multiple knife-edges. Deygout method is based on a hierarchical knife-
edge sorting used to distinguish the main edges, which induce the largest losses, and secondary edges, which have a lesser
effect. The edge hierarchy depends on the obstruction parameter () value.
87
1 Obstacle
Figure 2.16: Deygout Construction – 1 Obstacle
A straight line between transmitter and receiver is drawn and the height of the obstacle above the Tx-Rx axis, hi, is calculated.
The obstruction position, di, is also recorded. i are evaluated from these data. The point with the highest  value is termed
the principal edge, p, and the corresponding loss is J(p).
Therefore, we have
DiffractionLoss = J   P 
3 Obstacles
Then, the main edge (point p) is considered as a secondary transmitter or receiver. Therefore, the profile is divided in two
parts: one half profile, between the transmitter and the knife-edge section, another half, constituted by the knife-edge-
receiver section.
Figure 2.17: Deygout Construction – 3 Obstacles
The same procedure is repeated on each half profile to determine the edge with the higher . The two obstacles found, (points
t and r), are called ‘secondary edges’. Losses induced by the secondary edges, J(t) and J(r), are then calculated.
Once the edge hierarchy is determined, the total loss is evaluated by adding all the intermediary losses obtained.
Therefore, if  P  0
we have DiffractionLoss = J   P  + J   t  + J   r 
Otherwise, If  P  – 0.7 , DiffractionLoss = J   P 
88
In case of ITU 526-5 and WLL propagation models, Diffraction loss term is determined as
follows:
• If  P  – 0.78 , we have DiffractionLoss = J   P  +  J   t  + J   r    t
• Otherwise DiffractionLoss = 0
J  P  
Here, t = min  -----------
- 1
6 
2.13.3 Epstein-Peterson Method

The Epstein-Peterson construction is limited to a maximum of three edges. First, Deygout construction is applied to determine
the three main edges over the whole profile as described above. Then, the main edge height, hp, is recalculated according to
the Epstein-Peterson construction. hp is the height above a straight line connecting t and r points. The main edge position dp
is recorded and p and J(p) are evaluated from these data.
Figure 2.18: Epstein-Peterson Construction
Therefore, we have
DiffractionLoss = J   P  + J   t  + J   r 
2.13.4 Deygout Method with Correction

The Deygout method with correction (ITU 526-5) is based on the Deygout construction (3 obstacles) plus an empirical
correction, C.
Therefore, If  P  0 ,
we have DiffractionLoss = J   P  + J   t  + J   r  + C
Otherwise DiffractionLoss = J   P  + C
In case of ITU 526-5 propagation model, Diffraction loss term is determined as follows:
• If  P  – 0.78 , we have DiffractionLoss = J   P  + t   J   t  + J   r  + C 
• Otherwise DiffractionLoss = 0
J  P  
Here, t = min  -----------
- 1 and C = 8.0 + 0.04d with d = distance stated in km between

6
the transmitter and the receiver.
2.13.5 Millington Method

The Millington construction, limited to a single edge, is applied over the entire profile. Two horizon lines are drawn at the
transmitter and at the receiver. A straight line between the transmitter and the receiver is defined and the height of the
89
intersection point between the two horizon lines above the Tx-Rx axis, hh, is calculated. The position dh is recorded and then,
from these values, h and J(h) are evaluated using the same previous formulas.
Therefore, we have
DiffractionLoss = J   h 
Figure 2.19: Millington Construction
2.14 Shadow Fading Model

Propagation models predict the mean path loss as a function of transmission and reception parameters such as frequency,
antenna heights, and distance, etc. Therefore, the predicted path loss between a transmitter and a receiver is constant, in a
given environment and for a given distance. However, in reality different types of clutter may exist in the transmitter-receiver
path. Therefore, the path losses for the same distance could be different along paths that pass throught different types of
environments. The location of the receiver in different types of clutter causes variations with respect to the mean path loss
values given by the path loss models. Some paths undergo more loss while others are less obstructed and may have higher
received signal strength. The variation of path loss with respect to the mean path loss values predicted by the propagation
models, depending on the type of environment is called shadow fading (shadowing) or slow fading. "Slow" fading implies that
the variations in the path loss due to shadow fading occur comparatively slower than the fast fading effect (Rayleigh fading),
which is due to the mobile receiving multipath copies of a signal.
Different types of clutter (buildings, hills, etc.) make large shadows that cause variations in the path loss over long distances.
As a mobile passes under a shadow, the path loss to the mobile keeps varying from point to point. Shadow fading varies as
the mobile moves, while fast fading can vary even if the mobile remains at the same location or moves over very small
distances. It is crucial to account for the shadow fading in order to predict the reliability of coverage provided by any mobile
cellular system.
The shadowing effect is modelled by a log-normal (Gaussian) distribution, as shown in Figure 2.20 on page 90, whose standard
deviation  depends on the type of clutter.
Figure 2.20: Log-normal Probability Density Function
90
Different clutter types have different shadowing effects. Therefore, each clutter type in Atoll can have a different standard
deviation representing its shadowing characteristics. For different standard deviations, the shape of the Gaussian distribution
curve remains similar, as shown in Figure 2.20 on page 90.
The accuracy of this model depends upon:
• The suitability of the range of standard deviation used for each clutter class,
• The definition (bin size) of the digital map,
• How up-to-date the digital map is,
• The number of clutter classes,
• The accuracy of assignment of clutter classes.
Shadowing is applied to the predicted path loss differently depending on the technology, and whether it is applied to
predictions or simulations. The following sections explain how shadowing margins are calculated and applied to different
technology documents.
Shadowing margins are calculated for a given cell edge coverage probability. The cell edge coverage probability is the
probability of coverage at a pixel located at the cell edge, and corresponds to the reliability of coverage that you are planning
to achieve at the cell edge. For example, a cell edge coverage probability of 75 % means that the users located at the cell edge
will receive adequate signal level during 75 % of the time. Therefore, a coverage prediction with a cell edge coverage
probability of x % means that the signal level predicted on each pixel is reliable x % of the time, and the overall predicted
coverage area is reliable at least x % of the time.
References:
• [1] Saunders S. “Antennas and propagation for Wireless Communication Systems”
pp. 180-198
• [2] Holma H., Toskala A. “WCDMA for UMTS”
• [3] Jhong S., Leonard M. “CDMA systems engineering handbook” pp. 309-315, 1051-
1053”
• [4] Remy J.G., Cueugnet J., Siben C. “Systèmes de radiocommunications avec les
mobiles” pp. 309-310
• [5] Laiho J., Wacker A., Novosad T. “Radio network planning and optimisation for
UMTS” pp. 80-81
GSM GPRS EGPRS Documents
The shadowing margins are calculated as explained in "Shadowing Margin Calculation in Predictions" on page 95, and applied
to signal level or C/I as explained below.
• Signal Level-Based Predictions
Signal level-based predictions include coverage predictions (Coverage by Transmitter, Coverage by Signal Level, and
Overlapping Zones) and calculations in point analysis tabs (Profile and Reception) that require calculation of the
received signal level only, and do not depend on interference.
In these calculations (signal level calculations), a shadowing margin ( M Shadowing – model ) is applied to the received
signal level calculated for each pixel. The shadowing margin is calculated for a given cell edge coverage probability,
and depends on the model standard deviation (  model in dB) associated to the clutter class where the receiver is
located.
• Interference-Based Predictions
Interference-based predictions include coverage predictions (Coverage by C/I Level, Interfered Zones, Coverage by
GPRS/EDGE Coding Scheme, RLC Throughout/Timeslot, Application Throughput/Timeslot, Circuit Quality Indicator
Analysis) and calculations in point analysis window’s Interference tab that require calculation of the received signal
level and interference received from other base stations.
In these calculations, ( C  I calculations), the shadowing margin ( M Shadowing – C  I ) is applied to the ratio of the carrier
power (C) and the interfering signal levels (I) received from the interfering base stations. This shadowing margin is
calculated for a given cell edge coverage probability and depends on the C/I standard deviation (  C  I in dB) associated
to the clutter class where the receiver is located.
UMTS HSPA and CDMA2000 1xRTT 1xEV-DO Documents
The shadowing margins are calculated as explained in "Shadowing Margin Calculation in Predictions" on page 95 and
"Shadowing Margin Calculation in Monte-Carlo Simulations" on page 97, and applied to signal level, Ec/I0, or Eb/Nt as
explained below.
91
located.
• Interference+noise-Based Predictions
Interference+noise-based predictions include coverage predictions (Pilot Quality Analysis, Downlink Total Noise,
Service Area Analyses, Handoff Status, etc.) and point analysis (AS Analysis tab) that require calculation of the received
signal level and interference and noise received from other base stations.
In these calculations, the shadowing margins ( M Shadowing – Ec  Io , M Shadowing –  Eb  Nt  , or M Shadowing –  Eb  Nt  )

DL UL
are applied to Ec/I0 or Eb/Nt. These shadowing margins are calculated for a given cell edge coverage probability and
depend on the Ec/I0 or Eb/Nt standard deviations (  Ec  Io ,   Eb  Nt  , or   Eb  Nt  , in dB) associated to the clutter
DL UL
class where the receiver is located.
• Macro-Diversity Gains
UL DL
Atoll calculates the uplink and downlink macro-diversity gains ( G macro – diversity and G macro – diversity ) depending on
the receiver handover status. These gains are respectively taken into account to evaluate the uplink Eb/Nt in case of
soft handover and the downlink Ec/Io from best server. For detailed description of the calculation of macro-diversity
gains, please refer to "Macro-Diversity Gains Calculation" on page 97.
• Monte-Carlo Simulations
Random values for shadowing margins are calculated for each transmitter-receiver link and applied to the predicted
signal level. A shadowing margin for each transmitter-receiver link in each simulation is obtained by taking a random
value from the probability density distribution for the appropriate clutter class. The probability distribution is a log-
normal distribution as explained above.
TD-SCDMA Documents
"Shadowing Margin Calculation in Monte-Carlo Simulations" on page 97, and applied to signal level or interference+noise
predictions as explained below.
Signal level-based predictions include coverage predictions (Best Server and RSCP P-CCPCH Coverages, P-CCPCG
Pollution Analysis, Baton Handover Coverage, DwPCH and UpPCH Coverages, Cell to Cell Interference, and Scrambling
Code Interference) and calculations in point analysis tabs (Profile and Reception) that require calculation of the
located.
Interference+noise-based predictions include coverage predictions (P-CCPCH Eb/Nt and C/I Coverages, Service Area
Analsyses for downlink and uplink Eb/Nt and C/I, etc.) that require calculation of the received signal level and
interference received from other base stations.
In these calculations, the shadowing margins ( M Shadowing –  Eb  Nt  , M Shadowing –  Eb  Nt  , or

P – CCPCH DL
M Shadowing –  Eb  Nt  ) are applied to Eb/Nt. These shadowing margins are calculated for a given cell edge coverage
UL
probability and depend on the Eb/Nt standard deviations (   Eb  Nt  ,   Eb  Nt  , or   Eb  Nt  , in dB)

P – CCPCH DL UL
associated to the clutter class where the receiver is located.
92
WiMAX Documents
"Shadowing Margin Calculation in Monte-Carlo Simulations" on page 97 , and applied to signal level or C/(I+N) as explained
below.
located.
Interference-based predictions include coverage predictions (Coverage by C/(I+N) Level, Coverage by Bearer,
Coverage by Throughput, etc.) that require calculation of the received signal level and interference.
In these calculations, (C/(I+N) calculations), in addition to the shadowing margin ( M Shadowing – model ) applied to the
received signal level calculated for each pixel, the ratio M Shadowing – model – M Shadowing – C  I is applied to the
interfering signal levels (I). M Shadowing – C  I is calculated for a given cell edge coverage probability and depends on the
C/I standard deviation (  C  I in dB) associated to the clutter class where the receiver is located.
The reason why the ratio M Shadowing – model – M Shadowing – C  I is used can be understood from the following
derivation (linear, not it dB):
Inputs
• C P : The predicted received carrier power without any shadowing margin.

• I P : The predicted received interference power without any shadowing margin.
M
Shadowing – model
----------------------------------------------------------
10
• m C : Shadowing margin based on the model standard deviation ( 10 )
M Shadowing – C  I
--------------------------------------------------
10
• m C  I : Shadowing margin based on the C/I standard deviation ( 10 )
• N : Thermal noise
Calculations
The effective received carrier power is given by:
C = mC  CP
The effective C/I is given by:
C CP
--- = m C  I  -----
I IP
The above equations lead to:
mC  CP
C - = ---------------------- mC
I = ---------------------- -I
- = ----------
CP CP mC  I P
m C  I  ----- m C  I  -----
IP IP
mC
Where ----------- corresponds to M Shadowing – model – M Shadowing – C  I in dB.
mC  I
Therefore, the effective C/(I+N) is given by:
mC  CP
C - = ------------------------------------
--------------- -
I + N m
 ----------
C 
-I +N
 mC  I P 
93
LTE Documents
"Shadowing Margin Calculation in Monte-Carlo Simulations" on page 97 , and applied to signal level or C/(I+N) as explained
below.
In these calculations (signal level calculations), a shadowing margin ( M Shadowing – model ) is applied to the signal level
calculated for each pixel. The shadowing margin is calculated for a given cell edge coverage probability, and depends
on the model standard deviation (  model in dB) associated to the clutter class where the receiver is located.
Interference-based predictions include coverage predictions (Coverage by C/(I+N) Level, Coverage by Bearer,
Coverage by Throughput, etc.) that require calculation of the received signal level and received interference.
In these calculations, (C/(I+N) calculations), in addition to the shadowing margin ( M Shadowing – model ) applied to the
signal level calculated for each pixel, the ratio M Shadowing – model – M Shadowing – C  I is applied to the interfering signal
levels (I). M Shadowing – C  I is calculated for a given cell edge coverage probability and depends on the C/I standard
deviation (  C  I in dB) associated to the clutter class where the receiver is located.
The reason why the ratio M Shadowing – model – M Shadowing – C  I is used can be understood from the following
derivation (linear, not it dB):
Inputs
• C P : The predicted received carrier power without any shadowing margin.

• I P : The predicted received interference power without any shadowing margin.
M Shadowing – model
----------------------------------------------------------
10
• m C : Shadowing margin based on the model standard deviation ( 10 )
M Shadowing – C  I
--------------------------------------------------
10
• m C  I : Shadowing margin based on the C/I standard deviation ( 10 )
• N : Thermal noise
Calculations
The effective received carrier power is given by:
C = mC  CP
The effective C/I is given by:
C C
--- = m C  I  ----P-
I IP
The above equations lead to:
mC  CP
C - = ---------------------- mC
I = ---------------------- -I
- = ----------
CP CP mC  I P
m C  I  ----- m C  I  -----
IP IP
mC
Where ----------- corresponds to M Shadowing – model – M Shadowing – C  I in dB.
mC  I
Therefore, the effective C/(I+N) is given by:
94
mC  CP
C - = ------------------------------------
--------------- -
I + N m
 ----------
C
-  I P + N
 mC  I 
2.14.1 Shadowing Margin Calculation

The following sections describe the calculation method used for determining different shadowin margins.
The following shadowing margins are calculated using the method described below:
Network Type Standard Deviation MShadowing Applied to

 model M Shadowing – model C
GSM GPRS EGPRS
C  I M Shadowing – C  I C/I
 Ec  Io M Shadowing – Ec  Io Ec/I0
UMTS HSPA
  Eb  Nt  M Shadowing –  Eb  Nt  Eb/Nt (DL)
DL DL
  Eb  Nt  M Shadowing –  Eb  Nt  Eb/Nt (UL)

UL UL
 Ec  Io M Shadowing – Ec  Io Ec/I0
CDMA2000
DL DL

UL UL
  Eb  Nt  M Shadowing –  Eb  Nt  Eb/Nt P-CCPCH

P – CCPCH P – CCPCH
TD-SCDMA
DL DL

UL UL
 model M Shadowing – model C and C/(I+N)

WiMAX
C  I M Shadowing – C  I C/(I+N)
 model M Shadowing – model C and C/(I+N)

LTE
C  I M Shadowing – C  I C/(I+N)
2.14.1.1 Shadowing Margin Calculation in Predictions

Shadowing margins, MShadowing, are calculated from standard deviation values defined for the clutter class where the pixel
(probe mobile) is located, and required cell edge coverage probability, and applied to the path loss, Lpath.
Shadowing Error PDF (1 Signal)
The measured path loss in dB can be expressed as a Gaussian random variable:

L = Lpath +  dB  G  0 1 
where,
• Lpath is the predicted path loss,
95
• dB is the user-defined standard deviation of the error,

• G(0,1) is a zero-mean unit-variance Gaussian random variable.
Therefore, the probability density function (pdf) for the random (shadowing) part of path loss is:
2
x
– --------------
2
1 2
dB
p L  x  = --------------------  e
 dB 2
The probability that the shadowing error exceeds z dB is

2
x -
  – -------------
2
1 2 dB
PL  x  z  =  pL  x  dx = --------------------  e
 dB 2  dx
z z
Normalising x by dividing it bydB:
 2
x
– -----
dx = Q  --------
1 z

2
P L  x  z  = ----------  e
2   dB
z
----------
 dB
where Q is the complementary cumulative function.

To ensure a given cell edge coverage probability, R L , for the predicted value, a shadowing margin, M Shadowing , is added to
the link budget.
Confidence in the prediction can be expressed as:
C d = P' Tx – L  P rec  L  P' Tx – P rec  G  0 1    dB  M Shadowing
where,
• P rec is the signal level predicted at the receiver. P rec = P' Tx – L path – M Shadowing
• P' Tx = EIRP + G antRx – L Rx
• EIRP is the effective isotropic radiated power of the transmitter.
• L Rx are receiver losses.
• G antRx is the receiver antenna gain.
The shadowing margin is calculated such that:

M Shadowing
P  C d  P rec  = R L  M Shadowing  = 1 – P L  x – M Shadowing  0  = 1 – Q  --------------------------
-
  dB 
A lookup table is used for mapping the values of Q vs. a set of cell edge coverage probabilities.
M Shadowing
Figure 2.21: Normalised Margin M arg in = --------------------------
-
 dB
In interference-based predictions, where signal to noise ratio is calculated, the shadowing margin is only applied to the signal
from the interfered transmitter (C). We consider that the interference value is not altered by the shadowing margin. Random
variations also exist in the interfering signals, but taking only the average interference gives accurate results. [3] explains how
a certain level of interference is maintained by congestion control in CDMA-based networks.
96
2.14.1.2 Shadowing Margin Calculation in Monte-Carlo Simulations

Shadowing margins, MShadowing, are calculated from standard deviation values defined for the clutter class where the pixel
(probe mobile) is located, and required cell edge coverage probability, and added to the path loss, Lpath.
Random values are generated during Monte-Carlo simulation. Each user is assigned a service, a mobility type, an activity
status, a geographic position and a random shadowing value.
For each link, path loss (L) can be broken down to L = L path +  .
Here,  is a zero mean gaussian random variable G  0  dB  representing variation due to shadowing. It can be expressed as
the sum of two uncorrelated zero mean gaussian random variables,  L and  P .  L models the error related to the receiver’s
location (surrounding environment), and remains the same for all links between the receiver and the base stations from which
it is receiving signals.  P models the error related to the path between the transmitter and the receiver.
Therefore, in case of two links, we have:

1
 1 =  L +  P for link 1
2
 2 =  L +  P for link 2
i
Standard deviations of  L   L  and  P   P  can be calculated from  i , the model standard deviation   model  , and the
correlation coefficient    between  1 and  2 .
Assuming all  P have the same standard deviations, we have:
2 2 2
 model =  L +  P
2
L
 = ---------------
-
2
 model
Therefore,
2 2
 P =  model   1 –  
2 2
 L =  model  
 is set to 0.5 in Atoll, which gives:
 model  model
 L = ---------------
- and  P = ---------------
-
2 2
Receiver
Therefore, to model shadowing error common to all the signals received at a receiver ( E Shadowing – model ), values are
randomly generated for each receiver. These values have a zero-mean gaussian distribution with a standard deviation of
 model
- , where   model  is the model standard deviation associated with the receiver’s clutter class.
---------------
 2 
Next, Atoll generates another random value for each transmitter-receiver pair. This values represents the shadowing error
Path
not related to the location of the receiver ( E Shadowing – model ). These values also have a zero-mean gaussian distribution with
 model
a standard deviation  ---------------
- .
 2 
So, we have:
Receiver Path
E Shadowing – model = E Shadowing – model + E Shadowing – model
Random shadowing error has its mean value at zero. Hence, this shadowing modelling method has no impact on the simulated
network load. On the other hand, as shadowing errors on the transmitter-receiver links are uncorrelated, the method
influences the calculated macro-diversity gain in case the mobile is in soft handover.
2.14.2 Macro-Diversity Gains Calculation

The following sections explain how uplink and downlink macro-diversity gains are calculated in UMTS HSPA and CDMA2000
1xRTT 1xEV-DO documents for predictions and AS Analysis tab of the point analysis tool.
97
The calculation and use of macro-diversity gains can be disabled through the Atoll.ini file.
For more information, see the Administrator Manual.
2.14.2.1 Uplink Macro-Diversity Gain Evaluation

In UMTS HSPA and CDMA2000 1xRTT 1xEV-DO, mobiles may be in soft handoff (mobile connected to cells located on different
sites). In this case, we can consider the shadowing error pdf described below.
2.14.2.1.1 Shadowing Error PDF (n Signals)

For each link, path loss (L) can be broken down as:
L = L path + 
 is a zero mean gaussian random variable G  0  dB  representing variation due to shadowing. It can be expressed as the
sum of two uncorrelated zero mean gaussian random variables,  L and  P .  L models error related to the receiver local
environment; it is the same whichever the link.  P models error related to the path between transmitter and receiver.

1
 1 =  L +  P for the link 1
2
Knowing  i , the uplink Eb/Nt standard deviation    Eb  Nt   and the correlation coefficient  between  1 and  2 , we can
UL
calculate standard deviations of  L   L  and  P   P  (assuming all  P have the same standard deviations).
We have:
2 2 2
  Eb  Nt  = L + P
UL
2
L
 = -----------------------
-
2
  Eb  Nt 
UL
Therefore,
2 2
 P =   Eb  Nt   1 – 
UL
2 2
 L =   Eb  Nt  
UL
2 Signals Without Recombination
In technologies supporting soft handoff (UMTS and CDMA2000), cell is interference limited. As for one link, to ensure a
required cell edge coverage probability R L for the prediction, we add to each link budget a shadowing margin,
2signals
M Shadowing –  Eb  Nt  .
UL
Prediction reliability in order to have Eb/Nt higher or equal to Eb/Nt from the best server can be expressed as:
Cd
1 1
-------1 = P' Tx1 – L 1 – N 1  CI pred   1  P' Tx1 – L path – N 1 – CI pred
N1 1
or
Cd
1 1
-------2 = P' Tx2 – L 2 – N 2  CI pred   2  P' Tx2 – L path – N 2 – CI pred
N2 2
where
i
CI pred is the quality level (signal to noise ratio) predicted at the receiver for link i.
Ni is the noise level for link i.

We note:
98
2signals i
M Shadowing –  Eb  Nt  = P' Txi – L path – N i – CI pred
UL i
and
2 1 2
 1 = CI pred – CI pred
2
 1 is the minimum needed margin on each link.
Therefore, the probability of having a quality at least equal to the best predicted one is:
noMRC 2signals  Cd 1
Cd
1 
RL  M Shadowing –  Eb  Nt   = 1 – P L1 L2  -------1  CI pred -------2  CI pred
UL  1N N 2 
noMRC 2signals 2signals 2signals 2

RL  M Shadowing –  Eb  Nt   = 1 – P    1  M Shadowing –  Eb  Nt    2  M Shadowing –  Eb  Nt  – 1 
UL 1  2 UL UL
1 2
We can express it using  L ,  P and  P
P 
 
2signals
M Shadowing –  Eb  Nt    2  M Shadowing –  Eb  Nt 
2signals
–  1  L =  L
2
1  2  1 UL UL
1 2signals 2 2signals 2
= P  L   P 1 2  P  M Shadowing –  Eb  Nt  –  L  P  M Shadowing –  Eb  Nt  – 1 – L 
L  P  P UL UL
P 
 
2signals
M Shadowing –  Eb  Nt    2  M Shadowing –  Eb  Nt 
2signals
–  1 L =  L
2
1  2  1 UL UL 
= P    L   P    P  M Shadowing –  Eb  Nt  –  L  P    P  M Shadowing –  Eb  Nt  – 1 –  L 
L P UL P UL
noMRC 2signals
RL  M Shadowing –  Eb  Nt  
UL

 
=  1 – P    L   P    P  M Shadowing –  Eb  Nt  –  L   P    P  M Shadowing –  Eb  Nt  –  1 –  L  d L

 L P UL P UL 
 – 
i 2signals
P    P  M Shadowing –  Eb  Nt  – L 
P UL
2
 
–x -
--------- 2signals

 2 P
2
 M Shadowing –  Eb  Nt  UL –  L 
 1
=  -----------------  e dx = Q  ----------------------------------------------------------------
P  
  P 2  2signals 
 M –  L 
 Shadowing –  Eb  Nt  UL 
Then, we have:
noMRC 2signals
UL
 2signals 2signals 2
  M Shadowing –  Eb  Nt UL –  L  M Shadowing –  Eb  Nt  UL –  1 –  L 
=  1 – P    L   Q  ----------------------------------------------------------------  Q  ---------------------------------------------------------------------------- d L

 L  P   P  
 – 
If we introduce user defined standard deviation    Eb  Nt   and correlation coefficient    , and consider that P  is a
UL L
Gaussian pdf:
noMRC 2signals
UL
2
  –xL 2signals
 M Shadowing – x L   Eb  Nt  
2signals
 M Shadowing – x L   Eb  Nt   –  1 
2
 --------- –  Eb  Nt  –  Eb  Nt 
 Q  -----------------------------------------------------------------------------------------------  Q  ------------------------------------------------------------------------------------------------------------ dx L
  
1

2 UL UL UL UL
=  1 – ---------- e
 2    Eb  Nt  1–     Eb  Nt  1–  
 – UL UL
99
n Signals Without Recombination
We can generalize the previous expression to n signals (n is the number of available signals - Atoll may consider up to 3
signals):
noMRC nsignals
UL
2
  –xL
 M nsignals – x L   Eb  Nt    M nsignals – x L   Eb  Nt   –  1 
2
 --------- Shadowing –  Eb  Nt  Shadowing –  Eb  Nt 
 Q  -----------------------------------------------------------------------------------------------  Q  ------------------------------------------------------------------------------------------------------------ dx L
  
1

2 UL UL UL UL
=  1 – ---------- e
 2    Eb  Nt  1–     Eb  Nt  1–  
 – UL UL
The case where softer handoff occurs (two signals from co-site cells) is equivalent to the one signal case. The Softer/soft case
is equivalent to the two signals case. For the path associated with the softer recombination, we will use combined SNR to
calculate the availability of the link.
Correlation Coefficient Determination
There is currently no agreed model for predicting correlation coefficient    between  1 and  2 . Two key variables influence
correlation:
• The angle between the two signals. If this angle is small, correlation is high.
• The relative values of the two signal lengths. If angle is 0 and lengths are the same, correlation is zero. Correlation is
different from zero when path lengths differ.
A simple model has been found [1]:
T 
 =  -----
D1
------- when  T    
 D2
 T is a function of the mean size of obstacles near the receiver and  is also linked to the receiver environment.
In a normal handover status, assuming a hexagonal design for sites,  is close to  (+/- /3) and D1/D2 is close to 1.

In [1,5],  = 0.5 when  = 0.3 and  T = ------ .
10
In Atoll,  is set to 0.5.
2.14.2.1.2 Uplink Macro-Diversity Gain

UL
Atoll determines the uplink macro-diversity gain ( G macro – diversity ) from the shadowing margins calculated in case of one
signal and n signals.
Therefore, we have:
UL nsignals
G macro – diversity = M Shadowing –  Eb  Nt  – M Shadowing –  Eb  Nt 
UL UL
Where n is the number of cell-mobile signals.
2.14.2.2 Downlink Macro-Diversity Gain Evaluation

In UMTS HSPA and CDMA2000 1xRTT 1xEV-DO, in case of soft handoff, mobiles are able to switch from one cell to another if
the best pilot drastically fades. To model this function, we have to consider the probability of fading over the shadowing
margin, both for the best signal and for all the other available signals, in the shadowing margin calculation.
Let us consider the shadowing error pdf described below.
2.14.2.2.1 Shadowing Error PDF (n Signals)

For each link, path loss (L) can be broken down as:
L = L path + 
 is a zero mean gaussian random variable G  0  dB  representing variation due to shadowing. It can be expressed as the
sum of two uncorrelated zero mean gaussian random variables,  L and  P .  L models the error related to the receiver local
environment, which is the same for all links.  P models the error related to the path between the transmitter and the
receiver.
100
1
2
Knowing  i , the Ec/Io standard deviation   Ec  I o  and the correlation coefficient  between  1 and  2 , we can calculate
standard deviations of  L   L  and  P   P  (assuming all  P have the same standard deviations).
We have:
2 2 2
 Ec  I o =  L +  P
2
L
 = -------------
-
2
 Ec  I o
Therefore,
2 2
 P =  Ec  I o   1 –  
2 2
 L =  Ec  I o  
2 Available Signals
In technologies supporting soft handoff (UMTS and CDMA2000) cells are interference limited. As for one link, to ensure a
2signals
required cell edge coverage probability R L for the prediction, we add a shadowing margin, M Shadowing – Ec  Io , to each link
budget.
Prediction reliability to have -----   -----

Ec Ec
for the best server can be expressed as:
Io Io pred
Ec Ec 1 Ec 1
-------1- = P pilot – L 1 – Io   -----   1  P pilot – L m – Io –  -----
Io 1  Io  pred 1 1  Io  pred
Or
Ec Ec 1 Ec 1
-------2- = P pilot – L 2 – Io   -----   2  P pilot – L m – Io –  -----
Io 2  Io  pred 2 2  Io  pred
We note:
1
M Shadowing – Ec  Io = P pilot – L m – Io –  -----
2signals Ec
i i  Io  pred
Ec 1 Ec 2
 1 =  ----- –  -----
2
Io pred Io pred
2
 1 is the minimum needed margin on each link.
Therefore, probability of having a quality at least equal to the best predicted one is:
Ec 1 Ec 1 Ec 2 Ec 1 
 M Shadowing – Ec  Io  = 1 – P L1 L2  --------   -----  --------   -----
noMRC 2signals
RL
Io Io pred Io Io pred
noMRC 2signals 2signals 2signals 2

RL  M Shadowing – Ec  Io  = 1 – P 1 2   1  M Shadowing – Ec  Io  2  M Shadowing – Ec  Io –  1 
1 2
We can express it by using  L ,  P and  P
2signals 2signals 2
P 1 2   1  M Shadowing – Ec  Io  2  M Shadowing – Ec  Io –  1  L =  L 
= P  L   P 1 2  P  M Shadowing – Ec  Io –  L P  M Shadowing – Ec  Io –  1–  L 
L  P  P
2signals 2signals 2
P 1 2   1  M Shadowing – Ec  Io  2  M Shadowing – Ec  Io –  1  L =  L
= P    L   P    P  M Shadowing – Ec  Io –  L   P    P  M Shadowing – Ec  Io –  1 –  L 
L P P
101
noMRC 2signals
RL  M Shadowing – Ec  Io 

 PL  L   PP  P  MShadowing – Ec  Io – L   PP  P  MShadowing – Ec  Io – 1 – L  dL

= 1–
–
2
–x -
 ---------
2 2signals
1 2  M Shadowing – Ec  Io –  L

i 2signals P
P    P  M Shadowing – Ec  Io –  L  = ----------------- e dx = Q  -----------------------------------------------------
-
P  P 2  P 
 SHO –  L
Then, we have:
 2signals 2signals 2
 M Shadowing – Ec  Io –  L  M Shadowing – Ec  Io –  1 –  L

noMRC 2signals
RL  M Shadowing – Ec  Io  = 1– P    L   Q  ------------------------------------------------------  Q  ------------------------------------------------------------------ d L
L   P   P 
–
If we introduce a user defined Ec/Io standard deviation    and a correlation coefficient    and consider that P  is a
L
Gaussian pdf:
noMRC 2signals
2
 –xL
2signals 2signals 2
1
---------  M Shadowing – Ec  Io – x L  Ec  I o   M Shadowing – Ec  Io –  1 – x L  Ec  I o 

2
= 1 – ---------- e  Q  ----------------------------------------------------------------------------  Q  ---------------------------------------------------------------------------------------- dx L
2   Ec  I o 1 –     Ec  I o 1 –  
–
n Available Signals
We can generalize the previous expression for n signals (n is the number of available signals - Atoll may consider up to 3
signals):
noMRC nsignals
2
 –xL n
nsignals nsignals i
1
---------  M Shadowing – Ec  Io – x L  Ec  I o   M Shadowing – Ec  Io –  1 – x L  Ec  I o 
 
2
= 1 – ---------- e  Q  ---------------------------------------------------------------------------
- x Q  ---------------------------------------------------------------------------------------- dx L
2   Ec  I o 1 –     Ec  I o 1 –  
– i=2
2
 1 =1 dB
2
 1 =5 dB
2
 1 =10 dB
Figure 2.22: Margin - Probability (Case of 2 Signals)
102
2 signals
3
 1 =5 dB
3
 1 =10 dB
Figure 2.23: Margin - Probability (Case of 3 Signals with sigma = 8dB, delta1 = 1dB)
2 signals
3
 1 =5 dB
3
 1 =10 dB
Figure 2.24: Margin - Probability (Case of 3 Signals with sigma = 8dB, delta1 = 2dB)
Correlation Coefficient Determination
For further information about determination of the correlation coefficient, please see "Correlation Coefficient Determination"
on page 103.
2.14.2.2.2 Downlink Macro-Diversity Gain

DL
Atoll determines the downlink macro-diversity gain ( G macro – diversity ) from the shadowing margins calculated in case of one
signal and n signals.
Therefore, we have:
DL nsignals
G macro – diversity = M Shadowing – Ec  Io – M Shadowing – Ec  Io
Where n is the number of available signals.
2.15 Path Loss Matrices

Atoll is able to calculate two path loss matrices per transmitter, a first matrix over a smaller radius computed with a high
resolution and a propagation model (main matrix), and a second matrix over a larger radius computed with a low resolution
and another propagation model (extended matrix).
To be considered for calculations, a transmitter must fulfil the following conditions:
• It must be active,
• It must satisfy filter criteria defined in the Transmitters folder, and
• It must have a calculation area.
In the rest of the document, a transmitter fulfilling the conditions detailed above will be called TBC transmitter.
The path loss matrix size of a TBC transmitter depends on its calculation area. Atoll determines a path loss value ( L path ) on
each calculation bin (calculation bin is defined by the resolution) of the calculation area of the TBC transmitter. You may have
one or two path loss matrices per TBC transmitter.
103
2.15.1 Calculation Area Determination

Transmitter calculation area is made of a rectangle or a square depending on transmitter calculation radius and the
computation zone.
Calculation radius enables Atoll to define a square around the transmitter. One side of the square equals twice the entered
calculation radius.
Since the computation zone can be made of one or several polygons, transmitter calculation area corresponds to the
intersection area between its calculation square and the rectangle containing the computation zone area(s).
Figure 2.25: Example 1: Single Calculation Area
Figure 2.26: Example 2: Multiple Calculation Areas
Computation zone
Rectangle containing the computation zone(s)
Calculation area defined (square)
Transmitter
Actual calculation area on which Atoll calculates path losses
2.15.2 Validity of Path Loss Matrices

Most geographic data modifications and some radio data changes make path loss matrices invalid. This table lists these
modifications and also changes that have an impact only on coverage predictions.
Modification Matrix validity Impact on Calculate Force calculation

Path loss
Frequency Invalid Sufficient Not necessary
matrices
Antenna coordinates Path loss
Invalid Sufficient Not necessary
(site coordinate: X and Y, Dx and Dy) matrices
Path loss
Antennaa height Invalid
matrices
Sufficient Not necessary
104
Modification Matrix validity Impact on Calculate Force calculation

Path loss
Antennaa pattern Invalid
matrices
Path loss
Downtilta Invalid
matrices
Path loss
Azimutha Invalid
matrices
Path loss
% Power (secondary antennas) Invalid Sufficient Not necessary
matrices
Path loss
Site position/altitude Invalid Sufficient Not necessary
matrices
Path loss
Grid resolution (main or extended) Invalid Sufficient Not necessary
matrices
Path loss
Propagation model (main or extended) Invalid Sufficient Not necessary
matrices
Path loss
Propagation model parameters Invalid Sufficient Not necessary
matrices
Calculation areas Coverage
Valid Sufficient Not necessary
(Calculation areas gets smaller) predictions
Calculation areas Path loss
Invalid Sufficient Not necessary
(Calculation areas gets larger) matrices
Path loss
Receiver height Invalid Sufficient Not necessary
matrices
Coverage
Receiver losses Valid Sufficient Not necessary
predictions
Coverage
Receiver gain Valid Sufficient Not necessary
predictions
Coverage
Receiver antenna Valid Sufficient Not necessary
predictions
Path loss
Geographic layer order Invalid
matrices Insufficientb Necessary
Path loss
Geographic file resolution Invalid
Path loss
New DTM map Invalid
Path loss
Clutter class edition Invalid
Coverage
Coverage prediction resolution Valid Sufficient Not necessary
predictions
Coverage
Cell edge coverage probability Valid Sufficient Not necessary
predictions
Coverage
Coverage prediction conditions Valid Sufficient Not necessary
predictions
Coverage
Coverage prediction display options Valid Sufficient Not necessary
predictions
a. Modification of any parameter related to main or other antennas makes matrix invalid.
b. Except if this action has an impact on the site positions/altitudes.
2.15.3 Path Loss Tuning

Atoll can tune path loss matrices obtained from propagation results by the use of real measurements (CW Measurements or
Test Mobile Data). For each measured transmitter, Atoll tries to merge measurements and predictions on the same points
and to smooth the surrounding points of the path loss matrices for homogeneity reasons. A transmitter path loss matrix can
be tuned several times by the use of several measurement paths. All these tuning paths are stored in a catalogue. This
catalogue is stored under a .tuning folder containing a .dbf file and one .pts file per tuned transmitter. Since a tuning file can
contain several measurement paths, all these measurements are added to the tuning file.
105
For more information on the tuning files, see the Administrator Manual.
2.15.3.1 Transmitter Path Loss Tuning

The same algorithm is used for CW Measurement and Test Mobile Data. It is also the same for main and extended matrices.
Path Losses tuning will be done using two steps.
1. Total matrix correction
A mean error is calculated between each measured value and the corresponding bin in the pathloss matrix. Mean
error is calculated for each pathloss matrix (main and extended) of each transmitter. This mean error is then applied
to all the matrix bins. This tuning is done to smooth the local corrections (step 2) of measured values and not the tuned
bins.
2. Local correction for each measured value
For each measured value, an ellipse is used to define the pathloss area which has to be tuned. The main axis of the
ellipse is oriented to the transmitter.The ellipse is user-defined by two parameters:
• The radius of the axis parallel to the Profile (A)
• The radius of the axis perpendicular to the Profile (B)
Let’s take M a measurement value and P i the path loss value at point i, before any tuning.
M is limited by the minimum measurement threshold defined in the interface.
The squared elliptic distance between i and M is given by:

2 2
 Xi – XM   Yi – YM 
D i = -----------------------
- + -----------------------
-
2 2
A B
Where:
X i and X M are the X-coordinates of i and M respectively
Y i and Y M are the Y-coordinates of i and M respectively
The mean error for the first tuning is given by:
E =  --- 
1
 n  ei
i
Where e i is the error between measurement and prediction at point i
E is limited by the maximum total correction defined in the interface.
Then, the path loss value is tuned using E:

Pi = Pi +E
new old
Finally, a second tuning ( R i ) is applied where:
Ri =  1 – Di    M – g – Pi  so R i =  1 – D i    M – g –  P i + E
new old
Where g is (measurement gain - losses).
R i is limited by the maximum local correction defined in the interface.
So, the final tuned path loss is:

Pi = Pi + R i so P i = Pi + E + Ri
tuned new tuned old
106
When several ellipses overlap a pathloss bin, the final tuned path loss is given by:
 
 
  1 – d j P j
tuned


j
Pi = -------------------------------------------------
tuned  
n –
  d j

j
Where n is the number of overlapping ellipses
2.15.3.2 Repeater Path Loss Tuning

In the case of repeaters, Atoll provides only a composite measured value per pixel which is a combination of the contribution
of both a transmitter and one or several repeaters. In order to tune the path loss matrices of donor transmitters and repaters,
its is mandatory to split the contribution of each element in the measured value as starting point.
Let’s take M the measured value.
M = Md + Mr
where :
M d represents the contribution of the donor transmitter in the measured value.
M r represents the contribution of the repeater in the measured value.
All the values are used in Watts.
If C d and C r represent respectively the filtered signal level from the donor transmitter and the repeater on a pixel, one can
define the contribution of each element as follows:
Cd Cr
M d = M  ---------------- and M r = M  ---------------- .
Cd + Cr Cd + Cr
Following the path loss tuning process described in "Transmitter Path Loss Tuning" on page 106, the donor transmitter (resp.
the repeater) is then tuned using M d (resp. M r ) values.
2.16 File Formats

2.16.1 Path Loss Matrix File Format
When path loss matrices are stored externally, i.e., outside the ATL file, the path loss matrices folder contains a ‘pathloss.dbf’
file containing the calculation parameters of the transmitters and one LOS (path loss results) file per calculated transmitter.
The path loss matrices folder also contains a LowRes folder with another pathloss.dbf file and one LOS (path loss results) file
per transmitter that has an extended path loss matrix.
The formats of the pathloss.dbf and LOS files are described here.
2.16.1.1 Pathloss.dbf File Format

The pathloss.dbf file has a standard DBF (dBase III) format. The file can be opened in Microsoft Access, but it should not be
modified without consulting the Forsk customer support.
For general information, the format of DBF files in any Xbase language is as follows:
Notations used in the following tables: FS = FlagShip; D3 = dBaseIII+; Fb = FoxBase; D4 =

dBaseIV; Fp = FoxPro; D5 = dBaseV; CL = Clipper
107
DBF Structure
Byte Description Remarks

0...n DBF header (see next part for size, byte 8)
1st record of fixed length (see next parts); 2nd record (see next part for size, byte10) …; last
n+1 If .dbf is not empty
record
last optional: 0x1a (eof byte)
DBF Header
The DBF header size is variable and depends on the field count.
Byte Size Contents Description Applies to

00 1 0x03 plain .dbf FS, D3, D4, D5, Fb, Fp, CL
0x04 plain .dbf D4, D5 (FS)
0x05 plain .dbf D5, Fp (FS)
0x43 with .dbv memo var size FS
0xB3 with .dbv and .dbt memo FS
0x83 with .dbt memo FS, D3, D4, D5, Fb, Fp, CL
0x8B with .dbt memo in D4 format D4, D5
0x8E with SQL table D4, D5
0xF5 with .fmp memo Fp
01 3 YYMMDD Last update digits All
04 4 ulong Number of records in file All
08 2 ushort Header size in bytes All
10 2 ushort Record size in bytes All
12 2 0,0 Reserved All
14 1 0x01 Begin transaction D4, D5
0x00 End Transaction D4, D5
0x00 ignored FS, D3, Fb, Fp, CL
15 1 0x01 Encrypted D4, D5
0x00 normal visible All
16 12 0 (1) multi-user environment use D4,D5
28 1 0x01 production index exists Fp, D4, D5
0x00 index upon demand All
29 1 n language driver ID D4, D5
0x01 codepage437 DOS USA Fp
0x02 codepage850 DOS Multi ling Fp
0x03 codepage1251 Windows ANSI Fp
0xC8 codepage1250 Windows EE Fp
30 2 0,0 reserved All
32 n*32 Field Descriptor (see next paragraph) all
+1 1 0x0D Header Record Terminator all
Field descriptor array in the DBF header (32 bytes for each field):
0 11 ASCI field name, 0x00 termin all

11 1 ASCI field type (see next paragraph) all
12 4 n,n,n,n Fld address in memory D3
108

n,n,0,0 offset from record begin Fp
0,0,0,0 ignored FS, D4, D5, Fb, CL
16 1 byte Field length, bin (see next paragraph) all \ FS,CL: for C field type
17 1 byte decimal count, bin all / both used for fld lng
18 2 0,0 reserved all
20 1 byte Work area ID D4, D5
0x00 unused FS, D3, Fb, Fp, CL
21 2 n,n multi-user dBase D3, D4, D5
0,0 ignored FS, Fb, Fp, CL
23 1 0x01 Set Fields D3, D4, D5
0x00 ignored FS, Fb, Fp, CL
24 7 0...0 reserved all
31 1 0x01 Field is in .mdx index D4, D5
Field type and size in the DBF header, field descriptor (1 byte):
Size Type Description/Storage Applies to

ASCII (OEM code page chars)
C 1...n Char all
rest= space, not \0 term.
n = 1...64kb (using deci count) FS
n = 1...32kb (using deci count) Fp, CL
n = 1...254 all
D8 Date 8 ASCII digits (0...9) in the YYYYMMDD format all
ASCII digits (-.0123456789)
F 1...n Numeric variable pos. of float.point FS, D4, D5, Fp
n = 1...20
ASCII digits (-.0123456789)
N 1...n Numeric all
fix posit/no float.point
n = 1...20 FS, Fp, CL
n = 1...18 D3, D4, D5, Fb
L1 Logical ASCII chars (YyNnTtFf space) FS, D3, Fb, Fp, CL
ASCII chars (YyNnTtFf?) D4, D5 (FS)
10 digits repres. the start block posit. in .dbt file, or 10
M 10 Memo all
spaces if no entry in memo
Variable, bin/asc data in .dbv
4bytes bin= start pos in memo
4bytes bin= block size
V 10 Variable FS
1byte = subtype
1byte = reserved (0x1a)
10 spaces if no entry in .dbv
binary data in .ftp
P 10 Picture Fp
structure like M
binary data in .dbt
B 10 Binary D5
structure like M
OLE objects
G 10 General D5, Fp
structure like M
22 short int binary int max +/- 32767 FS
44 long int binary int max +/- 2147483647 FS
88 double binary signed double IEEE FS
Each DBF record (fixed length):
109
Byte Size Description Applies to

0 1 deleted flag "*" or not deleted " " All
x-times contents of fields, fixed length, unterminated.
1…n 1… All
For n, see (2) byte 10…11
2.16.1.2 Pathloss.dbf File Contents

The DBF file provides information that is needed to check validity of each path loss matrix.
Field Type Description

TX_NAME Text Name of the transmitter
FILE_NAME Text Name (and optionally, path) of .los file
MODEL_NAME Text Name of propagation model used to calculate path loss
Signature (identity number) of model used in calculations. You can check it in the
propagation model properties (General tab).
The Model_SIG is used for the purpose of validity. A unique Model_SIG is assigned to
MODEL_SIG Text each propagation model. When model parameters are modified, the associated model
ID changes. This enables Atoll to detect path loss matrix invalidity. In the same way, two
identical propagation models in different projects do not have the same model IDa.
ULXMAP Float X-coordinate of the top-left corner of the path loss matrix upper-left pixel
ULYMAP Float Y-coordinate of the top-left corner of the path loss matrix upper-left pixel
RESOLUTION Float Resolution of path loss matrix in metre
NROWS Float Number of rows in path loss matrix
NCOLS Float Number of columns in path loss matrix
FREQUENCY Float Frequency band
TILT Float Transmitter antenna mechanical tilt
AZIMUTH Float Transmitter antenna azimuth
TX_HEIGHT Float Transmitter height in metre
TX_POSX Float X-coordinate of the transmitter
TX_POSY Float Y-coordinate of the transmitter
ALTITUDE Float Ground height above sea level at the transmitter in metre
RX_HEIGHT Float Receiver height in metre
Logical number referring to antenna pattern. Antennas with the same pattern will have
ANTENNA_SI Float
the same number.
Maximum path loss stated in 1/16 dB. This information is used, when no calculation
MAX_LOS Float
radius is set, to check the matrix validity.
CAREA_XMIN Float Lowest x-coordinate of centre pixel located on the calculation radiusb
CAREA_XMAX Float Highest x-coordinate of centre pixel located on the calculation radius
CAREA_YMIN Float Lowest y-coordinate of centre pixel located on the calculation radius
CAREA_YMAX Float Highest y-coordinate of centre pixel located on the calculation radius
WAREA_XMIN Float Lowest x-coordinate of centre pixel located in the computation zonec
WAREA_XMAX Float Highest x-coordinate of centre pixel located in the computation zone
WAREA_YMIN Float Lowest y-coordinate of centre pixel located in the computation zone
WAREA_YMAX Float Highest y-coordinate of centre pixel located in the computation zone
Locking status
LOCKED Boolean 0: path loss matrix is not locked
1: path loss matrix is locked.
Atoll indicates if losses due to the antenna pattern are taken into account in the path
loss matrix.
INC_ANT Boolean
0: antenna losses not taken into account
1: antenna losses included
110
a. In order to benefit from the calculation sharing feature, users must retrieve the propagation models from the same
central database. This can be done using the Open from database command for a new document or the Refresh
command for an existing one. Otherwise, Atoll generates different model_ID (even if same parameters are applied on
the same kind of model) and calculation sharing become unavailable due to inconsistency.
b. These coordinates enable Atoll to determine the area of calculation for each transmitter.
c. These coordinates enable Atoll to determine the rectangle including the computation zone.
2.16.1.3 LOS File Format

The LOS (path loss results) files are binary files with a standard row-column structure. Data are stored starting from the
southwest to the northeast corner of the area. The file contains 16-bit signed integer values in the range [-32768; +32767]
with a 1/16 dB precision. "No data" values are represented by +32767.
2.16.2 Path Loss Tuning File Format

Atoll can tune path losses calculated by propagation models using CW measurements or drive test Data. Path losses are tuned
by merging measurement data with propagation results on pixels corresponding to the measurement points and the pixels in
the vicinity. Path losses surrounding the measurement points are smoothed for homogeneity.
Measuremment paths that are used for path loss tuning are stored as a catalogue in a folder containing a pathloss.dbf file and
one PTS (path loss tuning) file per transmitter. A tuning file can contain several measurement paths.
For more information on the path loss tuning algorithm, see the Technical Reference Guide.
2.16.2.1 Pathloss.dbf File Format

See "Pathloss.dbf File Format" on page 107.
2.16.2.2 Pathloss.dbf File Contents

The DBF file provides information about the measured transmitters involved in the tuning.

FILE_NAME Text Name (and optionally, path) of .pts file
AREA_XMIN Float Not used
AREA_XMAX Float Not used
AREA_YMIN Float Not used
AREA_YMAX Float Not used
2.16.2.3 PTS File Format

The PTS (path loss tuning) files contain a header and the list of measurement points.
Header:
• 4 bytes: version
• 4 bytes: flag (can be used to manage flags like active flag)
• 50 bytes: GUID
• 4 bytes: number of points
• 255 bytes: original measurement name (with prefix "Num" for drive test data and "CW" for CW measurements)
• 256 bytes: comments
• 4 bytes: X_RADIUS
• 4 bytes: Y_RADIUS
• 4 bytes: gain = measurement gain - losses
• 4 bytes: global error
• 4 bytes: rx height
• 4 bytes: frequency
• 8 bytes: tx Position
List of measurement points:
• 4 bytes: X
• 4 bytes: Y
• 4 bytes: measurement value
• 4 bytes: incidence angle.
111
2.16.3 Interference Matrix File Formats

Interference matrices are used by GSM, LTE, and WiMAX AFPs (automatic frequency planning tools). Interference matrices
can be imported and exported using the following formats:
• GSM: CLC, IM0, IM1, IM2
• LTE and WiMAX: IM2, TXT, CSV
Interference matrix files must contain interference probability values between 0 and 1, and not in precentage (between 0 and
100%). When interference matrix files are imported, Atoll does not check their validity and imports interference probability
values for loaded transmitters only.
In the following format descriptions and samples, lines starting with the "#" are considered as comments.
In GSM interference matrices:

• The interferer TRX type is not specified and is always considered to be BCCH.
Subcells have different powers defined as offsets with respect to the BCCH. For
subcells other than the BCCH, if the power offset of a subcell is X dB, then its
interference histogram will be shifted by X dB with respect to the BCCH interference
histogram.
• If no power offset is defined on the interfered TRX type, it is possible to set "All".
• For each interfered subcell-interferer subcell pair, Atoll saves probabilities for
several C/I values (6 to 24 values), including five fixed ones: –9, 1, 8, 14, and 22 dB.
Between two fixed C/I value, there can be up to three additional values (this number
depends on the probability variation between the fixed values). The C/I values have
0.5 dB accuracy and probability values are calculated and stored with an accuracy of
0.002 for probabilities between 1 and 0.05, and with an accuracy of 0.0001 for
probabilities lower than 0.05.
2.16.3.1 CLC Format (One Value per Line)

The CLC format uses wo ASCII text files: a CLC file and a DCT file. Interference matrices are imported by selecting the CLC file
to import. Atoll looks for the associated DCT file in the same directory and uses it to decode transmitter identifiers. If no DCT
file is available, Atoll assumes that the transmitter identifiers are the transmitter names, and the columns 1 and 2 of the CLC
file must contain the names of the interfered and interfering transmitters instead of their identification numbers.
2.16.3.2 CLC File Format

The CLC file consists of two parts. The first part is a header used for format identification. It must start with and contain the
following lines:
# Calculation Results Data File.

# Version 1.1, Tab separated format. Commented lines start with #.
The second part details interference histogram of each interfered subcell-interfering subcell pair. The lines after the header
are considered as comments if they start with "#". If not, they must have the following format:
<Column1><tab><Column2><tab><Column3><tab><Column4><tab><Column5><newline>
The 5 tab-separated columns are defined in the table below:

Column Name Description
Identification number of the interfered transmitter. If the column is empty,
Column1 Interfered transmitter
its value is identical to the one of the line above.
Identification number of the interferer transmitter. If the column is null, its
Column2 Interfering transmitter
value is identical to the one of the line above.
Interfered subcell. If the column is null, its value is identical to the one of the
Column3 Interfered TRX type line above. In order to save storage, all subcells with no power offset are not
duplicated (e.g. BCCH, TCH).
Column4 C/I threshold C/I value. This column cannot be null.
Probability to have C/I the value specified in column 4 (C/I threshold). This
Column5 Probability C/I > Threshold
field must not be empty.
The columns 1, 2, and 3 must be defined only in the first line of each histogram.
112
Sample

# Remark: C/I results do not incorporate power offset values.
# Fields are:
##------------#------------#------------#-----------#------------------#
#| Interfered | Interfering| Interfered | C/I | Probability |
#| Transmitter| Transmitter| Trx type | Threshold | C/I >= Threshold |
##------------#------------#------------#-----------#------------------#
#
# Warning, The parameter settings of this header can be wrong if
# the "export" is performed following an "import". They
# are correct when the "export" follows a "calculate".
#
# Service Zone Type is "Best signal level of the highest priority HCS layer".
# Margin is 5.
# Cell edge coverage probability 75%.
# Traffic spreading was Uniform
##---------------------------------------------------------------------#
1 2 TCH_INNER 8 1
9 0.944
10 0.904
11 0.892
14 0.844
15 0.832
16 0.812
17 0.752
22 0.316
25 0.292
1 2 BCCH,TCHa 8 1
9 0.944
10 .904
13 0.872
14 0.84
17 0.772
a. If the TCH and BCCH histograms are the same, they are not repeated. A single record indicates that the histograms
belong to TCH and BCCH both.
2.16.3.3 DCT File Format

The .dct file is divided into two parts. The first part is a header used for format identification. It must start with and contain
the following lines:
# Calculation Results Dictionary File.

The second part provides information about transmitters taken into account in AFP. The lines after the header are considered
as comments if they start with "#". If not, they must have the following format:
<Column1><tab><Column2><newline>
113
Column Name Type Description

Column1 Transmitter name Text Name of the transmitter
Column2 Transmitter Identifier Integer Identification number of the transmitter
Column3 BCCH during calculation Integer BCCH used in calculations
Column4 BSIC during calculation Integer BSIC used in calculations
Column5 % of vic’ coverage Float Percentage of overlap of the victim service area
Column6 % of int’ coverage Float Percentage of overlap of the interferer service area
The last four columns describe the interference matrix scope. One transmitter per line is described separated with a tab
character.
Sample
# Calculation Results Dictionary File.

# Fields are:
##-----------#-----------#-----------#-----------#---------#---------#
#|Transmitter|Transmitter|BCCH during|BSIC during|% of vic'|% of int'|
#|Name |Identifier |calculation|calculation|coverage |coverage |
##-----------#-----------#-----------#-----------#---------#---------#
#
#
# Service Zone Type is "Best signal level per HCS layer".
# Margin is 5.
# Cell edge coverage probability is 75%.
# Traffic spreading was Uniform (percentage of interfered area)
##---------------------------#
Site0_0 1 -1 -1 100 100
Site0_1 2 -1 -1 100 100
Site0_2 3 -1 -1 100 100
Site1_0 4 -1 -1 100 100
Site1_1 5 -1 -1 100 100
Site1_2 6 -1 -1 100 100
Site2_0 7 -1 -1 100 100
Site2_1 8 -1 -1 100 100
2.16.3.4 IM0 Format (One Histogram per Line)

This file contains one histogram per line for each interfered/interfering subcell pair. The histogram is a list of C/I values with
associated probabilities.
The .im0 file consists of two parts. The first part is a header used for format identification. It must start with and contain the
following lines:
The second part details interference histogram of each interfered subcell-interferer subcell pair. The lines after the header
<Column1><tab><Column2><tab><Column3><tab><Column4><newline>
114

Column1 Interfered transmitter Name of the interfered transmitter.
Column2 Interfering transmitter Name of the interferer transmitter.
Interfered subcell. In order to save storage, all subcells with no power offset
Column3 Interfered TRX type
are not duplicated (e.g. BCCH, TCH).
C/I value and the probability associated to this value separated by a space
Column4 C/I probability
character. This entry cannot be null.
Sample

# Fields are:
#------------------------------------------------------------------------
#Transmitter Interferer TRX type {C/I Probability} values
#------------------------------------------------------------------------
#
#
# Margin is 5.
##---------------------------------------------------------------------#
#
Site0_2 Site0_1 BCCH,TCH-10 1 -9 0.996 -6 0.976 -4 0.964 -1 0.936
0 0.932 1 0.924 4 0.896 7 0.864 8 0.848
9 0.832 10 0.824 11 0.804 14 0.712 17 0.66
Site0_2 Site0_3 BCCH,TCH-10 1 -9 0.996 -6 0.976 -4 0.972 -1 0.948
0 0.94 1 0.928 4 0.896 7 0.856 8 0.84
11 0.772 13 0.688 14 0.636 15 0.608 18 0.556
Site0_3 Site0_1 BCCH,TCH-10 1 -9 0.996 -6 0.98 -3 0.948 0 0.932
1 0.924 4 0.892 7 0.852 8 0.832 9 0.816
10 0.784 11 0.764 14 0.644 15 0.616 18 0.564
Site0_3 Site0_2 BCCH,TCH-9 1 -6 0.972 -3 0.964 -2 0.96 0 0.94
1 0.932 4 0.904 7 0.876 8 0.86 9 0.844
11 0.804 13 0.744 14 0.716 15 0.692 18 0.644
2.16.3.5 IM1 Format (One Value per Line, TX Name Repeated)

This file contains one C/I threshold and probability pair value per line for each interfered/interfering subcell pair. The
histogram is a list of C/I values with associated probabilities.
The .im1 file consists of two parts. The first part is a header used for format identification. It must start with and contain the
following lines:

The second part details interference histogram of each interfered subcell-interferer subcell pair. The lines after the header
115
<Column1><tab><Column2><tab><Column3><tab><Column4><tab><Column5><newline>

Column1 Interfered transmitter Name of the interfered transmitter.
Column2 Interfering transmitter Name of the interferer transmitter.
Interfered subcell. In order to save storage, all subcells with no power offset
Column3 Interfered TRX type
are not duplicated (e.g. BCCH, TCH).
Column4 C/I threshold C/I value. This column cannot be null.
Probability to have C/I the value specified in column 4 (C/I threshold). This
Column5 Probability C/I > Threshold
field must not be empty.
Sample

# Fields are:
#------------------------------------------------------------------------
#Transmitter Interferer TRX type C/I Probability
#------------------------------------------------------------------------
#
#
# Margin is 5.
##---------------------------------------------------------------------#
Site0_2 Site0_1 BCCH,TCH -10 1
Site0_2 Site0_1 BCCH,TCH -9 0.996
Site0_2 Site0_1 BCCH,TCH 0 0.932
...
2.16.3.6 IM2 Format (Co- and Adjacent-channel Probabilities)

IM2 files contain co-channel and adjacent-channel interference probabilities for each interfered transmitter – interfering
transmitter pair. In GSM, there is only one set of values for all the subcells of the interfered transmitter. Each line must have
the following format:
<Column1><SEP><Column2><SEP><Column3><SEP><Column4><newline>
116
Where the separator (<SEP>) can either be a tab or a semicolon.

The four columns are defined in the table below:
Column1 Interfered transmitter Name of the interfered transmitter
Column2 Interfering transmitter Name of the interferer transmitter
Column3 Co-channel probability Co-channel interference probability
Column4 Adjacent-channel probability Adjacent channel interference probability
Sample

# Fields are:
#------------------------------------------------------------------------
#Transmitter Interferer Co-channel Adjacent channel
#------------------------------------------------------------------------
#
#
# Margin is 5.
##---------------------------------------------------------------------#
Site0_2 Site0_1 0.226667 0.024
Site0_2 Site0_3 0.27 0.024
Site0_3 Site0_1 0.276 0.02
Site0_3 Site0_2 0.226 0.028
The columns in the sample above are separated with a tab. These columns can also be separated with a semilcolon:
Site0_2;Site0_1;0.226667;0.024
Site0_2;Site0_3;0.27;0.024
Site0_3;Site0_1;0.276;0.02
Site0_3;Site0_2;0.226;0.028
2.16.4 "Per Transmitter" Prediction File Format

When a coverage prediction is calculated by value intervals it is stored externally, i.e., outside the ATL file. A corresponding
’<doc_name>\{<GUID>}’ folder is actually created where the ATL document is located, as soon as the latter is saved.
The calculation of the coverage prediction is either global or "per transmitter".
• When the calculation is global, the results are stored in two files for the entire prediction: one HDR file and one BIL
file (both identified by the prediction name).
• When a calculation is "per transmitter", one HDR file and one BIL file are created for each transmitter in the prediction
(both identified by the transmitter’s name). In some "per transmitter" predictions, an additional DBF file is created for
the entire prediction (identified by the prediction name). The DBF file contains information on each transmitter and a
pointer to each transmitter’s specific HDR and BIL files.
In LTE, when a Cell Identifier Collision Zones (DL) prediction is calculated by value intervals with the display type set
"No. of interferers per cell", the HDR file and the BIL file are created for each cell in the prediction (both identified by
the cell’s name).
117
The format and the content of the DBF file is described here.
In both cases, an XML file describing the prediction is also created in the corresponding
’<doc_name>\{<GUID>}’ folder.
2.16.4.1 <per_transmitter_prediction>.dbf File Format

The format of ‘<per_transmitter_prediction>.dbf’ files is identical to the format described in "Pathloss.dbf File Format" on
page 107.
2.16.4.2 <per_transmitter_prediction>.dbf File Contents

The ‘<per_transmitter_prediction>.dbf’ files generated in specific ’{<GUID>}’ folders provide information that is needed to
check the validity of each "per transmitter" prediction> calculated by value intervals.

FILE_NAME Text Name of the transmitter’s BIL result file
RESOLUTION Float Resolution of the calculation, same as ’xdim’ and ’ydim’ in the HDR file
AREA_XMIN Float Same as ’ulxmap’ in the HDR file
AREA_XMAX Float Same as ’ulxmap’ + ’xdim’ * ’ncols’ in the HDR file
AREA_YMIN Float Same as ’ulymap’ in the HDR file
AREA_YMAX Float Same as ’ulymap’ + ’ydim’ * ’nrows’ in the HDR file
NBITS Float Same as ’nbits’ in the HDR file
NBANDS Float Same as ’nbands’ in the HDR file
BYTE_ORDER Float Same as ’byteorder’ in the HDR file
BAND_ROW_BYTES Float Same as ’bandrowbytes’ in the HDR file
TOTAL_ROW_BYTES Float Same as ’totalrowbytes’ in the HDR file
SKIP_BYTES Float Same as ’skipbytes’ in the HDR file
DATA_TYPE Text Same as ’datatype’ in the HDR file
NO_DATA_VALUE Same as ’nodatavalue’ in the HDR file
2.16.5 Coverage Prediction Export and Reports

2.16.5.1 Filtering Coverage Predictions at Export
Raster and vector coverage predictions can be filtered at export in order to exclude holes and islands. Predictions are filtered
by setting the colour of a pixel to the dominant colour of the bounding box, i.e., surrounding pixels, using a dispersion factor:
2
exp  – D   X  2   .
Here, D is the distance from the pixel to be coloured to each pixel within the bounding box and X is the value at that pixel.
In other words, the pixel will be coloured by the most representative value within this bounding box.
118
Figure 2.27: Bounding box for prediction filtering
The user-defined filtering percentage Y gives the size of the bounding box: Y  10 pixels in each direction. In other words, the
bounding box is increased by one pixel every 10 % (since Y is defined as a percentage).
2.16.5.2 Smoothing Coverage Predictions at Export

Vector coverage predictions can be smoothed at export in orer to simplify its contours. Predictions are smoothed by reducing
the number of points defining the contours of the polygons using a vertex reduction routine that successively reduces the
number of closely clustered vertices (vertex reduction within tolerance of prior vertex cluster, Douglas-Peucher polyline
simplification).
Two smoothing methods exist for defining the degree of coverage smoothing: smoothing by percentage and smoothing by
the maximum number of points.
Smoothing by Percentage
2 Z
The user-defined smoothing percentage Z gives the approximation tolerance: -------  R  ------ , where R is the user-defined
2 20
export resolution. Tolerance is the interval within which Atoll tries to reduce the number of points.
Figure 2.28: Smoothing Tolerance Definition
For example, for three successive points, A1, A2, and A3 as shown in Figure 2.29 on page 119, A2 will be deleted if within this
tolerance (and A1 and A3 will be directly linked) and A2 will be conserved if outside this tolerance.
A2 outside the tolerance interval A2 inside the tolerance interval

Figure 2.29: Smoothing by Percentage
119
Smoothing by Number of Points
The second method consists in defining a maximum number of points to be deleted. This number of points helps the algorithm
to determine the optimised tolerance (see "Smoothing by Percentage" on page 119) such that, with this obtained tolerance,
the number of points to be deleted will be lower than this value.
Let’s consider the following example ( 1 ). Starting from the maximum possible tolerance, the number of points to be filtered
out are estimated (circled in red in the following example ( 2 )). If this number is greater than the maximum number of points
defined by the user, Atoll reduces the tolerance until reaching the requested maximum number of points or less ( 3 ). The
first the number of points respecting the constraint is obtained, smoothing is applied by deleting these points and linking the
remaining closest points ( 4 ).
1 2
3 4
Figure 2.30: Smoothing by Number of Points
2.16.5.3 Examples of Prediction Export Filtering and Smoothing

Figure 2.31 on page 121 shows the original signal level coverage prediction whose filtered and smoothed exported results are
presented in Figure 2.32 on page 121.
120
Figure 2.31: Bounding box for prediction filtering
Filtering Percentage: 0 % Filtering Percentage: 0 %

Smoothing Percentage: 0 % Smoothing Percentage: 100 %
Filtering Percentage: 100 % Filtering Percentage: 100 %

Smoothing Percentage: 0 % Smoothing Percentage: 100 %
Figure 2.32: Exported prediction with filtering and smoothing
2.16.5.4 Coverage Prediction Reports Over Focus/Computation Zones

Statistics are calculated in coverage prediction reports over the focus zone or the computation zone, if no focus zone exists,
or the covered area, if neither zone exists.
If the reference surface area for the statistics is based on a focus or computation zone, there may be minute inaccuracies in
the calculated statistics because of the difference in the surface area calculation methods:
• The surface areas of the zones (polygons) are calculated by triangulation.
• The surface area of a coverage predictions is calculated by counting the number of covered pixels and multiplying this
number with the area of one pixel, calculated from resolution of the coverage prediction.
At the border of the focus or computation zone, a pixel is considered inside the zone if its centre is inside. Otherwise, the pixel
is considered outside the zone. This estimation may give rise to inaccuracies.
121
122
Chapter 3
GSM GPRS EDGE
Networks
• "Signal Level Calculation" on page 125
• "Interference-based DL Calculations" on page 131
• "GPRS/EDGE Calculations" on page 136
• "Codec Mode Selection and CQI Calculations" on
page 145
• "UL Coverage Predictions" on page 152
• "Traffic Analysis" on page 159
• "Network Dimensioning" on page 171
• "Key Performance Indicators Calculation" on page 181
• "Simulations" on page 184
• "Automatic Neighbour Allocation" on page 191
• "AFP Appendices" on page 196
© Forsk 2015
124
AT330_TRR_E1 Chapter 3: GSM GPRS EDGE Networks
3 GSM GPRS EDGE Networks

This chapter describes all the calculations performed in Atoll GSM/GPRS/EDGE documents. The first four sections describe the
signal level, interference, GPRS/EDGE-specific, and CQI calculations, respectively. The following three sections explain the
traffic analysis, network dimensioning, and KPI calculation processes. The last section describes the neighbour allocation
process in GSM.
• All the calculations are performed on TBC (to be calculated) transmitters. For the
definition of TBC transmitters please refer to "Path Loss Matrices" on page 103.
• Logarithms used in this chapter (Log function) are base-10 unless stated otherwise.
3.1 Signal Level Calculation

3.1.1 DL Signal Level
Three parameters can be studied in point analysis (Profile tab) and in signal level-based coverage predictions:
Studied Parameter Formulas

Signal level received from a transmitter on a TRX type
Txi
DL Signal level ( P rec ) Txi Txi
P rec  tt  = EIRP  tt  – P  tt  – L path – M Shadowing – model – L Indoor +  G ant – L Term 
Term
Txi Txi
Path loss ( L path ) L path = L model + L ant
Tx
Txi Txi Txi

DL Total losses ( L total ) L total =  L path + M Shadowing – model + L Indoor + L Tx – DL + L Term  –  G ant + G ant 
Tx Term
Here,
• EIRP is the effective isotropic radiated power of the transmitter,
• L model is the loss on the transmitter-terminal path (path loss) calculated by the propagation model,
• L ant is the transmitter antenna attenuation (from antenna patterns),
Tx
• M Shadowing – model is the shadowing margin. This parameter is taken into account when the option “Shadowing taken
into account” is selected,
• L Indoor are the indoor losses, taken into account when the option “Indoor coverage” is selected,
• L Term are the terminal losses,
• L Tx – DL are the total transmitter DL losses,
• G ant is the terminal antenna gain,
Term
• P is the power offset defined for the selected TRX type in the transmitter property dialogue,
• tt is the TRX type (in the GSM GPRS EDGE.mdb document template, there are three possible TRX types, BCCH, TCH and
inner TCH).
3.1.2 UL Signal Level

Two parameters can be studied in UL signal level-based coverage predictions:
Studied Parameter Formulas

Signal level received from a terminal at a transmitter
Term
UL Signal level ( P rec ) Term max Txi
P rec = P Term + G ant – L path – M Shadowing – model – L Indoor +  G ant – L Tx – UL – L Term 
Tx Term
Term Term Txi

UL Total losses ( L total ) L total =  L path + M Shadowing – model + L Indoor + L Tx – UL + L Term  –  G ant + G ant 
Tx Term
Here,
max
• P Term is the maximum terminal allowed power,
125
Chapter 3: GSM GPRS EDGE Networks ©Forsk 2015
Txi
• L path is the path loss on the transmitter-terminal link, identical in DL and in UL (see above),
• M Shadowing – model is the shadowing margin. This parameter is taken into account when the option “Shadowing taken
into account” is selected,
• L Indoor are the indoor losses, taken into account when the option “Indoor coverage” is selected,
• L Term are the terminal losses,
• L Tx – UL are the total transmitter UL losses,
• G ant is the transmitter antenna gain,
Tx
• G ant is the terminal antenna gain.

Term
3.1.3 Point Analysis

3.1.3.1 Profile Tab
Txi Txi
For a selected transmitter, it is possible to display the signal level received from a TRX type ( P rec  tt  ), the path loss, L path ,
Txi
or the DL total losses, L total – DL . Path loss and DL total losses are the same for all TRX types.
If the power reduction values defined for all the subcells are the same, the received signal level from the selected transmitter
will be the same for all TRX types.
3.1.3.2 Reception Tab

Analysis provided in the Reception tab is based on path loss matrices. Therefore, it is possible to display the signal levels
received from TBC transmitters for which path loss matrices have been calculated over their calculation areas.
Txi Txi
For each transmitter, Atoll can display the signal level received from a TRX type ( P rec  tt  ), the path loss, L path , or the total
Txi
losses, L total – DL . Path loss and total losses are the same for all TRX types.
If the power reduction values defined for all the subcells are the same, the received signal level from the selected transmitter
will be the same for all TRX types.
Reception level bars are displayed in the order of decreasing signal level. The number of displayed bars depends on the signal
level received from the best server. Bars are only displayed for transmitters whose signal level is within a 30 dB margin from
the best server signal level.
You can use a value other than 30 dB for the margin from the best server signal level, for
example a smaller value for improving the calculation speed. For more information on
defining a different value for this margin, see the Administrator Manual.
3.1.4 Signal Level-based DL Coverage Predictions

For each TBC transmitter, Txi, Atoll calculates the selected parameter on each pixel inside the Txi calculation area. In other
words, each pixel inside the Txi calculation area is considered a probe (non-interfering) receiver.
Coverage prediction parameters to be set are:
• The coverage conditions in order to determine the service area of each TBC transmitter, and
• The display settings to select the displayed parameter and its shading levels.
3.1.4.1 DL Service Area Determination

Atoll uses parameters entered in the Condition tab of the coverage prediction properties dialogue to determine the areas
coverage will be displayed.
We can distinguish eight cases as below. Let us assume that:
• Each transmitter, Txi, belongs to a Hierarchical Cell Structure (HCS) layer, k, with a defined priority and a defined
reception threshold.
• No max range is set.
3.1.4.1.1 All Servers

The service area of Txi corresponds to the pixels :
126
Txi
MinimumThreshold  P rec  tt   MaximumThreshold
Txi
For pure signal level-based calculations (not C/I or C/(I+N)), P rec  tt  can be replaced
Txi Txi
with L total – DL or L path .
3.1.4.1.2 Best Signal Level and a Margin

Txi
Txi
Txi Txi
And P rec  tt   Best  P rec  tt   – M

Txi Txj
ji
M is the specified margin (dB). The Best function considers the highest value from a list of values.
• If M = 0 dB, Atoll considers pixels the received signal level from Txi is the highest.
• If M = 2 dB, Atoll considers pixels the received signal level from Txi is either the highest or within a 2 dB margin from
the highest.
• If M = -2 dB, Atoll considers pixels the received signal level from Txi is 2 dB higher than the signal levels from
transmitters which are 2nd best servers.
3.1.4.1.3 Second Best Signal Level and a Margin

The service area of Txi corresponds to the pixels:
Txi
Txi
Txi Txi
Txi nd Txj
And P rec  tt   2 Best  P rec  tt   – M
ji
M is the specified margin (dB). The 2nd Best function considers the second highest value from a list of values.
• If M = 0 dB, Atoll considers pixels the received signal level from Txi is the second highest.
• If M = 2 dB, Atoll considers pixels the received signal level from Txi is either the second highest or within a 2 dB margin
from the second highest.
transmitters which are 3rd best servers.
3.1.4.1.4 Best Signal Level per HCS Layer and a Margin

For each HCS layer, k, the service area of Txi corresponds to the pixels :
Txi
Txi
Txi Txi
And P rec  BCCH   Best  P rec  BCCH   – M

Txi Txj
ji
127
the highest.
3.1.4.1.5 Second Best Signal Level per HCS Layer and a Margin
Txi
Txi
Txi Txi
Txi nd Txj
And P rec  BCCH   2 Best  P rec  BCCH   – M
ji
3.1.4.1.6 HCS Servers and a Margin

Txi
Txi
Txi Txi

Txi Txj
ji
Txi
And the received P rec  tt  exceeds the reception threshold defined per HCS layer.
the highest.
3.1.4.1.7 Highest Priority HCS Server and a Margin

Txi
Txi
Txi Txi
128

Txi Txj
ji
And Txi belongs to the HCS layer with the highest priority. The highest priority is defined by the priority field (0: lowest).
Txi
In the case two layers have the same priority, the traffic is served by the transmitter for
which the difference between the received signal strength and the HCS threshold is the
highest. The way the competition is managed between layers with the same priority can
be modified. For more information, see the Administrator Manual.
the highest.
3.1.4.1.8 Best Idle Mode Reselection Criterion (C2)

Such type of coverage is useful :
• To compare idle and dedicated mode best servers for voice traffic
• Display the GPRS/EDGE best server map (based on GSM idle mode)
The path loss criterion C1 used for cell selection and reselection is defined by:
Txi
C1 = P rec  BCCH  – MinimumThreshold  BCCH 
The path loss criterion (GSM03.22) is satisfied if C1  0 .

The reselection criterion C2 is used for cell reselection only and is defined by:
C2 = C1 + CELL_RESELECT_OFFSET
CELL_RESELECT_OFFSET is the Cell Reselect Offset defined for the transmitter.

Txi
MinimumThreshold  P rec  BCCH   MaximumThreshold
Txi
Txi Txi
 BCCH  = Best  C2  BCCH  

Txi Txj
And C2
j
The Best function considers the highest value from a list of values.
On each pixel, the transmitter with the highest C2 value is kept. It corresponds to the best server in idle mode. C2 is defined
as an integer in the 3GPP specifications, therefore, the C2 values in the above calculations are rounded down to the nearest
integer.
3.1.4.2 Coverage Display

3.1.4.2.1 Coverage Resolution
The resolution of the coverage prediction does not depend on the resolutions of the path loss matrices or the geographic data
and can be defined separately for each coverage prediction. Coverage predictions are generated using a bilinear interpolation
method from multi-resolution path loss matrices (similar to the one used to calculate site altitudes, see "Path Loss Calculation
Prerequisites" on page 57 for more information).
3.1.4.2.2 Display Types

It is possible to display the coverage predictions with colours depending on any transmitter attribute or other criteria such as:
129
Signal Level (in dBm, dBµV, dBµV/m)
Atoll calculates signal level received from the transmitter on each pixel of each transmitter service area. A pixel of a service
area is coloured if the signal level is greater than or equal to the defined minimum thresholds (pixel colour depends on signal
level). Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as many
layers as transmitter service areas. Each layer shows the different signal levels available in the transmitter service area.
Best Signal Level (in dBm, dBµV, dBµV/m)
Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. When other service
areas overlap the studied one, Atoll chooses the highest value. A pixel of a service area is coloured if the signal level is greater
than or equal to the defined thresholds (the pixel colour depends on the signal level). Coverage consists of several
independent layers whose visibility in the workspace can be managed. There are as many layers as defined thresholds. Each
layer corresponds to an area the signal level from the best server exceeds a defined minimum threshold.
Path Loss (dB)
Atoll calculates path loss from the transmitter on each pixel of each transmitter service area. A pixel of a service area is
coloured if path loss is greater than or equal to the defined minimum thresholds (pixel colour depends on path loss). Coverage
consists of several independent layers whose visibility in the workspace can be managed. There are as many layers as service
areas. Each layer shows the different path loss levels in the transmitter service area.
Total Losses (dB)
Atoll calculates DL total losses from the transmitter on each pixel of each transmitter service area. A pixel of a service area is
coloured if total losses is greater than or equal to the defined minimum thresholds (pixel colour depends on total losses).
Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as many layers
as service areas. Each layer shows the different total losses levels in the transmitter service area.
Best Server Path Loss (dB)
Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. When other service
areas overlap the studied one, Atoll determines the best transmitter and evaluates path loss from the best transmitter. A pixel
of a service area is coloured if the path loss is greater than or equal to the defined thresholds (pixel colour depends on path
loss). Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as many
layers as defined thresholds. Each layer corresponds to an area the path loss from the best server exceeds a defined minimum
threshold.
Best Server Total Losses (dB)
Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. service areas overlap
the studied one, Atoll determines the best transmitter and evaluates DL total losses from the best transmitter. A pixel of a
service area is coloured if the total losses is greater than or equal to the defined thresholds (pixel colour depends on total
losses). Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as many
layers as defined thresholds. Each layer corresponds to an area the total losses from the best server exceed a defined
minimum threshold.
Number of Servers
Atoll evaluates how many service areas cover a pixel in order to determine the number of servers. The pixel colour depends
on the number of servers. Coverage consists of several independent layers whose visibility in the workspace can be managed.
There are as many layers as defined thresholds. Each layer corresponds to an area the number of servers is greater than or
equal to a defined minimum threshold.
Cell Edge Coverage Probability (%)
On each pixel of each transmitter service area, the coverage corresponds to the pixels the signal level from this transmitter
fulfils signal conditions defined in Conditions tab with different cell edge coverage probabilities. There is one coverage area
per transmitter in the explorer.
Best Cell Edge Coverage Probability (%)
On each pixel of each transmitter service area, the coverage corresponds to the pixels the best signal level received fulfils
signal conditions defined in Conditions tab. There is one coverage area per cell edge coverage probability in the explorer.
Best C2 (dBm)
Atoll calculates C2 values received from transmitters on each pixel of each transmitter service area. When other service areas
overlap the studied one, Atoll chooses the highest value. A pixel of a service area is coloured if the C2 value is greater than or
130
equal to the defined thresholds (the pixel colour depends on the C2 value). Coverage consists of several independent layers
whose visibility in the workspace can be managed. There are as many layers as defined thresholds. Each layer corresponds to
an area the best C2 value exceeds a defined minimum threshold.
3.2 Interference-based DL Calculations

Interference-based calculations include all the calculations that involve the calculation of interference received from
interfering transmitters in addition to the signal level received from the server.
3.2.1 DL Carrier-to-Interference Ratio Calculation

MSA (Mobile Station Allocation) Definition
In order to understand the difference between each frequency hopping mode from the point of view of a mobile, it is inter-
esting to consider the Mobile Station Allocation. MSA is characterised by the pair (Channel list, MAIO). In the following, we
will use this notion to characterise the interference and resources set of a mobile.
For non-hopping (NH) mode, the channel list is 1 channel. For base-band hopping (BBH) or synthesized frequency hopping
(SFH), the channel list corresponds to the mobile allocation list (MAL).
For BBH, channels of MAL belong to the same TRX type.
Examples:
• Non-hopping (NH): An MSA is the channel assigned to a TRX used by a mobile.
TRX index Channel list MAIO MSA

1 53 - (53,-)
2 54 - (54,-)
• Baseband hopping (BBH): An MSA is the Mobile Allocation List (MAL) and the TRX index.

1 53 * ([53,54,55],0)
2 54 * ([53,54,55],1)
3 55 * ([53,54,55],2)
• Synthesised frequency hopping (SFH): An MSA is the Mobile Allocation List (MAL) and the Mobile Allocation Index
Offset (MAIO).

1 53 54 55 56 2 ([53,54,55,56],2)
2 53 54 55 56 3 ([53,54,55,56],3)
Therefore, from the point of view of a mobile station, BBH and SFH work in the same way. An MSA will be attached to each
mobile considered during the simulation and the level of interference will be evaluated on this MSA.
Notations and Assumptions
In the following description:

• v is a victim transmitter,
• MSAS(v) is the set of MSAs (Mobile Station Allocations) associated to v,
The number of MSAS(v) depends on TRX types to be analysed. You may study a given TRX type tt (there will be as many
MSA(v) as TRXs allocated to the subcell (v,tt)) or all the TRX types (the number of MSA(v) will correspond to the
number of TRXs allocated to v).
 C v  m 
Several MSAs, m, are related to a transmitter. Therefore, Atoll calculates the DL C/I  --------------
- for each victim
 Iv  m  
transmitter v with MSA m (m  MSAS(v)).
131
 C v  m 
Atoll considers the most interfered MSA, therefore, the displayed C/I or C/(I+N) are  --- = Min  --------------
C - or
 Iv k  Iv  m  
 C -  C  m  -
v
 ---------------------  = Min  ---------------------------------  , respectively. If the Detailed Results check box is selected, the C/I values for
 I + N Term tot 
k  I v  m  + N Term
tot
v
all MSAs are displayed.
• i is any potential interfering transmitter (TBC transmitters whose calculation areas intersect the service area of v),
• MSAS(i) is the set of MSAs related to potential interferers i,
• INT(v) is the set of transmitters that interfere v,
v
• C  m  is the carrier power level received from v on m,
v
• I  m  corresponds to the interference received from interfering transmitters i on m,
• M Shadowing used in the C/I calculation is based on the C/I standard deviation.
The C/I shadowing margin is applied on the carrier power level. The interference levels
are not changed.
Calculations
The carrier power level is the power received from the victim transmitter at the terminal.
v v
C  m  = P rec  m 
Term
If the interference conditions are based on C/(I+N), Atoll takes the total noise N tot into account. The total noise is the sum
Term
of the thermal noise N thermal (-121 dBm by default or user-defined), the terminal noise figure NF , and the inter-
v DL
technology downlink noise rise NR inter – techno log y .
Term Term v DL
N tot = N thermal + NF + NR inter – techno log y
Interference can be received from interfering transmitters i on co-channel and adjacent channels. Interference may also be
received from the transmitters of another technology.
v v v DL i v
Therefore, I m = I co m + I adj m + I inter – techno log y – G PC – G Div
i v
G PC is the average power control gain defined for the interfering transmitter i and G Div is the diversity gain defined for the
considered subcell.
Each interference component is explained below.
Co- and Adjacent Channel Interference:
v
I co  m  is the interference received at v on m on co-channel, given by:
 v i 
 
v i
I co  m  =  p m n  P rec  n   T i  n 
 
i  INT  v  n  MSAS  i  co
v
I adj  m  is the interference received at v on m on adjacent channels, given by:
i
 v i P rec  n  
 
v
I adj  m  =  p m n  ----------------
-  T i  n 
 F  adj
i  INT  v  n  MSAS  i 
i
Here, P rec  n  is the carrier power level received from i on n.
Ti(n) is occupancy of the MSA n:
i i
T i  n  = L traffic  n   f act  n 
i
L traffic  n  is the traffic load defined for the MSA n or i. It can be set to 100% in the coverage prediction properties.
132
i
f act  n  is the activity factor defined for the MSA n of i. If the subcell (i,tt) supports DTX, the value specified in the
coverage prediction properties is used. Otherwise, the activity factor is 1.
BCCH TRXs are always on. Therefore, DTX and traffic loads do not impact the interference
i i
from BCCH. In other words, f act  n  = 1 and L traffic  n  = 1 for the BCCH TRXs of the
interferers.
v i
p m n is the probability of having a co- or adjacent channel collision between MSAs n and m, depending on the used
frequency hopping mode.
• Collision Probability for Non Hopping Mode:
v i
p m n = 1
• Collision Probability for BBH and SFH Modes:

MSA m of v can be defined as the pair ([f1,f2,….fn], MAIO) and MSA n of i as the pair ([f’1,f’2,….f’n], MAIO’) ( f and
f’ are channels).
v i
An occurence OCCUR  f m f' n  refers to the event when a channel f of m encounters a channel f’ of n during
hopping. A collision occurs when f and f’ are co- or adjacent channels:
v i v i
Collision = OCCUR  f m f' n  such that f m – f' n = 0 or 1
The probability of collision is the ratio of the number of collisions to the number of occurences:
v i n collision
p m n = ----------------------
-
n occurence
The probibility of collision depends on the correlation between m and n. There can be two cases:
i. MSAs m and n are correlated
m and n must have identical HSN and synchronisation. The number of occurrences depends on the MAL
length, MAIO, and MAIO’.
Example:
Schematic view of hopping sequences

MSA m of v
34 37 39
([34 37 39], MAIO=0)
MSA n of i
38 36 34
([38 36 34], MAIO’=2)
Here, the number of occurrences is 3, the number of co-channel collisions is 1, and the number of adjacent
channel collisions is 1. Therefore,
v i v i
 p m n  co = 1
--- and  p m n  adj = 1
---
3 3
ii. MSAs m and n are not correlated
m and n do not have identical HSN and synchronisation. The probability of collision is the same for all the
channels.
Example:
Schematic view of hopping sequences

MSA m of v
34 37 39
([34 37 39], MAIO=0)
MSA n of i
38 36 34
([38 36 34], MAIO’=2)
Here, the number of occurrences is 9, the number of co-channel collisions is 1, and the number of adjacent
channel collisions is 3. Therefore,
133
v i v i
 p m n  co = 1
--- and  p m n  adj = 1
---
9 3
Diversity gain:
v
G Div is the diversity gain defined for the victim subcell.
Two types of diversity modes can be defined. In Tx Diversity, the signal is transmitted as many times that there are
antennas. In, the signal is successively transmitted on the various antennas.
For Tx Diversity mode, the diversity gain is defined as:
v Tx_Div
G Div = 3dB + Gclutter
Tx_Div
G clutter is the additional transmit diversity gain defined for the clutter class on which is located m.
For Antenna Hopping mode, the diversity gain is defined as:

v Ant_Div
G Div = G clutter
Ant_Div
G clutter is the antenna hopping gain defined for the clutter class on which is located m.
Inter-technology Downlink Interference:

DL
I inter – techno log y is the total inter-technology interference level on m due to transmitters in a linked Atoll document.
The interference from a transmitter Tx in a linked Atoll document is given as:

Tx
P Transmitted  ic i 
 -------------------------------------
DL
I inter – techno log y = -
Tx Tx
ni
L total  ICP ic i f
th Tx
ic i is the i frequency used by the transmitter Tx within its list of frequencies, P Transmitted  ic i  is the total
Tx Tx
transmitted Tx power on ic i , L total are the total losses between the transmitter Tx and the terminal, and ICP ic  f is
i
the inter-technology channel protection between the frequencies used by the transmitter Tx and the victim
transmitter v.
• In case of frequency hopping, the ICP value is weighted according to the fractional
load.
• In the ICP, the frequency gap is based on the defined base frequency for each
technology (e.g., 935 MHz in GSM 900)

Analysis provided in the Interference tab is based on path loss matrices. Therefore, it is possible to display the interference
levels received from TBC transmitters for which path loss matrices have been calculated over their calculation areas.
Atoll displays the following at the terminal:
• The carrier power level received from the victim transmitter v on the most interfered MAS m,
• Co-channel, adjacent channel, or both co- and adjacent channel interference received from interfering transmitters i
on MAS m (for further information about noise calculation, please refer to Signal to noise calculation: noise calculation
part),
Interferers are sorted in the order of descending carrier power levels.
• Neither DTX nor traffic load of TRXs are taken into account to evaluate interference
i i
levels. Therefore, we have T i  n  = L traffic  n   f act  n  = 1 .
• The C/I shadowing margin is applied on the carrier power level. The interference
levels are not changed.
134
3.2.3 Interference-based DL Coverage Predictions

Two interference-based DL coverage predictions are available:
• Coverage by C/I Level (DL): Provides a global analysis of the network quality.
Atoll calculates the C/I on each pixel within the service area of studied transmitters, determines the pixels where the
calculated C/I exceeds the defined minimum threshold, and colours these pixels depending on C/I value.
• Interfered Zones: Shows the areas a transmitter is interfered.
Atoll calculates the C/I on each pixel within the service area of studied transmitters, determines the pixels where the
calculated C/I is lower than the defined maximum threshold, and colours these pixels depending on colour of the
interfered transmitter.
• The coverage conditions in order to determine the service area of each TBC transmitter,
• The interference conditions to meet for a pixel to be covered, and
The thermal noise (N = -121 dBm, by default) is used in the calculations if the coverage prediction is based on C/(I+N). This
value can be modified by the user.
3.2.3.1 Service Area Determination

coverage will be displayed. Service areas are determined in the same manner as for signal level-based coverage predictions.
See "DL Service Area Determination" on page 126 for more information.
3.2.3.2 Coverage Area Determination

For each victim transmitter v, coverage area corresponds to pixels where DL  --- or  ----------- is between the lower and upper
C C
I v I+N v
thresholds defined in the coverage prediction properties.
The two options defining the thresholds are explained below.
3.2.3.2.1 Interference Condition Satisfied by At Least One TRX

In this case, the coverage area of a transmitter Txi corresponds to the pixels :
Minimum threshold   ---  Maximum threshold or Minimum threshold   -----------

C C
 Maximum threshold
I v I+N v
TRX TRX
j j
, TRXj is any TRX belonging to Txi.
3.2.3.2.2 Interference Condition Satisfied by The Worst TRX

In this case, the coverage area of a transmitter Txi corresponds to the pixels :
Minimum threshold   ---  Maximum threshold or Minimum threshold   -----------

C C
 Maximum threshold
 Iv  I + N v
TRX j TRX j
, TRXj is the TRX (belonging to Txi) with the worst C/I or C/(I+N) at the pixel.


135
C/I Level
Each pixel of the transmitter coverage area is coloured if the calculated DL C/I (or C/(I+N)) level is greater than or equal to the
specified minimum thresholds (pixel colour depends on DL C/I (or C/(I+N)) level). Coverage consists of several independent
layers whose visibility in the workspace can be managed. There are as many layers as transmitter coverage areas. Each layer
shows the different DL C/I levels available in the transmitter coverage area.
Max C/I Level
Atoll compares calculated DL C/I (or C/(I+N)) levels received from transmitters on each pixel of each transmitter coverage area
coverage areas overlap the studied one and chooses the highest value. A pixel of a coverage area is coloured if the DL C/I (or
C/(I+N)) level is greater than or equal to the specified thresholds (the pixel colour depends on the DL C/I (or C/(I+N)) level).
as defined thresholds. Each layer corresponds to an area the highest received DL C/I level exceeds a defined minimum
threshold.
Min C/I Level
Atoll compares DL C/I (or C/(I+N)) levels received from transmitters on each pixel of each transmitter coverage area the
coverage areas overlap the studied one and chooses the lowest value. A pixel of a coverage area is coloured if the DL C/I (or
C/(I+N)) level is greater than or equal to the specified thresholds (the pixel colour depends on the DL C/I (or C/(I+N)) level).
as defined thresholds. Each layer corresponds to an area the lowest received DL C/I level exceeds a defined minimum
threshold.
3.3 GPRS/EDGE Calculations

GPRS/EDGE calculations include coding scheme selection and throughput calculation. Coding schemes may be selected using
ideal link adaptation or without it. Once coding schemes have been selected, throughputs corresponding to these coding
schemes are readily determined from the look-up tables.
The following sections describe the two categories of calculations, i.e., with and without ideal link adaptations. Ideal link
adaptation implies that the selected coding scheme corresponds to the highest available throughput under the given radio
conditions.
GPRS/EDGE calculations may be based on signal levels (C) alone, on C/I, or on C/(I+N). For calculating the noise, either the
noise figure defined for the calculations or that of the selected terminal type is used.
Different GPRS/EDGE configurations may be defined for transmitter and terminals. In this case, Atoll only selects the coding
schemes that are common in the two, and gives priority to the thresholds defined in the terminal configuration. If no terminal
type is defined for the calculation, or if the terminal type does not have any GPRS/EDGE configuration assigned to it, Atoll only
uses the GPRS/EDGE configuration of the transmitter. If the transmitter does not have any GPRS/EDGE configuration assigned
to it, no coding scheme selection and throughput calculation is carried out.
In the following calculations, we assume that:
Txi
• P rec  TRX  is the signal level received from the selected TRX type (tt) or on all the TRXs of Txi on each pixel of the Txi
coverage area,
Txi
• P Backoff  TRX  is the Power Backoff defined for the subcell for 8PSK, 16QAM, or 32QAM modulations,
• CS is the set of all available coding schemes,
•  Reception Threshold  CS are the values of reception thresholds for the coding schemes available in the GPRS/EDGE
configuration,
• C
--- 
 I Threshold CS are the values of C/I thresholds for the coding schemes available in the GPRS/EDGE configuration,
•  -----------
C 
 I + N Threshold CS are the values of C/(I+N) thresholds for the coding schemes available in the GPRS/EDGE
configuration,
• The priorities of the coding scheme lists are as follows: DBS > DAS > MCS > CS.
When the calculations are based on C/I and C/(I+N):
• Atoll calculates the carrier-to-interference ratio for all the GPRS/EDGE TBC transmitters but takes into account all the
TBC transmitters (GSM and GPRS/EDGE) to evaluate the interference.
• The reception thresholds given for signal level C are internally converted to C/N thresholds ( N is the thermal noise
defined in the document database at -121 dBm by default) in order to be indexed by C/(I+N) values. C/I thresholds are
also indexed by the C/(I+N) value.
136
The selection of coding schemes is mainly based on the radio conditions mentionned
above. Nevertheless, you can optionally define some specific coding scheme graphs
accoding to a specific hopping mode, mobility type, frequency band and MAL. As an
example, you can model the gain due to longer MALs in coding scheme selection.
For more information on interference (I) calculation, see "DL Carrier-to-Interference Ratio Calculation" on page 131.
3.3.1 Coding Scheme Selection and Throughput Calculation

Without Ideal Link Adaptation
3.3.1.1 Calculations Based on C
Coding Scheme Selection
Atoll selects a coding scheme, cs, from among the coding schemes available in the GPRS/EDGE configuration, such that:
For each TRX type, tt, cs = Lowest  CS 

 Txi Txi
P rec  TRX  – P Backoff  TRX    Reception Threshold  CS

The selected coding scheme, cs, is the coding scheme with the lowest coding scheme number from the lowest priority coding
scheme list.
Throughput Calculation
Once the coding scheme cs is selected, Atoll reads the corresponding throughput value for the received signal level from the
Throughput=f(C) graph associated with cs.
3.3.1.2 Calculations Based on C/I

Atoll selects two coding schemes from among the coding schemes available in the GPRS/EDGE configuration, such that:
For each TRX type, tt, cs C = Lowest  CS 

 Txi Txi

 
 
And, cs C  I = Lowest  CS Txi Txi 
 P
rec
 TRX  – P
Backoff
 TRX 
 C 

 -------------------------------------------------------------------------  --- Threshold
I I  CS
csC is the coding scheme determined from the signal level, and csC/I is the coding scheme determined from the C/I level. Both
coding schemes are the coding schemes with the lowest coding scheme number from the lowest priority coding scheme list.
The selected coding scheme, cs, is the coding scheme with the lower coding scheme number among csC and csC/I:
cs = Min  cs C cs C  I  .
Throughput Calculation Based on the Worst Case Between C and C/I
For the coding scheme csC determined above, a throughput value, TPC, corresponding to the signal level is determined from
the TP = f(C) graph.
For the coding scheme csC/I determined above, a throughput value, TPC/I, corresponding to the C/I is determined from the TP
= f(C/I) graph.
The resulting throughput TP is the lower of the two values, TPC and TPC/I: TP = Min  TP C TP C  I  .
3.3.1.3 Calculations Based on C/(I+N)

137
 
 
For each TRX type, tt, cs C  N = Lowest  CS Txi Txi 
 P  TRX  – P  TRX  
-------------------------------------------------------------------------   ----------- Threshold 
rec Backoff C
 N I + N  CS
 
 
And, cs C   I + N  = Lowest  CS Txi Txi 
 P rec  TRX  – P Backoff  TRX  
-------------------------------------------------------------------------   ---------- 
C-
 I+N  I + N Threshold CS
csC/N is the coding scheme determined from the C/N, and csC/(I+N) is the coding scheme determined from the C/(I+N) level.
Both coding schemes are the coding schemes with the lowest coding scheme numbers from the lowest priority coding scheme
list.
The selected coding scheme, cs, is the coding scheme with the higher coding scheme number among csC/N and csC/(I+N):
cs = Max  cs C  N cs C   I + N   .
Throughput Calculation Based on Interpolation Between C/N and C/(I+N)
For the coding scheme csC/N determined above, the TP = f(C) graph is internally converted to TP = f(C/N) graph. A throughput
value, TPC/N, corresponding to the C/(I+N) is determined from the TP = f(C/N) graph.
For the coding scheme csC/(I+N) determined above, the TP = f(C/I) graph is internally converted to TP = f(C/(I+N)) graph. A
throughput value, TPC/(I+N), corresponding to the C/(I+N) is determined from the TP = f(C/(I+N)) graph.
The final throughput is computed by interpolating between the throughput values obtained from these two graphs. The
throughput interpolation method consists in interpolating TPC/N and TPC/(I+N) according to the respective weights of I and N
values.
The resulting throughput TP is given by: TP =   TP C  N +  1 –    TP C   I + N 
pN - , pN is the thermal noise power (value in Watts), and p(I+N) is the interferences + thermal noise power (value
 = -------------------
pI + N
in Watts).
3.3.2 Coding Scheme Selection and Throughput Calculation With

Ideal Link Adaptation
3.3.2.1 Calculations Based on C
Throughput Calculation
For the received signal level, and coding schemes whose reception thresholds are lower than the received signal level, Atoll
determines the highest throughput from the TP=f  C  graphs available in the GPRS/EDGE configuration.
 CS 
Txi Txi
TP C = Highest  TP=f  C = P rec  TRX  – P Backoff  TRX   
 Txi Txi

The selected coding scheme, cs, is the one corresponding to the highest throughput calculated above.
If there are more than one coding schemes providing the highest throughput at the pixel, the selected coding scheme, cs, is
the one with the lowest coding scheme number from the lowest priority coding scheme list.
3.3.2.2 Calculations Based on C/I

Throughput Calculation Based on Worst Case Between C and C/I
For the received signal level, and coding schemes whose reception thresholds are lower than the received signal level, Atoll
determines the highest throughput from the TP=f  C  graphs available in the GPRS/EDGE configuration.
138
 CS 
Txi Txi
TP C = Highest  TP=f  C = P rec  TRX  – P Backoff  TRX   
P
Txi
 TRX  – P
Txi
 TRX    Reception Threshold 

rec Backoff CS
For the received C/I, and coding schemes whose C/I thresholds are lower than the received C/I, Atoll determines the highest
throughput from the TP=f  C  I  graphs available in the GPRS/EDGE configuration.
Txi Txi
 
  P rec  TRX  – P Backoff  TRX    
TP C  I = Highest  TP=f  C  I = ----------------------------------------------------------------   CS Txi Txi 
I  P rec  TRX  – P Backoff  TRX 
 C 

 -------------------------------------------------------------------------  --- Threshold
I I  CS
The resulting throughput TP is the lower of the two values, TPC and TPC/I.
TP = Min  TP C TP C  I 
The selected coding scheme, cs, is the one corresponding to the lower of the two highest throughputs calculated above.
If there are more than one coding schemes providing the highest throughputs at the pixel, the selected coding scheme, cs, is
the one with the lowest coding scheme number from the lowest priority coding scheme list.

Throughput Calculation Based on Interpolation Between C/N and C/(I+N)
Atoll internally converts the TP = f(C) graphs into TP = f(C/N) graphs. For the received C/(I+N), and coding schemes whose C/
(I+N) thresholds are lower than the received C/(I+N), Atoll determines the highest throughput from the TP = f(C/N) graphs
available in the GPRS/EDGE configuration.
Txi Txi
 
  C P rec  TRX  – P Backoff  TRX    
TP C  N = Highest  TP=f  --- = ----------------------------------------------------------------   CS 
N I+N 
Txi
P rec  TRX  – P Backoff  TRX 
Txi

-------------------------------------------------------------------------   ----------- Threshold 
C
 I+N I + N  CS
Atoll internally converts the TP = f(C/I) graphs into TP = f(C/(I+N)) graphs. For the received C/(I+N), and coding schemes whose
C/(I+N) thresholds are lower than the received C/(I+N), Atoll determines the highest throughput from the TP = f(C/(I+N))
graphs available in the GPRS/EDGE configuration.
Txi Txi
 
  C P rec  TRX  – P Backoff  TRX    
TP C   I + N  = Highest  TP=f  ----------- = ----------------------------------------------------------------   CS 
I+N I+N 
Txi
P rec  TRX  – P Backoff  TRX 
Txi

-------------------------------------------------------------------------   ----------- Threshold 
C
 I+N I + N  CS
values.
pN -
 = ------------------- , pN is the thermal noise power (value in Watts), and p(I+N) is the interferences + thermal noise power (value
pI + N
in Watts).
The selected coding scheme, cs, is the one corresponding to the higher of the two highest throughputs calculated above.
If there are more than one coding schemes providing the highest throughputs at the pixel, the selected coding scheme, cs, is
the one with the highest coding scheme number from the highest priority coding scheme list.
3.3.3 Application Throughput Calculation

Application throughput is calculated from the effective RLC throughput as follows:
SF- – TP
TP Application = TP RLC  MAC  -------- Offset
100
139
TP RLC  MAC is the effective RLC throughput, and TP Offset and SF are the throughput offset (kbps) and the throughput scaling
factor (%) defined for the selected service.
3.3.4 BLER Calculation

Block error rate is calculated as follows:
 TP
 --------------- If  TP  TP MAX 
BLER =  TP MAX

 0 If  TP  TP MAX 
TP is the throughput per timeslot calculated for a pixel and TPMAX is the maximum throughput per timeslot read from the
GPRS/EDGE configuration used for the calculations.
3.3.5 GPRS/EDGE Coverage Predictions

Two GPRS/EDGE coverage predictions are available:
• Coverage by GPRS/EDGE Coding Scheme: Shows the areas various coding schemes are available.
• Packet Throughput and Quality Analysis: Shows the throughputs corresponding to the coding schemes available.
• The interference conditions to meet for a pixel to be covered, and

• Each transmitter, Txi, is GPRS/EDGE-capable.

Txi
SubcellReceptionThreshold  P rec  tt 

Txi

Txi Txj
ji
the highest.
140

Txi
Txi nd Txj
ji

Txi

Txi Txj
ji
the highest.
Txi
Txi nd Txj
ji

Txi

Txi Txj
ji
Txi
the highest.
141

Txi

Txi Txj
ji
Txi
the highest.
3.3.5.1.8 Best Idle Mode Reselection Criterion (C2)

Such type of coverage is useful:
Txi


Txi
SubcellReceptionThreshold  P rec  BCCH 

Txi Txj
And C2
j
integer.

142

It is possible to display the coverage predictions with colours depending on criteria such as:
Coverage by GPRS/EDGE Coding Scheme: Coding Schemes
Only the pixels with a coding scheme assigned are coloured. The pixel colour depends on the assigned coding scheme.
Coverage consists of several independent layers whose visibility in the map window can be managed. There are as many layers
as transmitter coverage areas. Each layer shows the coding schemes available in the transmitter coverage area.
Coverage by GPRS/EDGE Coding Scheme: Max Coding Schemes
On each pixel, Atoll chooses the highest coding scheme available from the TRXs of different transmitters covering that pixel.
Only the pixels with a coding scheme assigned are coloured. The pixel colour depends on the assigned coding scheme.
as possible coding schemes. Each layer shows the areas a given coding scheme can be used.
Packet Throughput and Quality Analysis: Effective RLC Throughput/Timeslot (kbps)
A pixel of the coverage area is coloured if the calculated effective RLC throughput per timeslot from any transmitter covering
that pixel exceeds the defined minimum threshold. The pixel colour depends on the effective RLC throughput per timeslot.
as transmitter coverage areas and throughput display thresholds. Each layer shows the effective RLC throughput that a
transmitter can provide on one timeslot.
Packet Throughput and Quality Analysis: Max Effective RLC Throughput/Timeslot (kbps)
A pixel of the coverage area is coloured if the calculated highest effective RLC throughput per timeslot from any transmitter
covering that pixel exceeds the defined minimum threshold. The pixel colour depends on the highest effective RLC throughput
per timeslot. Coverage consists of several independent layers whose visibility in the map window can be managed. There are
as many layers as throughput display thresholds. Each layer shows the best effective RLC throughput that any transmitter can
provide on one timeslot.
Packet Throughput and Quality Analysis: Average Effective RLC Throughput/Timeslot (kbps)
A pixel of the coverage area is coloured if the calculated average effective RLC throughput per timeslot from all the
transmitters covering that pixel exceeds the defined minimum threshold. The pixel colour depends on the average effective
RLC throughput per timeslot. Coverage consists of several independent layers whose visibility in the map window can be
managed. There are as many layers as throughput display thresholds. Each layer shows the average effective RLC throughput
that all the transmitters can provide on one timeslot.
Packet Throughput and Quality Analysis: Application Throughput/Timeslot (kbps)
A pixel of the coverage area is coloured if the calculated application throughput per timeslot from any transmitter covering
that pixel exceeds the defined minimum threshold. The pixel colour depends on the application throughput per timeslot.
as transmitter coverage areas and throughput display thresholds. Each layer shows the application throughput that a
transmitter can provide on one timeslot.
Packet Throughput and Quality Analysis: Best Application Throughput/Timeslot (kbps)
A pixel of the coverage area is coloured if the calculated highest application throughput per timeslot from any transmitter
covering that pixel exceeds the defined minimum threshold. The pixel colour depends on the highest application throughput
as many layers as throughput display thresholds. Each layer shows the best application throughput that any transmitter can
provide on one timeslot.
Packet Throughput and Quality Analysis: Average Application Throughput/Timeslot (kbps)
A pixel of the coverage area is coloured if the calculated average application throughput per timeslot from all the transmitters
covering that pixel exceeds the defined minimum threshold. The pixel colour depends on the average application throughput
as many layers as throughput display thresholds. Each layer shows the average application throughput that all the
transmitters can provide on one timeslot.
Packet Throughput and Quality Analysis: Effective RLC Throughput (kbps)
A pixel of the coverage area is coloured if the calculated effective RLC throughput from any transmitter covering that pixel
exceeds the defined minimum threshold. The pixel colour depends on the effective RLC for all the timeslots supported by the
143
selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). The number of DL timeslots is the
minimum between the number of DL timeslots defined in the selected terminal and service. Coverage consists of several
independent layers whose visibility in the map window can be managed. There are as many layers as transmitter coverage
areas and throughput display thresholds. Each layer shows the application throughput that a transmitter can provide on all
available timeslots in the terminal.
Packet Throughput and Quality Analysis: Max Effective RLC Throughput (kbps)
A pixel of the coverage area is coloured if the calculated highest effective RLC throughput from any transmitter covering that
pixel exceeds the defined minimum threshold. The pixel colour depends on the highest effective RLC throughput for all the
timeslots supported by the selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). The number
of DL timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service. Coverage
consists of several independent layers whose visibility in the map window can be managed. There are as many layers as
throughput display thresholds. Each layer shows the highest effective RLC throughput that any transmitter can provide on all
Packet Throughput and Quality Analysis: Average Effective RLC Throughput (kbps)
A pixel of the coverage area is coloured if the calculated average effective RLC throughput from all the transmitters covering
that pixel exceeds the defined minimum threshold. The pixel colour depends on the average effective RLC throughput for all
the timeslots supported by the selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). The
number of DL timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service.
as throughput display thresholds. Each layer shows the average effective RLC throughput that all the transmitters can provide
on all available timeslots in the terminal.
Packet Throughput and Quality Analysis: Application Throughput (kbps)
A pixel of the coverage area is coloured if the calculated application throughput from any transmitter covering that pixel
exceeds the defined minimum threshold. The pixel colour depends on the application throughput for all the timeslots
supported by the selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). The number of DL
timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service. Coverage consists
of several independent layers whose visibility in the map window can be managed. There are as many layers as transmitter
coverage areas and throughput display thresholds. Each layer shows the application throughput that a transmitter can provide
on all available timeslots in the terminal.
Packet Throughput and Quality Analysis: Max Application Throughput (kbps)
A pixel of the coverage area is coloured if the calculated highest application throughput from any transmitter covering that
pixel exceeds the defined minimum threshold. The pixel colour depends on the highest application throughput for all the
throughput display thresholds. Each layer shows the highest application throughput that any transmitter can provide on all
Packet Throughput and Quality Analysis: Average Application Throughput (kbps)
A pixel of the coverage area is coloured if the calculated average application throughput from all the transmitters covering
that pixel exceeds the defined minimum threshold. The pixel colour depends on the average application throughput for all the
throughput display thresholds. Each layer shows the average application throughput that all the transmitters can provide on
all available timeslots in the terminal.
Packet Throughput and Quality Analysis: Throughput per User (kbps)
A pixel of the coverage area is coloured if the calculated throughput per user from any transmitter covering that pixel exceeds
the defined minimum threshold. The pixel colour depends on the throughput per user for all the timeslots supported by the
selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). The number of DL timeslots is the
minimum between the number of DL timeslots defined in the selected terminal and service. The throughput per user is
calculated by applying the throughput reduction factor, determined using the selected dimensioning model, to the application
throughput. Coverage consists of several independent layers whose visibility in the map window can be managed. There are
as many layers as transmitter coverage areas and throughput display thresholds. Each layer shows the throughput per user
that a transmitter can provide on all available timeslots in the terminal.
144
Packet Throughput and Quality Analysis: Max Throughput per User (kbps)
A pixel of the coverage area is coloured if the calculated highest throughput per user from any transmitter covering that pixel
exceeds the defined minimum threshold. The pixel colour depends on the highest throughput per user for all the timeslots
supported by the selected terminal type (Number of Simultaneous Carriers x Number of DL Timeslots). The number of DL
timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service. The throughput
per user is calculated by applying the throughput reduction factor, determined using the selected dimensioning model, to the
application throughput. Coverage consists of several independent layers whose visibility in the map window can be managed.
There are as many layers as throughput display thresholds. Each layer shows the highest throughput per user that any
transmitter can provide on all available timeslots in the terminal.
Packet Throughput and Quality Analysis: Average Throughput per User (kbps)
A pixel of the coverage area is coloured if the calculated average throughput per user from all the transmitters covering that
pixel exceeds the defined minimum threshold. The pixel colour depends on the average throughput per user for all the
of DL timeslots is the minimum between the number of DL timeslots defined in the selected terminal and service. The
throughput per user is calculated by applying the throughput reduction factor, determined using the selected dimensioning
model, to the application throughput. Coverage consists of several independent layers whose visibility in the map window can
be managed. There are as many layers as throughput display thresholds. Each layer shows the average throughput per user
that all the transmitters can provide on all available timeslots in the terminal.
Packet Throughput and Quality Analysis: BLER (%)
A pixel of the coverage area is coloured if the calculated BLER from any transmitter exceeds the defined minimum threshold.
The pixel colour depends on the BLER. Coverage consists of several independent layers whose visibility in the map window
can be managed. There are as many layers as transmitter coverage areas and BLER display thresholds. Each layer shows the
BLERs that the covered pixels experience on one timeslot.
Packet Throughput and Quality Analysis: Max BLER (%)
A pixel of the coverage area is coloured if the calculated highest BLER from all the transmitters exceeds the defined minimum
threshold. The pixel colour depends on the BLER. Coverage consists of several independent layers whose visibility in the map
window can be managed. There are as many layers as BLER display thresholds. Each layer shows the BLER that the covered
pixels experience on one timeslot.
3.4 Codec Mode Selection and CQI Calculations

Atoll supports FR, HR, EFR, and AMR codec modes. A codec configuration contains codec mode adaptation thresholds and
quality graphs for circuit quality indicators. Atoll has the following circuit quality indicators included by default:
• FER or Frame Erasure Rate: The number of frames in error divided by the total number of frames. These frames are
usually discarded, in which case this can be called the Frame Erasure Rate.
• BER or Bit Error Rate: BER is a measurement of the raw bit error rate in reception before the decoding process begins.
Any factor that impacts the decoding performance, such as frequency hopping, will impact the correlation between
BER and FER, or the perceived end-user voice quality.
• MOS or Mean Opinion Score: Voice quality can be quantified using mean opinion score (MOS). MOS values can only
be measured in a test laboratory environment. MOS values range from 1 (bad) to 5 (excellent). Different voice codecs
have slightly different FER to MOS correlation since the smaller the voice codec bit rate is, the more sensitive it
becomes to frame erasures.
The default codec configurations in Atoll include default FER, BER, and MOS quality graphs with respect to the carrier to
interference ratio, and codec mode adaptation thresholds (calculated from the FER vs. C/I graphs for all codec modes at 5 %
FER).
145
Figure 3.1: FER vs. C/I Graphs
Figure 3.2: BER vs. C/I Graphs
Figure 3.3: MOS vs. C/I Graphs
146
The graphs are based on:

[1] T. Halonen, J. Romero, J. Melero; GSM, GPRS and EDGE performance – Evolution
towards 3G/UMTS, John Wiley and Sons Ltd.
[2] J. Wigard, P. Mogensen; A simple mapping from C/I to FER and BER for a GSM type of
air interface.
[3] 3GPP Specifications TR 26.975 V6.0.0; Performance characterization of the Adaptive
Multi-Rate (AMR) speech codec (Release 6)
3.4.1 Circuit Quality Indicator Calculations

Circuit quality indicator calculations include codec mode selection and CQI calculation. Codec modes may be selected using
ideal link adaptation or without it. Once codec modes have been selected, CQI corresponding to these codec modes are
determined from the look-up tables.
The following sections describe the two categories of calculations, i.e., with and without ideal link adaptations. Ideal link
adaptation implies that the selected codec mode corresponds to the best value of the reference CQI under the given radio
conditions. Without ideal link adaptation, the codec mode is selected based on the codec adaptation thresholds.
CQI calculations may be based on C/N or on C/(I+N). For calculating the noise, either the noise figure defined for the
calculations or that of the selected terminal type is used.
Different codec configurations may be defined for transmitter and terminals. In this case, Atoll only selects the coding
schemes that are common in the two, and gives priority to the thresholds defined in the terminal configuration. If no terminal
type is defined for the calculation, or if the terminal type does not have any codec configuration assigned to it, Atoll only uses
the codec configuration of the transmitter. If the transmitter does not have any codec configuration assigned to it, no codec
mode selection and CQI calculation is carried out.
If more than one codec modes satisfy the C/N or C/I conditions, Atoll selects the higher priority codec mode.
Txi
• P rec  TRX  is the signal level received from the selected TRX type (tt) or on all the TRXs of Txi on each pixel of the Txi
coverage area,
• CM is the set of all available codec modes,
•  Adaptation Threshold  CM are the values of adaptation thresholds for the codec modes available in the codec
configuration,
The computed noise N is compared to the codec configuration reference noise N Ref . If the values are the same, the
defined graphs are used as is, otherwise the graphs are downshifted by the difference N – N Ref .
When the calculations are based on C/(I+N):

• Atoll calculates the carrier-to-interference ratio for all the TBC transmitters with codec configurations assigned, but
takes into account all the TBC transmitters (with and without codec configurations) to evaluate the interference.
The selection of codec modes is mainly based on the radio conditions mentionned above.
Nevertheless, you can optionally define some specific codec mode graphs accoding to a
specific hopping mode, mobility type, frequency band and MAL. As an example, you can
model the gain due to longer MALs in codec mode selection.
For more information on interference (I) calculation, see "DL Carrier-to-Interference Ratio Calculation" on page 131.
Ideal link adaptation for circuit quality indicator studies is defined at the codec configuration level. If the ideal link adaptation
option is checked, Atoll will select the codec mode, for the transmitter under study, according to the codec quality graphs (CQI
= f(C/N) and CQI = f(C/I)) related to the defined reference CQI, which may be different from the CQI being calculated.
Otherwise, Atoll will use the adaptation thresholds defined in the Adaptation Thresholds tab to determine the codec mode
to be used in the studies.
3.4.2 CQI Calculation Without Ideal Link Adaptation

3.4.2.1 Calculations Based on C/N
Atoll selects the highest priority codec mode, cm, from among the codec modes available in the codec configuration:
147
 
For each TRX type, tt, cm = Highest Priority  CM Txi


P  TRX 
 ---------------------------   Adaptation Threshold  
rec
N CM
Txi
P rec  TRX 
For ------------------------ , Atoll determines the CQI from the CQI=f(C/N) graph associated to the selected codec mode, cm.
N

 
For each TRX type, tt, cm = Highest Priority  CM 
 Txi
P rec  TRX 

 ---------------------------   Adaptation Threshold  
I+N CM
Txi
P rec  TRX 
For -----------------------
- , Atoll determines the CQI from the CQI=f(C/I) graph associated to the selected codec mode, cm.
I+N
3.4.3 CQI Calculation With Ideal Link Adaptation

3.4.3.1 Calculations Based on C/N
Ideal link adaptation is used by a codec configuration according to a defined reference CQI (MOS by default).
Atoll calculates signal level received from Txi on each pixel of Txi coverage area and converts it into C/N values as described
earlier. Then, Atoll filters all the codec modes that satisfy the C/N criterion (defined by the CQI = f(C/N) graphs for the
reference CQI) and are common between the transmitter and the terminal type codec configuration.
The selected codec mode among these filtered codec modes will be,
 
 
For each TRX type, tt, cm = Highest Priority  CM  , for MOS
   P
Txi
 

 TRX
CQI Ref = Highest  CQI=f  ---= --------------------------- 
C rec 
  N N tot   
  
 
 
Or, cm = Highest Priority  CM  , for BER and FER
   P
Txi
  

 TRX
CQI Ref = Lowest  CQI=f  ---= --------------------------- 
C rec 
 N N tot 
  
, cm is the codec mode with the highest priority among the set of codec modes CM for which the reference CQI gives the
Txi
P rec  TRX 
highest or the lowest value at the received C/N level, -----------------------
- .
N tot
If more than one codec mode graphs give the same value for reference CQI, then Atoll selects the codec mode with the highest
priority.
From the CQI = f(C/N) graph associated to the selected codec mode cm, Atoll evaluates the CQI for which the study was
Txi
P rec  TRX 
performed corresponding to ------------------------ for the selected codec mode.
N tot

Ideal link adaptation is used by a codec configuration according to a defined reference CQI (MOS by default).
Atoll calculates the C/I level received from the transmitter on each pixel of Txi coverage area, for each TRX and converts it into
C/(I+N). Then, Atoll filters all the codec modes that satisfy the C/(I+N) criteria (defined by the CQI = f(C/I) graphs for the
reference CQI) and are common between the transmitter and the terminal type codec configuration.
The selected codec mode among these filtered codec modes will be,
148
 
 
For each TRX type, tt, cm = Highest Priority  CM  , for MOS
   P Txi  TRX  

 = Highest  CQI=f  --- = --------------------------- 
C rec 
 CQI
Ref  I I+N 
  tot
 
 
Or, cm = Highest Priority  CM  , for BER and FER
   P
Txi
 TRX  

 CQI Ref = Lowest  CQI=f  --- = --------------------------- 
C rec 
  I I+N 
  
tot
, cm is the codec mode with the highest priority among the set of codec modes CM for which the reference CQI gives the
Txi
P rec  TRX 
highest or the lowest value at the received C/(I+N) level, -----------------------
-.
I + N tot
If more than one codec mode graphs give the same value for reference CQI, then Atoll selects the codec mode with the highest
priority.
From the CQI = f(C/I) graph associated to the selected codec mode cm (indexed with the C/(I+N) values), Atoll evaluates the
Txi
P rec  TRX 
CQI for which the study was performed corresponding to -----------------------
- for the selected codec mode.
I + N tot
3.4.4 Circuit Quality Indicators Coverage Predictions

The Circuit Quality Indicators coverage predictions show the areas BER, FER, and MOS values in the transmitter coverage
areas.
• The interference and quality indicator conditions to meet for a pixel to be covered, and

We can distinguish seven cases as below. Let us assume that:
• Each transmitter, Txi, has a codec configuration assigned.

Txi

Txi

Txi Txj
ji
149
the highest.

Txi
Txi nd Txj
ji

Txi

Txi Txj
ji
the highest.
Txi
Txi nd Txj
ji

Txi

Txi Txj
ji
Txi
And the received P rec  BCCH  exceeds the reception threshold defined per HCS layer.
150
the highest.

Txi

Txi Txj
ji
Txi
the highest.


It is possible to display the coverage predictions with colours depending on criteria such as:
BER
Only the pixels with a codec mode assigned are coloured. The pixel colour depends on the BER value. Coverage consists of
several independent layers whose visibility in the map window can be managed. There are as many layers as transmitter
coverage areas and BER display thresholds. Each layer shows the BER in the transmitter coverage area.
FER
Only the pixels with a codec mode assigned are coloured. The pixel colour depends on the FER value. Coverage consists of
coverage areas and FER display thresholds. Each layer shows the FER in the transmitter coverage area.
MOS
Only the pixels with a codec mode assigned are coloured. The pixel colour depends on the MOS value. Coverage consists of
coverage areas and MOS display thresholds. Each layer shows the MOS in the transmitter coverage area.
151
Max BER
Only the pixels with a codec mode assigned are coloured. The pixel colour depends on the highest BER value among the BER
values for all the transmitters covering the pixel. Coverage consists of several independent layers whose visibility in the map
window can be managed. There are as many layers as BER display thresholds. Each layer shows the BER value.
Max FER
Only the pixels with a codec mode assigned are coloured. The pixel colour depends on the highest FER value among the FER
window can be managed. There are as many layers as FER display thresholds. Each layer shows the FER value.
Max MOS
Only the pixels with a codec mode assigned are coloured. The pixel colour depends on the highest MOS value among the MOS
window can be managed. There are as many layers as MOS display thresholds. Each layer shows the MOS value.
3.5 UL Coverage Predictions

For each TBC transmitter, Txi, Atoll calculates the selected parameter at each Txi inside its calculation area. Results are shown
on each pixel, each pixel acting as a transmitting terminal. Hence, transmitters are here (non-interfering) receivers.
• The coverage conditions in order to determine the DL service area of each TBC transmitter, and
Two interfaced predictions are available:
• One prediction which shows on each pixel UL losses or UL signal levels
• One prediction which shows on each pixel UL C/I levels.
Additional studies such as codec modes and coding schemes predictions are used during simulations but are not graphically
available.
3.5.1 DL Service Area Determination

Atoll uses parameters entered in the Condition tab of the coverage prediction properties dialogue to determine the service
areas of the TBC transmitters.
3.5.1.1 All Servers

Txi
Txi
Txi Txi
3.5.1.2 Best Signal Level and a Margin

Txi
Txi
Txi Txi
152

Txi Txj
ji
the highest.
3.5.1.3 Second Best Signal Level and a Margin

The service area of Txi corresponds to the pixels:
Txi
Txi
Txi Txi
Txi nd Txj
ji
3.5.1.4 Best Signal Level per HCS Layer and a Margin

Txi
Txi
Txi Txi

Txi Txj
ji
the highest.
3.5.1.5 Second Best Signal Level per HCS Layer and a Margin
Txi
153
Txi
Txi Txi
Txi nd Txj
ji
3.5.1.6 HCS Servers and a Margin

Txi
Txi
Txi Txi

Txi Txj
ji
Txi
the highest.
3.5.1.7 Highest Priority HCS Server and a Margin

Txi
Txi
Txi Txi

Txi Txj
ji
Txi
154
the highest.
3.5.1.8 Best Idle Mode Reselection Criterion (C2)

Such type of coverage is useful :
Txi


Txi
MinimumThreshold  P rec  BCCH   MaximumThreshold
Txi
Txi Txi

Txi Txj
And C2
j
integer.
3.5.2 Coverage by UL Signal Level

3.5.2.1 Coverage Resolution
3.5.2.2 Display Types

UL signal levels and UL losses calculations are explained in "UL Signal Level" on page 125. It is possible to display the coverage
by UL signal level with colours depending on any transmitter attribute or other criteria such as:
3.5.2.2.1 UL Signal Level (in dBm, dBµV, dBµV/m)

Atoll calculates the signal level received at each transmitter on its service area from surrounding pixels. A pixel of a service
area is coloured if the UL signal level is greater than or equal to the defined minimum thresholds (pixel colour depends on
signal level). Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as
many layers as transmitter service areas. Each layer shows the different UL signal levels at the transmitter on its service area.
3.5.2.2.2 Best UL Signal Level (in dBm, dBµV, dBµV/m)

Atoll calculates the signal level received at each transmitter on its service area from surrounding pixels. When other service
areas overlap the studied one, Atoll chooses the highest value. A pixel of a service area is coloured if the UL signal level is
155
greater than or equal to the defined thresholds (the pixel colour depends on the signal level). Coverage consists of several
independent layers whose visibility in the workspace can be managed. There are as many layers as transmitter service areas.
Each layer shows the different UL signal levels at the transmitter on its service area.
3.5.2.2.3 UL Total Losses (dB)

Atoll calculates total losses from the terminal at each transmitter on its service area. A pixel of a service area is coloured if UL
total losses are greater than or equal to the defined minimum thresholds (pixel colour depends on UL total losses). Coverage
areas. Each layer shows the different UL total losses at the transmitter on its service area.
3.5.2.2.4 Minimum UL Total Losses (dB)

Atoll calculates total losses from the transmitter on each pixel of each transmitter service area. When other service areas
overlap the studied one, Atoll chooses the lowest value. A pixel of a service area is coloured if UL total losses are greater than
or equal to the defined minimum thresholds (pixel colour depends on total losses). Coverage consists of several independent
layers whose visibility in the workspace can be managed. There are as many layers as service areas. Each layer shows the
different total losses levels in the transmitter service area.
3.5.3 Coverage by UL C/I

An UL C/I coverage predictions is available. It provides the UL C/I level at the transmitter level caused by surrounding uplink
traffic.
3.5.3.1 Coverage Resolution

3.5.3.2 UL C/I Evaluation

The UL C/I level can be computed as follows, for a given MSA
C MSA Term Tx
--- = P rec – N tot
I UL
,
Tx Tx MSA DL
• N tot = N thermal + NF + NR Intra – techno log y is the UL total noise at transmitter on the considered MSA
Term
• P rec is the received signal level at the transmitter,
• N thermal is the thermal noise (-121 dBm by default or user-defined)
Tx
• NF is the transmitter noise figure
MSA DL
• NRIntra – techno log y is the intra-technology UL noise rise at the considered MSA. Since UL noise rise are defined per
TRX, Atoll takes the TRX UL noise rise in case of non-hopping or extracts a mean noise rise from the several TRXs
composing the MSA in case of Base Band Hopping or Synthesized Frequency Hopping.
MSA
For a given transmitter having several MSAs, all possible C --- are displayed in case the detailed results box is selected. If
I UL
not, the worst results (the min C/I per transmitter) are retained.
3.5.3.3 Coverage Area Determination

MSA
For each MSA, coverage area corresponds to pixels where C
--- is between the lower and upper thresholds defined in the
I UL
coverage prediction properties.
3.5.3.4 Display Types

3.5.3.4.1 C/I Level

Each pixel of the transmitter coverage area is coloured if the calculated UL C/I level is greater than or equal to the specified
minimum thresholds (pixel colour depends on UL C/I level). Coverage consists of several independent layers whose visibility
156
in the workspace can be managed. There are as many layers as transmitter service areas. Each layer shows the different UL
C/I levels available in the transmitter coverage area.
3.5.3.4.2 Max C/I Level

Atoll compares calculated UL C/I levels received from transmitters on each pixel of each transmitter coverage area coverage
areas overlap the studied one and chooses the highest value. A pixel of a coverage area is coloured if the UL C/I level is greater
than or equal to the specified thresholds (the pixel colour depends on the UL C/I level). Coverage consists of several
layer corresponds to an area the highest received UL C/I level exceeds a defined minimum threshold.
3.5.3.4.3 Min C/I Level

Atoll compares UL C/I levels received from transmitters on each pixel of each transmitter coverage area the coverage areas
overlap the studied one and chooses the lowest value. A pixel of a coverage area is coloured if the UL C/I level is greater than
or equal to the specified thresholds (the pixel colour depends on the UL C/I level). Coverage consists of several independent
layers whose visibility in the workspace can be managed. There are as many layers as defined thresholds. Each layer
corresponds to an area the lowest received UL C/I level exceeds a defined minimum threshold.
3.5.4 Coverage by UL Coding Schemes

An UL Coding Scheme coverage prediction is implemented in order to be used in simulations. The prediction itself does not
have any interface. for the simulations, settings are hard coded and are described hereafter.
These calculations are based on C/(I+N). Coding schemes are selected without using ideal link adaptation.
Different GPRS/EDGE configurations may be defined for transmitter and terminals. In this case, Atoll only selects the coding
schemes that are common in the two, and gives priority to the thresholds defined in the transmitter configuration. If no
terminal type is defined for the calculation, or if the terminal type does not have any GPRS/EDGE configuration assigned to it,
Atoll only uses the GPRS/EDGE configuration of the transmitter. If the transmitter does not have any GPRS/EDGE
configuration assigned to it, no coding scheme selection and throughput calculation is carried out.
Txi
• P rec  TRX  is the DL signal level received from the BCCH of Txi on each pixel of the Txi coverage area,
Term
• P rec is the UL the signal level received at each transmitter on its service area from surrounding pixels
• CS is the set of all available coding schemes,
•  Reception Threshold  CS are the values of reception thresholds for the coding schemes available in the GPRS/EDGE
configuration,
•  -----------
C
Threshold
I + N  CS are the values of C/(I+N) thresholds for the coding schemes available in the GPRS/EDGE
configuration,
• The priorities of the coding scheme lists are as follows: DBS > DAS > MCS > CS.
Since the calculations are based on C/I and C/(I+N):
• Atoll calculates the UL C/I to all the GPRS/EDGE TBC transmitters.
• The reception thresholds given for signal level C are internally converted to C/N thresholds (N is the thermal noise
defined in the document database at -121 dBm by default) in order to be indexed by C/(I+N) values. C/I thresholds are
also indexed by the C/(I+N) value.
For more information on UL C/I calculation, see "Coverage by UL C/I" on page 156.

Atoll uses hard-coded parameters for simulations. In that case, the DL service area is based on the option "HCS servers" with
a margin of 4 dB.
Txi
SubcellReceptionThreshold  Prec  tt 
And P rec  BCCH   Best  P rec  BCCH   – 4dB

Txi Txj
ji
Txi
157
3.5.4.2 Coding Scheme Selection

 
 
For each MSA , cs C  N = Lowest  CS Term 
 P rec 
-----------------   ---------- 
C-
 N I + N
Threshold
 CS
 
 
And, cs C   I + N  = Lowest  CS Term 
 P rec 
-----------------   ----------- Threshold 
C
 I+N I + N  CS
csC/N is the coding scheme determined from the C/N, and csC/(I+N) is the coding scheme determined from the C/(I+N) level.
Both coding schemes are the coding schemes with the lowest coding scheme numbers from the lowest priority coding scheme
list.
The selected coding scheme, cs, is the coding scheme with the higher coding scheme number among csC/N and csC/(I+N):
cs = Max  cs C  N cs C   I + N   .
3.5.4.3 Throughput Calculation

For the coding scheme csC/N determined above, the TP = f(C) graph is internally converted to TP = f(C/N) graph. A throughput
value, TPC/N, corresponding to the C/(I+N) is determined from the TP = f(C/N) graph.
For the coding scheme csC/(I+N) determined above, the TP = f(C/I) graph is internally converted to TP = f(C/(I+N)) graph. A
throughput value, TPC/(I+N), corresponding to the C/(I+N) is determined from the TP = f(C/(I+N)) graph.
values.
pN -
 = ------------------- , pN is the thermal noise power (value in Watts), and p(I+N) is the interferences + thermal noise power (value
pI + N
in Watts).
3.5.5 Coverage by UL Codec Modes

An UL Codec Mode coverage prediction is implemented in order to be used in simulations. The prediction itself does not have
any interface. for the simulations, settings are hard coded and are described hereafter.
Circuit quality indicator calculations include codec mode selection and CQI calculation. Codec modes are selected according
to C/(I+N) quality without using ideal link adaptation. Once codec modes have been selected, CQI and number of used
timeslots (0.5 in case of HR) corresponding to these codec modes are determined from the look-up tables.
Different codec configurations may be defined for transmitter and terminals. In this case, Atoll only selects the coding
schemes that are common in the two, and gives priority to the thresholds defined in the transmitter configuration. If no
terminal type is defined for the calculation, or if the terminal type does not have any codec configuration assigned to it, Atoll
only uses the codec configuration of the transmitter. If the transmitter does not have any codec configuration assigned to it,
no codec mode selection and CQI calculation is carried out.
If more than one codec modes satisfy the quality conditions, Atoll selects the higher priority codec mode.
Txi
• P rec  TRX  is the DL signal level received from the BCCH of Txi on each pixel of the Txi coverage area,
Term
• P rec is the UL the signal level received at each transmitter on its service area from surrounding pixels
• CM is the set of all available codec modes,
•  Adaptation Threshold  CM are the values of adaptation thresholds for the codec modes available in the codec
configuration,
The computed noise N is compared to the codec configuration reference noise N Ref . If the values are the same, the
defined graphs are used as is, otherwise the graphs are downshifted by the difference N – N Ref .
158
Since the calculations are based on C/I and C/(I+N):

• Atoll calculates the UL C/I to all the GPRS/EDGE TBC transmitters.
For more information on UL C/I calculation, see "Coverage by UL C/I" on page 156.

Atoll uses hard-coded parameters for simulations. In that case, the DL service area is based on the option "HCS servers" with
a margin of 4 dB.
Txi
SubcellReceptionThreshold  Prec  BCCH 
And P rec  BCCH   Best  P rec  BCCH   – 4dB

Txi Txj
ji
Txi
3.5.5.2 Codec Mode Selection

 
For each MSA, cm = Highest Priority  CM 
 Term
P rec

 -----------------   Adaptation Threshold  
I+N CM
Term
P rec
For ------------
- , Atoll determines the CQI from the CQI=f(C/I) graph associated to the selected codec mode, cm.
I+N
3.6 Traffic Analysis

When starting a traffic analysis, Atoll distributes the traffic from maps to transmitters of each layer according to the
compatibility criteria defined in the transmitter, services, mobility type, terminal type properties. Transmitters considered in
traffic analysis are the active and filtered transmitters that belong to the focus zone.
• If no focus zone exists in the .atl document, Atoll takes into account the computation
zone.
• For details of the average timeslot capacity calculation, see the Network
Dimensioning section (calculation of minimum reduction factor).
3.6.1 Traffic Distribution

3.6.1.1 Normal Cells (Nonconcentric, No HCS Layer)
3.6.1.1.1 Circuit Switched Services
A user with a given circuit switched service, c, a terminal, t, and a mobility type, m, will be distributed to the BCCH and TCH
subcells of a transmitter if:
• The terminal, t, works on the frequency band used by the BCCH subcell,
• The terminal, t, works on the frequency band used by the TCH subcell.
3.6.1.1.2 Packet Switched Services

A user with a given packet switched service, p, a terminal, t, and a mobility type, m, will be distributed to the BCCH and TCH
subcells of a transmitter if:
• The transmitter is an GPRS/EDGE station (option specified in the transmitter property dialogue),
• The terminal, t, is technologically compatible with the transmitter,
• The terminal, t, works on the frequency band used by the TCH subcell.
159
3.6.1.2 Concentric Cells

In case of concentric cells, TCH_INNER TRX type has the highest priority to carry traffic.

A user with a given circuit switched service, c, a terminal, t, and a mobility type, m, will be distributed to the TCH_INNER, BCCH
and TCH subcells of a transmitter if:
• The terminal, t, works on the frequency band(s) used by the TCH_INNER and TCH subcells.

A user with a given packet switched service, p, a terminal, t, and a mobility type, m, will be distributed to the TCH_INNER,
BCCH and TCH subcells of a transmitter if:
• The terminal, t, works on the frequency band(s) used by the TCH_INNER and TCH subcells.
3.6.1.3 HCS Layers

For each HCS layer, k, you may specify the maximum mobile speed supported by the transmitters of the layer.

A user with a given circuit switched service, c, a terminal, t, and a mobility type, m, will be distributed to the BCCH and TCH
subcells (and TCH_INNER in case of concentric cells) of a transmitter if:
• The terminal, t, works on the frequency band(s) used by the TCH_INNER and TCH subcells,
• The user’s mobility, m, is less than the maximum speed supported by the layer, k.

A user with a given packet switched service, p, a terminal, t, and a mobility type, m, will be distributed to the BCCH and TCH
subcells (and TCH_INNER in case of concentric cells) of a transmitter if:
• The terminal, t, works on the frequency band(s) used by the TCH_INNER and TCH subcells,
• The user mobility, m, is less than the maximum speed supported by the layer, k.
3.6.2 Calculation of the Traffic Demand per Subcell

Here we assume that:
• Users considered for evaluating the traffic demand fulfil the compatibility criteria defined in the transmitter, services,
mobility, terminal properties as explained above.
• Atoll distributes traffic on subcell service areas, which are determined using the option “Best signal level per HCS
layer” with a 0dB margin and the subcell reception threshold as lower threshold.
• Same traffic is distributed to the BCCH and TCH subcells.
3.6.2.1 User Profile Traffic Maps

3.6.2.1.1 Normal Cells (Nonconcentric, No HCS Layer)
Number of subscribers ( X up m ) for each TCH subcell (Txi, TCH), per user profile up with a given mobility m, is inferred as:
X up m  Txi TCH  = S up m  Txi TCH   D
Sup,m is the TCH service area containing the user profile up with the mobility m and D is the user profile density.
For each behaviour described in the user profile up, Atoll calculates the probability for the user to be connected with a given
service using a terminal t.
Circuit Switched Services
For a circuit switched service c, we have:
160
N call  d
p up  c t  = -------------------
-
3600
Ncall is the number of calls per hour and d is the average call duration (in seconds).
Then, Atoll evaluates the traffic demand, D up  c t  m , in Erlangs for the subcell (Txi, TCH) service area.
D up  c t  m  Txi TCH  = X up m  Txi TCH   p up  c t 
Packet Switched Services (Max Rate)
For a max rate packet switched service p, we have:
N call  V  8
p up  p t  = -----------------------------
-
3600
Ncall is the number of calls per hour and V is the transmitted data volume per call (in Kbytes).
Then, Atoll evaluates the traffic demand, D up  p t  m , in kbits/s for the subcell (Txi, TCH) service area.
D up  p t  m  Txi TCH  = X up m  Txi TCH   p up  p t 
Packet Switched Services (Constant Bit Rate)
For a constant bit packet switched service p, we have:
N call  d
p up  p t  = -------------------
-
3600
Ncall is the number of calls per hour and d is the average call duration (in seconds).
Then, Atoll evaluates the traffic demand, D up  p t  m , in kbits/s for the subcell (Txi, TCH) service area.
D up  p t  m  Txi TCH  = X up m  Txi TCH   p up  p t 
3.6.2.1.2 Concentric Cells

In case of concentric cells, Atoll distributes a part of traffic on the TCH_INNER service area (TCH_INNER is the highest priority
traffic carrier) and the remaining traffic on the outer ring served by the TCH subcell. The traffic spread over the TCH_INNER
subcell may overflow to the TCH subcell. In this case, the traffic demand is the same on the TCH_INNER subcell but increases
on the TCH subcell.
• Traffic overflowing from the TCH_INNER to the TCH is not uniformly spread over the
TCH service area. It is still located on the TCH_INNER service area.
Number of subscribers ( X up m ) for each TCH_INNER (Txi, TCH_INNER) and TCH (Txi, TCH) subcell, per user profile up with a
given mobility m, is inferred as:
X up m  Txi,TCH_INNER  = S up m  Txi,TCH_INNER   D
X up m  Txi,TCH  =  S up m  Txi,TCH  – S up m  Txi,TCH_INNER    D
S up m  Txi,TCH_INNER  and S up m  Txi,TCH  respectively refer to the TCH_INNER and TCH subcell service areas containing
the user profile up with the mobility m. D is the user profile density.
Figure 3.4: Representation of a Concentric Cell TXi
161
Circuit Switched Services
For each user of the user profile up using a circuit switched service c with a terminal t, Atoll calculates the probability
( p up  c t  ) of the user being connected. Calculations are detailed in "Circuit Switched Services" on page 159.
Then, Atoll evaluates the traffic demand, D up  c t  m , in Erlangs in the (Txi, TCH_INNER) and (Txi, TCH) subcell service areas.
D up  c t  m  Txi,TCH_INNER  = X up m  Txi,TCH_INNER   p up  c t 
D up  c t  m  Txi,TCH  = X up m  Txi,TCH   p up  c t  + D up  c t  m  Txi,TCH_INNER   O max  Txi,TCH_INNER 
O max  Txi,TCH_INNER  is the maximum rate of traffic overflow (in %) specified for the TCH_INNER subcell.
Packet Switched Services (Max Rate)
For each user of the user profile up using a max rate packet switched service p with a terminal t, probability of the user being
connected ( p up  p t  ) is calculated as explained in "Packet Switched Services" on page 159.
Atoll evaluates the traffic demand, D up  p t  m , in kbits/s in the (Txi, TCH_INNER) and (Txi, TCH) subcell service areas.
D up  p t  m  Txi,TCH_INNER  = X up m  Txi,TCH_INNER   p up  p t 
D up  p t  m  Txi,TCH  = X up m  Txi,TCH   p up  p t  + D up  p t  m  Txi,TCH_INNER   O max  Txi,TCH_INNER 
Packet Switched Services (Constant Bit Rate)
For each user of the user profile up using a constant bit packet switched service p with a terminal t, probability of the user
being connected ( p up  p t  ) is calculated as explained in "Packet Switched Services" on page 159.
Atoll evaluates the traffic demand, D up  p t  m , in kbits/s in the (Txi, TCH_INNER) and (Txi, TCH) subcell service areas.
D up  p t  m  Txi,TCH_INNER  = X up m  Txi,TCH_INNER   p up  p t 
D up  p t  m  Txi,TCH  = X up m  Txi,TCH   p up  p t  + D up  p t  m  Txi,TCH_INNER   O max  Txi,TCH_INNER 
3.6.2.1.3 HCS Layers

We assume two HCS layers: the micro layer has a higher priority than the macro layer. Txi belongs to the micro layer and Txj
to the macro. The traffic contained in the input traffic map can be assigned to all the HCS layers.
Normal Cells
Atoll distributes traffic on the TCH service areas. The traffic capture is calculated with the option “Best signal level per HCS
macro
layer” meaning that there is an overlap between HCS layers service areas. Let S overlapping  Txj TCH  denote this area (TCH
service area of the macro layer overlapped by the TCH service area of the micro layer). Traffic on the overlapping area is
distributed to the TCH subcell of the micro layer because it has a higher priority. On this area, traffic of the micro layer may
overflow to the macro layer. In this case, the traffic demand is the same on the TCH subcell of the micro layer but increases
on the TCH subcell of the macro layer.
Traffic overflowing to the macro layer is not uniformly spread over the TCH service area
of Txj. It is only located on the overlapping area.
162
Figure 3.5: Representation of Micro and Macro Layers
Atoll evaluates the traffic demand on the micro layer (higher priority) as explained above. For further details, please refer to
formulas for normal cells. Then, it proceeds with the macro layer (lower priority).
macro
Number of subscribers ( X up m ) for each TCH subcell (Txj, TCH) of the macro layer, per user profile up with the mobility m, is
inferred as:
macro macro macro
X up m  Txj TCH  =  S up m  Txj TCH  – S up m – overlapping  Txj TCH    D
macro
S up m  Txj TCH  is the TCH service area of Txj containing the user profile up with the mobility m and D is the profile density.
For each user described in the user profile up with the circuit switched service c and the terminal t, the probability for the user
being connected ( p up  c t  ) is calculated as explained in "Circuit Switched Services" on page 159.
macro
Then, Atoll evaluates the traffic demand, D up  c t  m , in Erlangs in the subcell (Txj, TCH) service area.
macro
macro macro micro S upm – overlapping  Txj TCH 
D up  c t m  Txj TCH  = X up m Txj TCH   p up  c t  + D up  c t m Txi TCH   ------------------------------------------------------------------
micro
 Omax Txi TCH 
S up m  Txi TCH 
For each user described in the user profile up with the packet switched service p and the terminal t, probability for the user
to be connected ( p up  p t  ) is calculated as explained in "Packet Switched Services" on page 159.
macro
Then, Atoll evaluates the traffic demand, D up  p t  m , in kbits/s in the subcell (Txj, TCH) service area.
macro
macro macro micro S upm – overlapping  Txj TCH 
D up  p t m  Txj TCH  = X up m Txj TCH   p up  p t  + D up  p t m Txi TCH   ------------------------------------------------------------------
micro
 Omax Txi TCH 
S up m  Txi TCH 
O max  Txi TCH  is the maximum rate of traffic overflow (stated in %) specified for the TCH subcell of Txi (micro
micro
layer) and S up m  Txi TCH  is the TCH service area of Txi containing the user profile up with the mobility m.
Concentric Cells
Atoll evaluates the traffic demand on the micro layer (higher priority HCS layer) as explained above. For further details, please
refer to formulas given in case of concentric cells. Then, it proceeds with the macro layer (lower priority HCS layer).
The traffic capture is calculated with the option “Best signal level per HCS layer”. It means that there are overlapping areas
between HCS layers traffic is spread according to the layer priority. On these areas, traffic of the higher priority layer may
overflow.
The TCH_INNER service area of the macro layer is overlapped by the micro layer. This area consists of two parts: an area
macro
overlapped by the TCH service area of the micro layer S overlapping –  Txi TCH   Txj,TCH_INNER  and another overlapped by
macro
the TCH_INNER service area of the micro layer S overlapping –  Txi,TCH_INNER   Txj,TCH_INNER  .
Let us consider three areas, S1, S2 and S3.

macro macro
S 1 = S up m  Txj,TCH_INNER  – S up m – overlapping –  Txi TCH   Txj,TCH_INNER 
macro
S 2 = S up m – overlapping –  Txi,TCH_INNER   Txj,TCH_INNER 
macro
S 3 = S up m – overlapping –  Txi TCH   Txj,TCH_INNER  – S 2
163
Figure 3.6: Concentric Cells
macro
S up m  Txj,TCH_INNER  is the TCH_INNER subcell service area of Txj containing the user profile up with the mobility m. We
only consider the overlapping areas containing the user profile up with the mobility m.
macro
On S1, the number of subscribers per user profile up with a given mobility m ( X up m ) is inferred:
macro
X up m  Txj,TCH_INNER  = S 1  D
D is the user profile density.

The traffic spread over the TCH_INNER service area of the micro layer may overflow on the TCH subcell. The traffic overflowing
to the TCH subcell is located on the TCH_INNER service area. On S2, the TCH subcell traffic coming from the TCH_INNER subcell
traffic overflow may overflow proportional to R2.
S2
R 2 = -------------------------------------------------------
-
micro
S up m  Txi,TCH_INNER 
The traffic spread over the ring served by the TCH subcell of the micro layer only may overflow on S3 proportional to R3.
S3
R 3 = --------------------------------------------------------------------------------------------------
-
micro micro
S up m  Txi,TCH  – S up m  Txi,TCH_INNER 
micro micro
S up m  Txi,TCH  and S up m  Txi,TCH_INNER  are the TCH and TCH_INNER service areas of Txi respectively containing the
user profile up with the mobility m.
For each user described in the user profile up with a circuit switched service c and a terminal t, the probability for the user
being connected ( p up  c t  ) is calculated as explained in "Circuit Switched Services" on page 159. Then, Atoll evaluates the
macro
traffic demand, D up  c t  m , in Erlangs in the subcell (Txj, TCH_INNER) service area.
macro
X up m  Txj,TCH_INNER   p up  c t  +
macro
D up  c t  m  Txj,TCH_INNER  = R  D micro  Txi,TCH_INNER   O  Txi,TCH_INNER   O  Txi,TCH  +
2 up  c t  m max max
micro
R 3  X up m  Txi TCH   p up  c t   O max  Txi TCH 
For each user described in the user profile up with a packet switched service p and a terminal t, probability for the user to be
connected ( p up  p t  ) is calculated as explained in "Packet Switched Services" on page 159.
macro
Then, Atoll evaluates the traffic demand, D up  p t  m , stated in kbits/s in the subcell (Txj, TCH_INNER) service area.
macro
X up m  Txj,TCH_INNER   p up  p t  +
macro
D up  p t  m  Txj,TCH_INNER  = R  D micro
2 up  p t  m  Txi,TCH_INNER   O max  Txi,TCH_INNER   O max  Txi,TCH  +
micro
R 3  X up m  Txi TCH   p up  p t   O max  Txi TCH 
164
O max  Txi TCH  and O max  Txi,TCH_INNER  are the maximum rates of traffic overflow (stated in %) specified for the TCH
and TCH_INNER subcells of Txi respectively.
The area of the TCH ring of the macro layer is overlapped by the micro layer. There are two parts: an area overlapped by the
macro
TCH service area of the micro layer S overlapping –  Txi TCH   Txj,TCH -- TCH_INNER  and another one by the TCH_INNER
macro
service area of the micro layer S overlapping –  Txi,TCH_INNER   Txj,TCH -- TCH_INNER  .
Let us consider three areas, S’1, S’2 and S’3.

macro macro macro
S' 1 = S up m  Txj,TCH  – S up m  Txj,TCH_INNER  – S up m – overlapping –  Txi TCH   Txj,TCH -- TCH_INNER 
macro
S' 2 = S up m – overlapping –  Txi,TCH_INNER   Txj,TCH -- TCH_INNER 
macro
S' 3 = S up m – overlapping –  Txi TCH   Txj,TCH -- TCH_INNER  – S' 2
macro macro
S up m  Txj,TCH  and S up m  Txj,TCH_INNER  are the TCH and TCH_INNER subcell service areas of Txj respectively. We only
consider the overlapping areas containing the user profile up with the mobility m.
macro
On S’1, the number of subscribers per user profile up with a given mobility m ( X up m ) is inferred:
macro
X up m  Txj,TCH  = S' 1  D
D is the user profile density.

The traffic spread over the TCH_INNER service area of the micro layer may overflow on the TCH subcell. The traffic overflowing
on the TCH subcell is located on the TCH_INNER service area. On S’2, the TCH subcell traffic coming from the TCH_INNER
subcell traffic overflow may overflow proportionally to R’2.
S' 2
R' 2 = -------------------------------------------------------
-
micro
S up m  Txi,TCH_INNER 
The traffic spread over the ring served by the TCH subcell of the micro layer only may overflow on S’3 proportional to R’3.
S' 3
R' 3 = --------------------------------------------------------------------------------------------------
-
micro micro
S up m  Txi,TCH  – S up m  Txi,TCH_INNER 
micro micro
S up m  Txi,TCH  and S up m  Txi,TCH_INNER  are the TCH and TCH_INNER service areas of Txi respectively containing the
user profile up with the mobility m.
For each user described in the user profile up with a circuit switched service c and a terminal t, the probability for the user
being connected ( p up  c t  ) is calculated as explained in "Circuit Switched Services" on page 159.
macro
Then, Atoll evaluates the traffic demand, D up  c t  m , in Erlangs in the subcell (Txj, TCH) service area.
macro
X up m  Txj TCH   p up  c t  +
macro
macro D up  c t  m  Txj,TCH_INNER   O max  Txj,TCH_INNER  +
D up  c t  m  Txj TCH  =
micro
R' 2  D up  c t  m  Txi,TCH_INNER   O max  Txi,TCH_INNER   O max  Txi,TCH  +
micro
R' 3  X up m  Txi TCH   p up  c t  m  O max  Txi TCH 
For each user described in the user profile up with a packet switched service p and a terminal t, the probability for the user
being connected ( p up  p t  ) is calculated as explained in "Packet Switched Services" on page 159.
macro
Then, Atoll evaluates the traffic demand, D up  p t  m , in kbits/s in the subcell (Txj, TCH) service area.
macro
X up m  Txj TCH   p up  p t  +
macro
macro D up  p t  m  Txj,TCH_INNER   O max  Txj,TCH_INNER  +
D up  p t  m  Txj TCH  =
micro
R' 2  D up  p t  m  Txi,TCH_INNER   O max  Txi,TCH_INNER   O max  Txi,TCH  +
micro
R' 3  X up m  Txi TCH   p up  p t  m  O max  Txi TCH 
165
O max  Txi,TCH  is the maximum rate of traffic overflow (stated in %) specified for the TCH subcell of Txi (micro layer),
O max  Txi,TCH_INNER  the maximum rate of traffic overflow indicated for the TCH_INNER subcell of Txi (macro layer),
O max  Txj,TCH_INNER  the maximum rate of traffic overflow indicated for the TCH_INNER subcell of Txj (macro layer) and
micro
X up m  Txi TCH  the number of subscribers with the user profile up and mobility m on the TCH service area of Txi (as
explained in "Concentric Cells" on page 160).
3.6.2.2 Sector Traffic Maps

We assume that the traffic map is built from a coverage by transmitter prediction calculated for the TCH subcells with options:
• “HCS Servers” and no margin if the network only consists of normal cells and concentric cells,
• “Highest Priority HCS Server” and no margin in case of HCS layers.
When creating the traffic map, you have to specify the traffic demand per transmitter and per service (throughput for a max
rate packet switched service and Erlangs for a circuit switched or constant bit rate packet switched service) and the global
distribution of terminals and mobility types.
Let E c  Txi TCH  denote the Erlangs for the circuit switched service, c, on the TCH subcell of Txi.
Let T p  Txi TCH  denote the throughput of the packet switched service (Max Bit Rate), p, on the TCH subcell of Txi.
Let E p  Txi TCH  denote the Erlangs for the packet switched service (Constant Bit Rate), p, on the TCH subcell of Txi.
We assume that 100% of users have the terminal, t, and the mobility type, m.
3.6.2.2.1 Normal Cells (Nonconcentric, No HCS Layer)

For each circuit switched service, c, Atoll evaluates the traffic demand, Dc,t,m, in Erlangs in the subcell (Txi, TCH) service area.
D c t m  Txi TCH  = E c  Txi TCH 
For each packet switched service (Max Bit Rate), p, Atoll evaluates the traffic demand, Dp,t,m, in kbits/s in the subcell (Txi, TCH)
service area.
D p t m  Txi TCH  = T p  Txi TCH 
For each packet switched service (Constant Bit Rate), p, Atoll evaluates the traffic demand, Dp,t,m, in kbits/s in the subcell (Txi,
TCH) service area.
D p t m  Txi TCH  = E p  Txi TCH   TP p GBR
TP p GBR is the guaranteed bit rate of the constant bit rate packet switched service p.
3.6.2.2.2 Concentric Cells

In case of concentric cells, Atoll distributes a part of traffic on the TCH_INNER service area (TCH_INNER is the highest priority
traffic carrier) and the remaining traffic, on the ring served by the TCH subcell only. The traffic spread over the TCH_INNER
subcell may overflow to the TCH subcell. In this case, the traffic demand is the same on the TCH_INNER subcell and rises on
the TCH subcell.
Traffic overflowing from the TCH_INNER to the TCH is not uniformly spread over the TCH
service area. It is only located on the TCH_INNER service area.
For each circuit switched service, c, Atoll evaluates the traffic demand, Dc,t,m, in Erlangs in the subcell, (Txi, TCH_INNER) and
(Txi, TCH), service areas.
S  Txi,TCH_INNER 
D c t m  Txi,TCH_INNER  = --------------------------------------------  E c  Txi TCH 
S  Txi TCH 
and
--------------------------------------------------------------------------------
S  Txi,TCH  – S  Txi,TCH_INNER  -
 E c  Txi TCH  +
D c t m  Txi,TCH  = S  Txi TCH 
D c t m  Txi,TCH_INNER   O max  Txi,TCH_INNER 
For each packet switched service (Max Bit Rate), p, Atoll evaluates the traffic demand, Dp,t,m, in kbits/s in the subcell, (Txi,
TCH_INNER) and (Txi, TCH), service areas.
166
D p t m  Txi,TCH_INNER  = --------------------------------------------  T p  Txi TCH 
and
 S  Txi,TCH  – S  Txi,TCH_INNER  
---------------------------------------------------------------------------------  T p  Txi TCH  +
D p t m  Txi,TCH  = S  Txi TCH 
D p t m  Txi,TCH_INNER   O max  Txi,TCH_INNER 
O max  Txi,TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell,
S  Txi,TCH  and S  Txi,TCH_INNER  are the TCH and TCH_INNER service areas of Txi respectively.
For each packet switched service (Constant Bit Rate), p, Atoll evaluates the traffic demand, Dp,t,m, in kbits/s in the subcell, (Txi,
TCH_INNER) and (Txi, TCH), service areas.
D p t m  Txi,TCH_INNER  = --------------------------------------------  E p  Txi TCH   TP p GBR
and
--------------------------------------------------------------------------------
S  Txi,TCH  – S  Txi,TCH_INNER  -
 E p  Txi TCH   TP p GBR +
D p t m  Txi,TCH  = S  Txi TCH 
D p t m  Txi,TCH_INNER   O max  Txi,TCH_INNER 
O max  Txi,TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell,
S  Txi,TCH  and S  Txi,TCH_INNER  are the TCH and TCH_INNER service areas of Txi respectively.
3.6.2.2.3 HCS Layers

We assume we have two HCS layers: the micro layer has a higher priority and the macro layer has a lower one. Txi belongs to
the micro layer and Txj to the macro one. The traffic contained in the input traffic map can be assigned to all the HCS layers.
Normal Cells
Atoll distributes traffic on the TCH service areas. The traffic capture is calculated with the option “HCS Servers”. It means that
macro
there is an overlapping area between HCS layers. Let S overlapping  Txj TCH  denote the TCH service area of the macro layer
overlapped by the TCH service area of the micro layer. Traffic on the overlapping area is distributed to the TCH subcell of the
micro layer (higher priority layer). On this area, traffic of the micro layer may overflow to the macro layer. In this case, the
traffic demand is the same on the TCH subcell of the micro layer but rises on the TCH subcell of the macro layer.
Traffic overflowing on the macro layer is not uniformly spread over the TCH service area
of Txj. It is only located on the overlapping area.
Atoll starts evaluating the traffic demand on the micro layer (highest priority HCS layer).
micro
For each circuit switched service, c, Atoll calculates the traffic demand, D c t m , in Erlangs in the subcell (Txi, TCH) service
area.
micro
D c t m  Txi TCH  = E c  Txi TCH 
micro
For each packet switched service (Max Bit Rate), p, Atoll calculates the traffic demand, D p t m , in kbits/s in the subcell (Txi,
TCH) service area.
micro
D p t m  Txi TCH  = T p  Txi TCH 
micro
For each packet switched service (Constant Bit Rate), p, Atoll calculates the traffic demand, D p t m , in kbits/s in the subcell
(Txi, TCH) service area.
micro
D p t m  Txi TCH  = E p  Txi TCH   TP p GBR
Then, Atoll proceeds with the macro layer (lower priority HCS layer).
macro
For each circuit switched service, c, Atoll calculates the traffic demand, D c t m , in Erlangs in the subcell (Txj, TCH) service area.
167
macro
macro micro S overlapping  Txj TCH 
D c t m  Txj TCH  = E c  Txj TCH  + D c t m  Txi TCH   ----------------------------------------------------
micro
-  O max  Txi TCH 
macro
For each packet switched service (Max Bit Rate), p, Atoll calculates the traffic demand, D p t m , in kbits/s in the subcell (Txj,
TCH) service area.
macro
D p t m  Txj TCH  = T p  Txj TCH  + D p t m  Txi TCH   ----------------------------------------------------
micro
O max  Txi TCH  is the maximum rate of traffic overflow (in %) specified for the TCH subcell of Txi (micro cell) and
micro
S  Txi TCH  the TCH service area of Txi.
macro
(Txj, TCH) service area.
macro
D p t m  Txj TCH  = E p  Txi TCH   TP p GBR + D p t m  Txi TCH   ----------------------------------------------------
micro
O max  Txi TCH  is the maximum rate of traffic overflow (in %) specified for the TCH subcell of Txi (micro cell) and
micro
S  Txi TCH  the TCH service area of Txi.
You can restrict the traffic assignement of each traffic map to a specific HCS layer in the
running options of the traffic capture. If you do so, no overflow occurs between HCS
layers and the only overflow which is considered occurs within concentric cells (See
"Concentric Cells" on page 160).
Concentric Cells
Atoll evaluates the traffic demand on the micro layer as explained above in case of concentric cells and then proceeds with
the macro layer (lower priority layer).
The traffic capture is calculated with the option “HCS Servers”. It means that there is overlapping areas between HCS layers
traffic is spread over according to the layer priority. On these areas, traffic of the higher priority layer may overflow.
Figure 3.7: Concentric Cells
The TCH_INNER service area of the macro layer is overlapped by the micro layer. This area consists of two parts: an area
macro
overlapped by the TCH service area of the micro layer S overlapping –  Txi TCH   Txj,TCH_INNER  and another overlapped by
macro
the TCH_INNER service area of the micro layer S overlapping –  Txi,TCH_INNER   Txj,TCH_INNER  .
Let us consider three areas, S1, S2 and S3.
168
macro macro
S1 = S  Txj,TCH_INNER  – S overlapping –  Txi TCH   Txj,TCH_INNER 
macro
S 2 = S overlapping –  Txi,TCH_INNER   Txj,TCH_INNER 
macro
S 3 = S overlapping –  Txi TCH   Txj,TCH_INNER  – S 2
macro
S  Txj,TCH_INNER  is the TCH_INNER subcell service area of Txj.
The traffic specified for Txj in the map description ( E c  Txj TCH  ) is spread over S1 proportionally to R1.
S1
R 1 = -------------------------------------
map
S  Txj TCH 
map
S  Txj TCH  is the TCH service area of Txj in the traffic map with the option “Best signal level of the highest priority
layer”.
The traffic spread over the TCH_INNER service area of the micro layer may overflow to the TCH subcell. The traffic overflowing
to the TCH subcell is located on the TCH_INNER service area. On S2, the TCH subcell traffic coming from the TCH_INNER subcell
traffic overflow may overflow proportional to R2.
S2
R 2 = -------------------------------------------------------
-
micro
The traffic spread over the ring only served by the TCH subcell of the micro layer may overflow on S3 proportional to R3.
S3
R 3 = --------------------------------------------------------------------------------------------------
-
micro micro
S  Txi,TCH  – S  Txi,TCH_INNER 
macro
For each circuit switched service, c, Atoll calculates the traffic demand, D c t m , in Erlangs in the subcell (Txj, TCH_INNER)
service area.
R 1  E c  Txj TCH  +
micro
macro R 2  D c t m  Txi,TCH_INNER   O max  Txi,TCH_INNER   O max  Txi TCH  +
D c t m  Txj,TCH_INNER  =
micro micro
S  Txi TCH  – S  Txi,TCH_INNER  
R 3  ----------------------------------------------------------------------------------------------------------
micro
-  E c  Txi TCH   O max  Txi TCH 
macro
For each packet switched service (Max Bit Rate), p, Atoll calculates the traffic demand, D p t m , in kbits/s in the subcell (Txj,
TCH_INNER) service area.
R 1  T p  Txj TCH  +
micro
macro R 2  D p t m  Txi,TCH_INNER   O max  Txi,TCH_INNER   O max  Txi TCH  +
D p t m  Txj,TCH_INNER  =
micro micro
S  Txi TCH  – S  Txi,TCH_INNER  -
R 3  ----------------------------------------------------------------------------------------------------------
micro
 T p  Txi TCH   O max  Txi TCH 
O max  Txi TCH  is the maximum rate of traffic overflow (stated in %) specified for the TCH subcell of Txi,
O max  Txi,TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txi and
micro
S  Txi TCH  is the TCH subcell service area of Txi.
macro
(Txj, TCH_INNER) service area.
R 1  E p  Txi TCH   TP p GBR +

micro
R 2  D p t m  Txi,TCH_INNER   O max  Txi,TCH_INNER   O max  Txi TCH  +
macro  micro 
D p t m  Txj,TCH_INNER  = S  Txi TCH  – S
micro
 Txi,TCH_INNER  -
 ---------------------------------------------------------------------------------------------------------- 
 S
micro
 Txi TCH  
R3   
  
 
 E p  Txi  TCH   TP p GBR  O max  Txi TCH  
169
O max  Txi,TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txi and
micro
S  Txi TCH  is the TCH subcell service area of Txi.
The area of the TCH ring of the macro layer is overlapped by the micro layer. There are two parts: an area overlapped by the
macro
TCH service area of the micro layer S overlapping –  Txi TCH   Txj,TCH -- TCH_INNER  and another overlapped by the
macro
TCH_INNER service area of the micro layer S overlapping –  Txi,TCH_INNER   Txj,TCH -- TCH_INNER  .
Let us consider three areas, S’1, S’2 and S’3.

macro macro macro
S' 1 = S  Txj TCH  – S  Txj,TCH_INNER  – S overlapping –  Txi TCH   Txj,TCH -- TCH_INNER 
macro
S' 2 = S overlapping –  Txi,TCH_INNER   Txj,TCH -- TCH_INNER 
macro
S' 3 = S overlapping –  Txi TCH   Txj,TCH -- TCH_INNER  – S' 2
macro macro
S  Txj TCH  and S  Txj,TCH_INNER  are the TCH and TCH_INNER subcell service areas of Txj respectively.
The traffic specified for Txj in the map description ( E c  Txj TCH  ) is spread over S’1 proportional to R’1.
S' 1
R' 1 = -------------------------------------
map
S  Txj TCH 
map
S  Txj TCH  is the TCH service area of Txj in the traffic map with the option “Best signal level of the highest priority
layer”.
The traffic spread over the TCH_INNER service area of the micro layer may overflow to the TCH subcell. The traffic overflowing
to the TCH subcell is located on the TCH_INNER service area. On S’2, the TCH subcell traffic coming from the TCH_INNER
subcell traffic overflow may overflow proportional to R’2.
S' 2
R' 2 = -------------------------------------------------------
-
micro
The traffic spread over the ring only served by the TCH subcell of the micro layer may overflow on S’3 proportional to R’3.
S' 3
R' 3 = --------------------------------------------------------------------------------------------------
-
micro micro
S  Txi,TCH  – S  Txi,TCH_INNER 
macro
For each circuit switched service, c, Atoll calculates the traffic demand, D c t m , in Erlangs in the subcell (Txj, TCH) service area.
R' 1  E c  Txj TCH  +

macro
D c t m  Txj,TCH_INNER   O max  Txj,TCH_INNER  +
macro
D c t m  Txj TCH  = micro
R' 2  D c t m  Txi,TCH_INNER   O max  Txi,TCH_INNER   O max  Txi TCH  +
micro micro
S  Txi,TCH  – S  Txi,TCH_INNER  
R' 3  -------------------------------------------------------------------------------------------------------
micro
-  E c  Txi TCH   O max  Txi TCH 
S  Txi,TCH 
macro
For each packet switched service (Max Bit Rate), p,Atoll calculates the traffic demand, D p t m , in kbits/s in the subcell (Txj,
TCH) service area.
R' 1  T p  Txj TCH  +

macro
macro
D p t m  Txj TCH  = micro
R' 2  D p t m  Txi,TCH_INNER   O max  Txi,TCH_INNER   O max  Txi TCH  +
micro micro
R' 3  -------------------------------------------------------------------------------------------------------
micro
-  T p  Txi TCH   O max  Txi TCH 
S  Txi,TCH 
170
O max  Txj,TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txj,
O max  Txi,TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txi,
micro micro
S  Txi,TCH  is the TCH subcell service area of Txi and S  Txi,TCH_INNER  is the TCH_INNER subcell service area of
Txi.
macro
For each packet switched service (Constant Bit Rate), p,Atoll calculates the traffic demand, D p t m , in kbits/s in the subcell
(Txj, TCH) service area.
R' 1  E p  Txi TCH   TP p GBR +

macro
micro
macro
R' 2  D p t m  Txi,TCH_INNER   O max  Txi,TCH_INNER   O max  Txi TCH  +
D p t m  Txj TCH  =
 micro micro 
 ------------------------------------------------------------------------------------------------------- 
-
 S
micro
 Txi,TCH  
R' 3   
  
 
 E p  Txi TCH   TP p GBR  O max  Txi TCH  
O max  Txj,TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txj,
O max  Txi,TCH_INNER  is the maximum rate of traffic overflow (stated in %) specified for the TCH_INNER subcell of Txi,
micro micro
S  Txi,TCH  is the TCH subcell service area of Txi and S  Txi,TCH_INNER  is the TCH_INNER subcell service area of
Txi.
3.7 Network Dimensioning

Atoll is capable of dimensioning a GSM GPRS EDGE network with a mixture of circuit and package switched services. This
section describes the technical details of Atoll’s dimensioning engine.
3.7.1 Dimensioning Models and Quality Graphs

In Atoll, a dimensioning model is an entity utilized by the dimensioning engine along with other inputs (traffic, limitations,
criteria, etc.) in the process of dimensioning. A dimensioning model defines the QoS KPIs to be taken into account when
dimensioning a network for both circuit and packet switched traffic. The user can define either to use Erlang B or Erlang C
queuing model for circuit switched traffic and can define which KPIs to consider when dimensioning the network for packet
switched traffic. The dimensioning engine will only utilize the quality curves of the KPI selected. The KPIs not selected are
supposed to be either already satisfactory or not relatively important.
3.7.1.1 Circuit Switched Traffic

The network dimensioning for circuit switched traffic is performed using the universally accepted and adopted Erlang B and
Erlang C formulas. The dimensioning criterion in these formulas is the Grade of Service or the allowed blocking probability of
the circuit switched traffic.
In the Erlang B approach, this Grade of Service is defined as the percentage of incoming circuit switched calls that are blocked
due to lack of resources or timeslots. This formula implies a loss system. The blocked calls are supposed to be lost and the
caller has to reinitiate it.
In the Erlang C approach, the Grade of Service is the percentage of incoming calls that are placed in a waiting queue when
there are no resources available, until some resources or timeslots are liberated. This queuing system has no lost calls. As the
load on the system increases, the average waiting time in the queue also increases.
These formulas and their details are available in many books. For example, Wireless Communications Principles and Practice
by Theodore S. Rappaport, Prentice Hall.
Following the common practice, network dimensioning in Atoll is based on the principle that a voice or GSM call has priority
over data transmission. Therefore, as explained later in the network dimensioning steps, Atoll first performs network
dimensioning according to the circuit switched traffic present in the subcell in order to ensure the higher priority service
availability before performing the same for the packet switched traffic.
171
3.7.1.2 Packet Switched Traffic

Since packet switched traffic does not occupy an entire timeslot the whole time, it is much more complicated to study than
circuit switched traffic. Packet traffic is intermittent and bursty. Whenever there is packet data to be transferred, a Temporary
Block Flow (TBF) is initiated for transferring these packets. Multiple TBFs can be multiplexed on the same timeslot. This implies
that there can be many packet switched service users that have the same timeslots assigned for packet data transfer but at
different intervals of time.
This multiplexing of a number of packet switched service users over the same timeslots incurs a certain reduction in the
throughput (data transfer rate) for each multiplexed user. This reduction in the throughput is more perceivable when the
system traffic load is high. The following parts describe the three most important Key Performance Indicators in GPRS/EDGE
networks and how they are modelled in Atoll.
3.7.1.2.1 Throughput
Throughput is defined as the amount of data delivered to the Logical Link Control Layer in a given unit of time. Each temporary
block flow (TBF), and hence each user, has an associated measured throughput sample in a given network. Each network will
have a different throughput probability distribution depending on the load and network configuration. Instead of using the
precise probability distributions, it is more practical to compute the average and percentile throughput values.
In GPRS, the resources are shared between the users being served, and consequently, the throughput is reduced as the
number of active users increases. This reduction in user perceived throughput is modelled through a reduction factor. The
throughput experienced by a user accessing a particular service can be calculated as:
User throughput = Number of allocated timeslots x Timeslot capacity x Reduction Factor
Or
User throughput per allocated timeslot = Timeslot capacity x Reduction Factor
Timeslot Capacity
The timeslot capacity is the average throughput per fully utilized timeslot. It represents the average throughput from the
network point of view. It mainly depends on the network’s propagation conditions and criteria in the coverage area of a
transmitter (carrier power, carrier-to-interference distribution, etc.). It is a measure of how much data the network is able to
transfer with 1 data Erlang, or in other words, how efficiently the hardware resources are being utilized by the network. It may
also depend on the RLC protocol efficiency.
Atoll computes the average timeslot capacity during the traffic analysis and is used to determine the minimum throughput
reduction factor. But since this information is displayed in the network dimensioning results (only due to relevance), this
information has been considered as a part of the network dimensioning process in this document.
Timeslot Utilisation
Timeslot utilization takes into account the average number of timeslots that are available for packet switched traffic. It is a
measure of how much the network is loaded with data services. Networks with timeslot utilisation close to 100% are close to
saturation and the end-user performance is likely to be very poor.
In Atoll this parameter is termed as the Load (Traffic load for circuit switched traffic and packet switched traffic load for packet
switched traffic). It is described in more detail in the Network dimensioning steps section.
Reduction Factor
Reduction factor takes into account the user throughput reduction due to timeslot sharing among many users. The figure
below shows how the peak throughput available per timeslot is reduced by interference and sharing.Reduction factor is a
function of the number of timeslots assigned to a user (Nu), number of timeslots available in the system (Ns) and the average
system packet switched traffic load (Lp) (utilization of resources in the system). Data Erlangs or data traffic is given by:
Data Erlangs = L P  N S
172
Figure 3.8: Reduction of Throughput per Timeslot
More precisely, the reduction factor is a function of the ratio Ns/Nu (Np). Np models the equivalent timeslots that are available
for the packet switched traffic in the system. For example, a 24-timeslot system with each user assigned 3 timeslots per
connection can be modelled by a single timeslot connection system with 8 timeslots in total.
The formula for reduction factor can be derived following the same hypotheses followed by Erlang in the derivation of the
blocking probability formulas (Erlang B and Erlang C).
Let X be a random variable that measures the reduction factor in a certain system state:
0 if n = 0
1 if 0 < n  NP
X
N
-----P- if n > NP
n
n is the instantaneous number of connections in the system. The throughput reduction factor is defined as:

P X= n
RF   X  --------------------
P X= 0
-
n=0
Or,

PX= n
RF =  X  ---------------------------

-
 P X= i
n=0
i 0
Here, P(X=n) is the probability function of having n connections in the system. Under the same assumptions as those of the
Erlang formulas, the probability function can be written as:
n
 LP  NP 
-----------------------
n!
P  X = n  = ---------------------------------------------------------------------------------------
N
if 0  n  N P
P 
i i
 LP  NP   LP  NP 
 ---------------------
i!
+  ------------------------------
 i – NP 
i=0 i=N + 1 N P!  N P
P
n
 LP  NP 
------------------------------
i – N 
P
N P!  N P
P  X = n  = ---------------------------------------------------------------------------------------
N
if n > N P
P 
i i
 LP  NP   LP  NP 
 ---------------------
i!
+  ------------------------------
 i – NP 
i=0 i = NP + 1 N P!  N P
Hence the reduction factor can finally be written as:
173
NP 
i i
 LP  NP   LP  NP  N
 --------------------- +  ------------------------------   -----P-
i! i – N   i 
P
i=1 i = N + 1 N P!  N P
RF = -------------------------------------------------------------------------------------------------------
N
P -
P 
i i
 LP  NP   LP  NP 
 --------------------- +
i!  ------------------------------
i – N 
P
i=1 i = NP + 1 N P!  N P
This formula is not directly applicable in any software application due to the summations up to infinity. Atoll uses the following
version of this formula that is exactly the same formula without the summation overflow problem.
NP  NP + 1  N
n  P
n
 LP  NP  NP  L

----------------------- – -------------------  ln  1 – L P  +  ----P
n! N P!  n
n = 1  n = 1 
RF = ------------------------------------------------------------------------------------------------------------------------
N
-
P NP
n
 LP  NP   LP  NP  LP
 ----------------------- + --------------------------  --------------
n! N P! 1 – LP
n=1
The default quality curves for the Reduction Factor have been derived using the above formula. Each curve is for a fixed
number of timeslots available for packet switched traffic (Np) describing the reduction factor at different values of packet
switched traffic load (Lp). The figure below contains all the reduction factor quality curves in Atoll. The Maximum reduction
factor can be 1, implying a maximum throughput, and the minimum can be 0, implying a saturated system with no data
throughput.
Figure 3.9: Reduction Factor for Different Packet Switched Traffic Loads (Lp, X-axis)
Each curve in the above figure represents an equivalent number of packet switched timeslots, NP.
3.7.1.2.2 Delay
Delay is the time required for an LLC PDU to be completely transferred from the SGSN to the MS, or vice versa. As the delay
is a function of the delays and the losses incurred at the packet level, the network parameters, such as the packet queue
length, and different protocol properties, such as the size of the LLC PDU, become important. It is also quite dependent upon
the radio access round trip time (RA RTT) and has a considerable impact on the application level performance viewed by the
user.
The delay parameter is a user level parameter rather than being a network level quantity, like throughput per cell, timeslot
capacity, TBF blocking and reduction factor, hence it is difficult to model and is currently under study. Hence, no default curve
is presently available for delay in Atoll.
174
3.7.1.2.3 Blocking Probability

In GPRS, there is no blocking as in circuit switched connections. If a new temporary block flow (TBF) establishment is requested
and there are already M users per timeslot, M being the maximum limit of multiplexing per timeslot (Multiplexing factor), the
request is queued in the system to be established later when resources become available.
Supposing that M number of users can be multiplexed over a single timeslot (PDCH), we can have a maximum of M * Np users
in the system. This implies that if a new TBF is requested when there are already M * Np users active, it will be blocked and
placed in a queue. So the blocking probability is the probability of having M * Np + 1 users in the system or more, meaning,
P X= n for n =  M  N P  + 1
as in this case n is always greater than Np, we have,
n
 LP  NP 
------------------------------
 i – NP 
N P!  N P
P  X = n  = ---------------------------------------------------------------------------------------
N
P 
i i
 LP  NP   LP  NP 
 --------------------- +
i!  ------------------------------
 i – NP 
i=0 i = NP + 1 N P!  N P
So, the Blocking Probability can be given as:


n
 LP  NP 
  -------------------------------
 1 – NP 
-
n = MN +1 P N !  N

P
BP = P  X = n  = ----------------------------------------------------------------------------------------
N
P -
P 
i i
n = MN+1  LP  NP   LP  NP 
 --------------------- +
i!  -------------------------------
 1 – NP 
-
i=0 i = NP + 1 N P!  N P
Eliminating the summations to infinity, the blocking probability can be stated in a simpler form:
M  NP
 LP  NP  LP
-  -------------
----------------------------------------- -
 M  NP – NP  1 – L
N P!  N P P
BP = ----------------------------------------------------------------------------------
N
P NP
i
 LP  NP   LP  NP  LP
 --------------------- + --------------------------  -------------
i! N P! 1 – LP
-
i=0
The above formula has been used to generate the default quality curves for blocking probability in Atoll.
These graphs are generated for a user multiplexing factor of 8 users per timeslot. Each curve represents an equivalent number
of packet switched timeslots, NP.
The curves depict the blocking probabilities for different number of available connections (Np) at different packet switched
traffic loads (Lp) for a fixed user multiplexing factor of 8. The figure below contains all the blocking probability curves for
packet switched traffic dimensioning in Atoll. The blocking probability increases with the packet switched traffic load, which
implies that as the packet switched traffic increases for a given number of timeslots, the system starts to get more and more
loaded, hence there is higher probability of having a temporary block flow placed in a waiting queue.
175
Figure 3.10: Blocking Probability for Different Packet Switched Traffic Loads (Lp, X-axis)
Reference: T. Halonen, J. Romero, J. Melero; GSM, GPRS and EDGE performance –

Evolution towards 3G/UMTS, John Wiley and Sons Ltd.
3.7.2 Network Dimensioning Process

The network dimensioning process is described below in detail. As the whole dimensioning process is in fact a chain of small
processes that have there respective inputs and outputs, with outputs of a preceding one being the inputs to the next, the
best method is to detail each process individually in form of steps of the global dimensioning process.
3.7.2.1 Network Dimensioning Engine

During the dimensioning process, Atoll first computes the number of timeslots required to accommodate the circuit switched
traffic. Then it calculates the number of timeslots to add in order to satisfy the demand of packet switched traffic. This is
performed using the quality curves entered in the dimensioning model used. If the dimensioning model has been indicated to
take all three KPIs in to account (throughput reduction factor, delay and blocking probability), the number of timeslots to be
added is calculated such that:
• The throughput reduction factor is greater than the minimum throughput reduction factor,
• Delay is less than the maximum permissible delay defined in the service properties, and
• The blocking probability is less than the maximum allowable blocking probability defined in the service properties.
The figure below depicts a simplified flowchart of the dimensioning engine in Atoll.
Figure 3.11: Network Dimensioning Process
176
On the whole, following are the inputs and outputs of the network dimensioning process:
3.7.2.1.1 Inputs
• Circuit switched traffic demand
• Packet switched traffic demand
• Timeslot configurations defined for each subcell
• Target traffic overflow rate and Half-rate traffic ratio for each subcell
• Service availability criteria: minimum required throughput per user, maximum permissible delay, maximum allowable
blocking probability etc.
• Dimensioning model parameters: Maximum number of TRXs per transmitter, dimensioning model for circuit switched
traffic, number of minimum dedicated packet switched timeslots per transmitter, maximum number of TRXs added
for packet switched services, KPIs to consider, and their quality curves.
3.7.2.1.2 Outputs
• Number of required TRXs per transmitter
• Number of required shared, circuit switched and packet switched timeslots
• Traffic load
• Served circuit switched traffic
• Served packet switched traffic
• Effective rate of traffic overflow
• Actual KPI values: throughput reduction factor, delay and blocking probability
3.7.2.2 Network Dimensioning Steps

This section describes the entire process step by step as it is actually performed in Atoll. Details of the calculations of the
parameters that are calculated during each step are described as well.
3.7.2.2.1 Step 1: Timeslots Required for CS Traffic

Atoll computes the number of timeslots required to accommodate the circuit switched traffic assigned to each subcell. Atoll
takes the circuit switched traffic demand (Erlangs) either user-defined or calculated in the traffic analysis and assigned to the
current subcell and the maximum blocking probability defined for the circuit switched service, and computes the required
number of timeslots to satisfy this demand using the Erlang B or Erlang C formula (as defined by the user).
If the user-defined target rate of traffic overflow per subcell, OTarget, is greater than the maximum blocking rate defined in the
services properties, it is going to be taken as the Grade of Service required for that subcell instead of the maximum blocking
rate of the service.
For the blocking probability GoS and circuit switched traffic demand TDC, Atoll determines the required number of timeslots
TSreq. C for each subcell using formulas described below. In fact, Atoll searches for TSreq. C value until the defined grade of
service is reached.
For Erlang B, we have:
TSreqC
 TD C 
----------------------------
 TS reqC !
GoS = TS
------------------------------
reqC
k
 TD C 
 ---------------
k!
-
k=0
For Erlang C, we have:

TS
reqC
 TD C 
GoS = ------------------------------------------------------------------------------------------------------------------------------------
TS –1
-
reqC
k
TS reqC TD C   TD C 
 TD C  +  TS reqC !   1 – --------------
-   ----------------
TS reqC k!
k=0
Atoll considers the effect of half-rate circuit switched traffic by taking into account a user-defined percentage of half-rate
traffic. Atoll computes the effective equivalent number of full-rate timeslots that will be required to carry the total traffic with
the defined percentage of half-rate traffic.
If the number of timeslots required to accommodate the full-rate circuit switched traffic is TSreq. FR, and the percentage of
half-rate traffic within the subcell is defined by HR, then the effective number of equivalent full-rate circuit switched timeslots
TSeff. that can carry this traffic mix is calculated by:
TS eff = TS reqFR   1 – HR
-------
 2
177
Atoll employs this simplified approach to integrating half-rate circuit switched traffic, which provides approximately the same
results as obtained by using the half-rate traffic charts.
3.7.2.2.2 Step 2: TRXs Required for CS Traffic and Dedicated PS Timeslots

This stage of the network dimensioning process computes the number of TRXs required to carry the circuit switched traffic
demand through the number of required timeslots calculated above and the timeslot configuration defined by the user in the
network settings. Atoll distributes the number of required circuit switched timeslots calculated in Step 1 taking into account
the presence of dedicated packet switched timeslots in each TRX according to the timeslot configurations.
If a timeslot configuration defines a certain number of dedicated packet switched timeslots pre-allocated in certain TRXs,
those timeslots will not be considered capable of carrying circuit switched traffic and hence will not be allocated. For example,
if 4 timeslots have been marked as packet switched timeslots in the first TRX and Atoll computes 8 timeslots for carrying a
certain circuit switched traffic demand, then the number of TRXs to be allocated cannot be 1 even if there is no packet
switched traffic considered yet.
The total numbers of timeslots that carry circuit switched and packet switched traffic respectively are the sums of respective
dedicated and shared timeslots:
TS P = TS S + TS P dedicated and TS C = TS S + TS C dedicated
3.7.2.2.3 Step 3: Effective CS Blocking, Effective CS Traffic Overflow and Served CS Traffic
In this step, the previously calculated number of required TRXs is used to compute the effective blocking rate for the circuit
switched traffic. This is performed by using the Erlang B or Erlang C formula with the circuit switched traffic demand and the
number of required TRXs as inputs and computing the Grade of Service (or blocking probability). It then calculates the
effective traffic overflow rate, Oeff..
In case of Erlang B formula, the effective rate of traffic overflow for the circuit switched traffic is the same as the circuit
switched blocking rate. While in case of the Erlang C model, the circuit switched traffic is supposed to be placed in an infinite-
length waiting queue. This implies that there is no overflow in this case.
From this data, it also computes the served circuit switched traffic. This is the difference of the circuit switched traffic demand
and the percentage of traffic that overflows from the subcell to other subcells calculated above. Hence, for an effective traffic
overflow rate of Oeff. and the circuit switched traffic demand of TDC, the served circuit switched traffic STC is computed as:
ST C = TD C   1 – O eff 
3.7.2.2.4 Step 4: TRXs to Add for PS Traffic

This step is the core of the dimensioning process for packet switched services. First of all, Atoll computes the number of TRXs
to be added to carry the packet switched traffic demand. This is the number of TRXs that contain dedicated packet switched
and shared timeslots.
To determine this number of TRXs, Atoll calculates the equivalent average packet switched traffic demand in timeslots by
studying each pixel covered by the transmitter. This calculation is in fact performed in the traffic analysis process or is user-
defined in the subcells table. Knowing the traffic demand per pixel of the covered area in terms of kbps and the maximum
attainable throughput per pixel (according to the C and/or C/I conditions and the coding scheme curves in the GPRS/EDGE
configuration), Atoll calculates the average traffic demand in packet switched timeslots by:
Traffic demand per pixel (kbps)-

TD P
Timeslots
=  --------------------------------------------------------------------------
Throughput per pixel (kbps)
pixel
The average timeslot capacity of a transmitter is calculated by dividing the packet switched traffic demand over the entire
coverage area (in kbps) by the packet switched traffic demand in timeslots calculated above.
With the number of timeslots required to serve the circuit switched traffic, the timeslots required for packet switched traffic
and their respective distributions according to the timeslot configurations being known, Atoll calculates the number of
timeslots available for carrying the packet switched traffic demand. These timeslots can be dedicated packet switched
timeslots and the shared ones. So, following the principle that shared timeslots are potential carriers of both traffic types,
TS P = TS S + TS P dedicated
TS C = TS S + TS C dedicated
The packet switched traffic load is calculated by the formula:
 ST C – TS C dedicated + TD P 
Timeslots
L P = ------------------------------------------------------------------------------------
-
TS P
The second important parameter for the calculation of Reduction Factor, Delay and Blocking Probability is the equivalent
number of available timeslots for packet switched traffic, i.e. NP. This is computed by dividing the total number of timeslots
178
available for carrying packet switched traffic by the number of downlink timeslots defined in the mobile terminal properties.
So, NP is calculated at this stage as:
TS P
N P = -----------------------
-
TS Terminal
, TSTerminal is the number of timeslots that a terminal will use in packet switched calls.
The number timeslots that a terminal can use in packet switched calls is the product of the number of available DL timeslots
for packet-switched services (on a frame) and the number of simultaneous carriers (in case of EDGE evolution).
The number of timeslots that a terminal will use in packet switched calls is determined by taking the lower of the maximum
number of timeslots on a carrier for packet switched service defined in the service properties and the maximum number of
timeslots that a mobile terminal can use for packet switched services (see above) on acarrier.
TS Terminal = min  TS Max Service TS Max TerminalType 
and TS Max TerminalType = TS DL TerminalType  Carriers DL TerminalType
Here, the min(X,Y) function yields the lower value among X and Y as result.
Now, knowing the packet switched traffic load, LP, and the equivalent number of available timeslots, NP, Atoll finds out the
KPIs that have been selected before launching the dimensioning process using the quality curves stored in the dimensioning
model.
This particular part of this step can be iterative if the KPIs to consider in dimensioning are not satisfied in the first try. If the
KPIs calculated above are within acceptable limits as defined by the user, it means that the dimensioning process has
acceptable results. If these KPIs are not satisfied, then Atoll increases the number of TRXs calculated for carrying packet
switched traffic by 1 (each increment adding 8 more timeslots for carrying packet switched traffic as the least unit that can be
physically added or removed is a TRX) and resumes the computations from Step 3. It then recalculates the packet switched
traffic load, LP, and the equivalent number of available timeslots, NP. Then it recomputes the KPIs with these new values of LP
and NP. If the KPIs are within satisfactory limits the results are considered to be acceptable. Otherwise, Atoll performs another
iteration to find the best possible results.
The calculated values of all the KPIs are compared with the ones defined in the service properties. The values for maximum
Delay and Blocking probability are defined directly in the properties but the minimum throughput reduction factor is
calculated by Atoll using the user’s inputs: minimum throughput per user and required availability. This calculation is in fact
performed during the traffic analysis process, but since it is relevant to the dimensioning procedure, it is displayed in a column
in the dimensioning results so that the user can easily compare the minimum requirement on the reduction factor KPI with
the resulting one. If dimensioning is not based on a traffic analysis, the minimum throughput reduction factor is a user-defined
parameter.
Minimum Throughput Reduction Factor Calculation
The minimum throughput reduction factor is computed using the input data: minimum required throughput per user defined
in the service properties, the average throughput per timeslot deduced from the throughput curves stored in the GPRS/EDGE
configuration properties for each coding scheme, the total number of downlink timeslots defined in the properties of the
mobile terminal (See TS Max TerminalType defintion above) and the required availability defined in the service properties.
It is at the stage of calculating the average timeslot capacity per transmitter that Atoll studies each covered pixel for carrier
power or carrier-to-interference ratio. According to the measured carrier power or carrier-to-interference ratio, Atoll deduces
the maximum throughput available on that pixel through the throughput vs. C or throughput vs. C/I curves of the GPRS/EDGE
configuration.
The throughput per timeslot per pixel TPTS, Pixel can be either a function of carrier power C, or carrier power C and the carrier-
to-interference ratio C/I, depending on the user-defined traffic analysis RF conditions criteria. Therefore,
TP TS Pixel = f  C 
Or
TP TS Pixel = f  C  and TP TS Pixel = f  ---

C
 i
The required availability parameter defines the percentage of pixels within the coverage area of the transmitter that must
satisfy the minimum throughput condition. This parameter renders user-manageable flexibility to the throughput
requirement constraint.
To calculate the minimum throughput reduction factor for the transmitter, Atoll computes the minimum throughput
reduction factor for each pixel using the formula:
179
TP user min
RF min Pixel = ----------------------------------------------------
-
TP TS Pixel  TS Terminal
Once the minimum reduction factor for each pixel is known, Atoll calculates the global minimum reduction factor that is
satisfied by the percentage of covered pixels defined in the required availability. The following example may help in
understanding the concept and calculation method.
Example: Let the total number of pixels, covered by a subcell S, be 1050. The reliability level set to 90%. This implies that the
required minimum throughput for the given service will be available at 90% of the pixels covered. This, in turn, implies that
there will be a certain limit on the reduction factor, i.e. if the actual reduction factor in that subcell becomes less than a
minimum required, the service will not be satisfactory.
Atoll computes the minimum reduction factor at each pixel using the formula mentioned above, and outputs the following
results:
RFmin Number of pixels
0.3 189
0.36 57
0.5 20
0.6 200
0.72 473
0.9 23
0.98 87
So for a reliability level of 90%, the corresponding RFmin will be the one provided at least 90% of the pixels covered, i.e. 945
pixels. The corresponding value of the resulting RFmin in this example hence turns out to be 0.9, since this value covers 962
pixels in total. Only 87 of the covered pixels imply an RFmin of 0.98. These will be the pixels that do not provide satisfactory
service.
This calculation is performed for each service type available in the subcell coverage area. The final minimum throughput
reduction factor is the highest one amongst all calculated for each service separately.
The minimum throughput reduction factor RFmin value is a minimum requirement that must be fulfilled by the network
dimensioning process when the Reduction Factor KPI is selected in the dimensioning model.
Figure 3.12: Minimum Throughput Reduction Factor
3.7.2.2.5 Step 5: Served PS Traffic

Atoll calculates the served packet switched traffic using the number of timeslots available to carry the packet switched traffic
demand. As the result of the above iterative step, Atoll always finds the best possible answer in terms of number of timeslots
required to carry the packet switched traffic demand unless the requirement exceeds the maximum limit on the number of
the packet switched traffic timeslots defined in the dimensioning model properties. Hence, there is no packet traffic overflow
unless the packet switched traffic demand requires more TRXs than the maximum allowed
3.7.2.2.6 Step 6: Total Traffic Load

This step calculates the final result of the dimensioning process, i.e. the total traffic load. The total traffic load L is calculated as:
180
ST C + ST P
L = ---------------------------------------------------------------------------------
-
TS C dedicated + TS P dedicated + TS S
,
• STC is the served circuit switched traffic
• STP is the served packet switched traffic
• TSC, dedicated is the number of dedicated circuit switched timeslots
• TSP, dedicated is the number of dedicated packet switched timeslots
• TSS is the number of shared timeslots
3.8 Key Performance Indicators Calculation

This feature calculates the current values for all circuit switched and packet switched Key Performance Indicators as a measure
of the current performance of the network. It can be used to evaluate an already dimensioned network in which recent traffic
changes have been made in limited regions to infer the possible problematic areas and then to improve the network
dimensioning with respect to these changes.
The concept of this computation is the inverse of that of the dimensioning process. In this case, Atoll has the results of the
dimensioning process already committed and known. Atoll then computes the current values for all the KPIs knowing the
number of required TRXs, the respective numbers of shared and dedicated timeslots and the circuit switched and packet
switched traffic demands.
The computation algorithm utilizes the parameters set in the dimensioning model properties and the quality curves for the
throughput reduction factor, delay and the blocking probability.
The following conventional relations apply:
If,
• TSC, dedicated is the number of timeslots dedicated to the circuit switched traffic,
• TSP, dedicated is the number of timeslots dedicated to the packet switched traffic,
• TSS is the number of shared timeslots for a transmitter,
Then, the number of timeslots available for the circuit switched traffic, TSC, is defined as:
TS C = TS S + TS C dedicated
And the number of timeslots available for the packet switched traffic, TSP, is given by:
TS P = TS S + TS P dedicated
3.8.1 Circuit Switched Traffic

For each subcell, Atoll has already calculated the effective traffic overflow rate and the blocking rate during the dimensioning
process. Also knowing the circuit switched traffic demand, TDC, and the number of timeslots available for circuit switched
traffic, TSC, the blocking probability can be easily computed using the Erlang formulas or tables.
3.8.1.1 Erlang B
Under the current conditions of circuit switched traffic demand, TDC, and the number of timeslots available for the circuit
switched traffic, TSC, the percentage of blocked circuit switched traffic can be computed through:
TS C
 TD C 
---------------------
 TS C !
% of blocked traffic = --------------------------
TS
C
k
 TD C 
 ----------------
k!
k=0
In a network dimensioning based on Erlang B model, the circuit switched traffic overflow rate, OC, is the same as the
percentage of traffic blocked by the subcell calculated above.
3.8.1.2 Erlang C
Similarly, under the current conditions of circuit switched traffic demand, TDC, and the number of timeslots available for the
circuit switched traffic, TSC, the percentage of delayed circuit switched traffic can be computed through:
181
TS
C
 TD C 
% of traffic delayed = ----------------------------------------------------------------------------------------------------------
TS – 1
-
C
k
TS TD  TD C 
+  TS C !   1 – --------C-  
C
 TD C  ---------------
-
TS C k!
k=0
If the circuit switched traffic demand, TDC, is higher than the number of timeslots available to accommodate circuit switched
traffic, the column for this result will be empty signifying that there is a percentage of circuit switched traffic actually being
rejected rather than just being delayed under the principle of Erlang C model.
The circuit switched traffic overflow rate, OC, will be 0 if the circuit switched traffic demand, TDC, is less than the number of
timeslots available for the circuit switched traffic, TSC.
If, on the other hand, the circuit switched traffic demand, TDC, is higher than the number of timeslots available to carry the
circuit switched traffic, TSC, then there will be a certain percentage of circuit switched traffic that will overflow from the
subcell. This circuit switched traffic overflow rate, OC, is calculated as:
TD C – TS C
O C = -----------------------
TD C
3.8.1.3 Served Circuit Switched Traffic

The result of the above two processes will be a traffic overflow rate for the circuit switched traffic for each subcell, OC. The
served circuit switched traffic, STC, is calculated as:
ST C = TD C   1 – O C 
3.8.2 Packet Switched Traffic

Identifying the total traffic demand, TDT, (circuit switched traffic demand + packet switched traffic demand) as:
TD T = TD C + TD P
The following two cases can be considered.
3.8.2.1 Case 1: Total Traffic Demand > Dedicated + Shared Timeslots

In the case the total number of timeslots available is less than the total traffic demand, there will be packet switched data
traffic that will be rejected by the subcell as it will not be able to accommodate it. The following results are expected in this
case:
3.8.2.1.1 Traffic Load

The traffic load will be 100%, as the subcell will have more traffic to carry than it can. This implies that the system will be
loaded to the maximum and even saturated. Hence the user level quality of service is bound to be very unsatisfactory.
3.8.2.1.2 Packet Switched Traffic Overflow

In a 100% loaded, or even saturated subcell, the packet switched data calls will start being rejected because of shortage of
available resources. Hence there will be a perceptible packet switched traffic overflow in this subcell, OP. This overflow rate
is calculated as show below:
  TS C dedicated + TS P dedicated + TS S  – ST C 
-  100
O P = 1 – -----------------------------------------------------------------------------------------------------------
TD P
3.8.2.1.3 Throughput Reduction Factor

The resulting throughput reduction factor for a 100% loaded or saturated subcell will be 0. Hence, the throughput perceived
by the packet switched service user will be 0, implying a very bad quality of service.
3.8.2.1.4 Delay
Again for a 100% loaded or saturated subcell, the delay at the packet switched service user end will be infinite as there is no
data transfer (throughput = 0).

All the data packets will be rejected by the system since it is saturated and has no free resources to allocate to incoming data
packets. Hence, the blocking probability will be 100%.
182
3.8.2.1.6 Served Packet Switched Traffic

With the packet switched data traffic overflowing from the subcell, there will be a part of that traffic that is not served. The
served packet switched data traffic, STP, is calculated on the same principle as the served circuit switched traffic:
ST P = TD P   1 – O P 
3.8.2.2 Case 2: Total Traffic Demand < Dedicated + Shared Timeslots

In the case the total traffic demand is less than the number of timeslots available to carry the traffic, the subcell will not be
saturated and there will be some deducible values for all the data KPIs. In a normally loaded subcell, the packet switched data
traffic will have no overflow percentage. This is due to the fact that the packet switched data traffic is rather placed in a waiting
queue than be rejected.
Therefore, there will be a within limits packet switched traffic load, LP, calculated as under:
 ST C – TS C dedicated + TD P 
Timeslots
L P = ------------------------------------------------------------------------------------
-
TS P
The second parameter for computing the KPIs from the quality curves of the dimensioning model is the number of equivalent
timeslots available for the packet switched data traffic, NP, which is calculated in the same manner as in the dimensioning
process as well:
TS P
N P = -----------------------
-
TS Terminal
These parameters calculated, now Atoll can compute the required KPIs through their respective quality curves.
3.8.2.2.1 Traffic Load

The traffic load is computed knowing the total traffic demand and the total number of timeslots available to carry the entire
traffic demand:
TD T
Traffic Load = ---------------------------------------------------------------------------------
-
TS C dedicated + TS P dedicated + TS S
3.8.2.2.2 Packet Switched Traffic Overflow

In a normally loaded subcell, no packet switched data calls will be rejected. The packet switched traffic overflow will,
therefore, be 0.
3.8.2.2.3 Throughput Reduction Factor

The resulting throughput reduction factor for a normally loaded subcell is calculated through the throughput reduction factor
quality curve for given packet switched traffic load, LP, and number of equivalent timeslots, NP.
3.8.2.2.4 Delay
The resulting delay the subcell is calculated through the delay quality curve for given packet switched traffic load, LP, and
number of equivalent timeslots, NP.

The resulting blocking probability for a normally loaded subcell is calculated through the blocking probability quality curve for
given packet switched traffic load, LP, and number of equivalent timeslots, NP.
3.8.2.2.6 Served Packet Switched Traffic

As there is no overflow of the packet switched traffic demand from the subcell under consideration, the served packet
switched traffic will be the same as the packet switched traffic demand:
ST P = TD P
183
3.9 Simulations
Once you have modelled the network services and users and have created traffic maps, you can create simulations. The simu-
lation process consists of two steps:
1. Obtaining a realistic user distribution: Atoll generates a user distribution using a Monte Carlo algorithm; this user
distribution is based on the traffic database and traffic maps and is weighted by a Poisson distribution between simu-
lations of the same group.
Each user is assigned a service, a mobility type, and an activity status by random trial, according to a probability law
that uses the traffic database.
The user activity status is an important output of the random trial and has direct consequences on the next step of
the simulation and on the network interferences. A user can be either active or inactive. Both active and inactive users
consume radio resources and create interference.
Finally, another random trial determines user positions in their respective traffic zone (possibly according to the clut-
ter weighting and the indoor ratio per clutter class).
2. Modelling network regulation mechanisms: Atoll manages the GSM resources as described in "Radio Resource Man-
agement in GSM" on page 184
3.9.1 Radio Resource Management in GSM

In this section, the following are explained:
• "MSA (Mobile Station Allocation) Definition" on page 131
• "GSM Simulation Process" on page 184.
3.9.1.1 GSM Simulation Process

Figure 3.13 shows the GSM simulation algorithm. The specific simulation process in GSM consists of the following steps:
184
Figure 3.13: GSM simulation algorithm
For each simulation, the simulation process,

1. It sets initial values for the following parameters:
a. Mobile transmission power is set to the maximum mobile power.
b. Cell traffic loads for each MSA and transmitter are set to their average current value in the Transmitters table (one
traffic load value per subcell).
For each iteration k, the simulation process
2. For each circuit-switched mobile
a. Determines the server and the MSA to which the circuit-switched mobile is attached.
b. Determines the downlink C/(I+N) for each of these mobiles as explained in "DL Carrier-to-Interference Ratio Cal-
culation" on page 131.
c. Determines the uplink C/(I+N) for each of these mobiles as explained "UL C/I Evaluation" on page 156
d. Determines MSA codec modes in downlink as explained in "Calculations Based on C/(I+N)" on page 148 part of
"CQI Calculation Without Ideal Link Adaptation" on page 147.
e. Determines MSA codec modes in uplink as explained in "Codec Mode Selection" on page 159.
f. Performs the corresponding target power controls.
185
See detailed information in "Servers Selection" on page 187 and "Codec Mode Assignment and DL Power Control" on
page 187.
3. For each packet-switched mobile
a. Determines the server and the MSA to which the packet-switched mobile is attached.
b. Determines the downlink C/(I+N) for each of these mobiles as explained in "DL Carrier-to-Interference Ratio Cal-
culation" on page 131.
c. Determines the uplink C/(I+N) for each of these mobiles as explained "UL C/I Evaluation" on page 156
d. Determines MSA coding scheme in downlink as explained in "Calculations Based on C/(I+N)" on page 137 part of
"Coding Scheme Selection and Throughput Calculation Without Ideal Link Adaptation" on page 137.
e. Determines MSA coding scheme in uplink as explained in "Coding Scheme Selection" on page 158.
f. Evaluates the number of necessary timeslots to reach the minimum downlink and uplink throughput demands
(defined in the requested service) of the users randomly ranked.
g. Performs the corresponding target power controls. The number of timeslots in DL and UL are obviously not linked.
See detailed information in "Servers Selection" on page 187 and "Coding Scheme Assignment, Throughput Evaluation
and DL Power Control" on page 188.
4. It equally shares the remaining resources to packet-switched users who did not reach their maximum throughput
demands. Resources and throughputs are finally assigned to each packet-switched user.
See detailed information in "Codec Mode Assignment and DL Power Control" on page 187.
5. It updates the traffic loads, Half-Rate traffic ratios, DL power control gains and DTX gains of all the subcells according
to the resources in use and the total resources.
See detailed information in "Subcell Traffic Loads Management" on page 189, "Half-Rate Traffic Ratio Management"
on page 189, "DL Power Control Gain Management" on page 189 and "DTX DL Gain Management" on page 190.
6. It updates the UL traffic loads of all the subcells and the UL noise rises of all the TRXs according to the resources in use
and the total resources.
7. Performs the convergence test to see whether the differences between the current and the new loads are within the
convergence thresholds.
The convergence criteria are evaluated at the end of each iteration k, and can be written as follows:
Subcell i Subcell Subcell

TL DL – GSM = Max  TL i i
– TL DL – GSM 
k All Subcell
 DL – GSM k k – 1
i

PCG DL – GSM =  PCG i i 
k
Max
 DL – GSM k – PCG DL – GSM k – 1
All Subcell
i

TL UL – GSM = Max  TL i i
– TL UL – GSM 
k All Subcell
 UL – GSM k k – 1
i
TRX i TRX TRX

Max NR UL – GSM – NRUL – GSM 
i i
NR UL – GSM =
k All TRX k k – 1
i
Subcell i Subcell i Subcell i TRX i

If TL DL – GSM , PCG DL – GSM , TL UL – GSM and NR UL – GSM are the simulation convergence
Req Req Req k
thresholds defined when creating the simulation, Atoll stops the simulation in the following cases.
Convergence: Simulation has converged between iteration k - 1 and k if:
Subcell i Subcell i Subcell i Subcell i Subcell i Subcell i
TL DL – GSM  TL DL – GSM AND PCG DL – GSM  PCG DL – GSM AND TL UL – GSM  TL UL – GSM AND
k Req k Req k Req
TRX i TRX i
NRUL – GSM  NR UL – GSM
k Req
No convergence: Simulation has not converged even after the last iteration, i.e., k = Max Number of Iterations defined
Subcell Subcell Subcell Subcell
i i i i
when creating the simulation, if: TL DL – GSM  TL DL – GSM OR PCGDL – GSM  PCG DL – GSM OR
k Req k Req
Subcell i Subcell i TRX i TRX i
TL UL – GSM  TL UL – GSM OR NR UL – GSM  NR UL – GSM .
k Req k Req
186
8. Repeats the above steps (from step 2.) for the iteration k+1 using the new calculated loads as the current loads until
convergence.
3.9.1.2 Servers Selection

For a given network, the service areas of each transmitter are evaluated in the same way than an HCS server study with 0 dB
margin. In other words, each pixel, is covered by the best server of each HCS layer, assuming the received signal strength is
greater than the reception threshold defined on that layer.
In addition to the coverage condition above, for a given mobile distribution, a mobile might be served by a transmitter if its
mobility (as assigned by Atoll at the beginning of the simulation) does not exceed the maximum speed permitted on that layer.
Finally the frequency band(s) in use in the transmitter have to be supported by the user terminal.
In none of these conditions are fulfilled, the mobile is rejected with the condition "No Coverage".
If these conditions are fulfilled, as a result, each mobile then has a list of potential servers, each server being on a different
HCS layer. For each mobile list, Atoll sorts the potential servers according to their HCS layer priority in decreasing order.
On the very first iteration of the simulation, the mobile selects the highest priority transmitter. During the iterative process,
if the mobile is regularly rejected from the highest priority transmitter, it will select the second highest priority transmitter
and so on, until convergence.
In addition, if the mobile is rejected from a layer, even after convergence, the algorithm will try to attach this mobile to a lower
priority layer until no solution can be found.
3.9.1.3 Codec Mode Assignment and DL Power Control

Two types of services can be assigned to users: circuit-switched and packet-switched ones. The network has been set up and
dimensioned in order to first serve circuit services, and then to serve packet services with the remaining resources.
When serving a circuit-switched user, depending on the computed radio conditions at the server location, a codec mode is
assigned to a user. Depending on this codec mode, the user will use either an entire timeslot (any AMR, EFR or FR codec mode)
or half a timeslot (HR codec mode).
As explained in "GSM Simulation Process" on page 184, the resource element assigned to a mobile station is an MSA.
Depending on the assigned MSA, the level of quality at the mobile might be different, and consequently, its served codec
mode so as the required number of timeslots.
Assuming a server is selected for each mobile, several MSAs are candidate. For each candidate MSA, a codec mode study is
run, using the computed C/(I+N) and based on the user terminal and mobility (See "Calculations Based on C/(I+N)" on
page 148 for more information). For each MSA, a codec mode is obtained. For each mobile, the list of candidate codec modes
is saved.
At the beginning of a simulation iteration, no traffic is attached to MSAs. Their load starts from 0 and is increased as traffic
increases and mobiles are attached to them. For a given user, within his MSA list, the MSA having currently the lowest load is
selected and, as a consequence, the load of this MSA is now increased. The effect of this mechanism results in a load balancing
of MSAs within a transmitter.
When MSAs are almost full, Atoll selects the MSAs the most optimised in term of timeslot occupancy. As an example, to
optimise the resource allocation, a codec mode costing half a timeslot might be chosen instead of a codec mode costing an
entire timeslot in the case the MSA with the lowest cost would have been chosen.
This mechanism is then reproduced for all the users requesting a circuit-switched service.
i
For each MSA k, the assigned codec mode i corresponds to a quality target:  C  I  Target . Due to the radio conditions, and using
k
the victim max power, a  C  I  Max is obtained.
k i
If  C  I  Max   C  I  Target , no codec mode can be served and the mobile is rejected with the condition "No Service".
k i
If  C  I  Max   C  I  Target , the corresponding codec mode is assigned to the mobile. If the MSA is on the BCCH, no power
k
control is applied. For any other TRX type, Atoll evaluates the minimum required power P Min in order to reduce the quality
i
at the user’s terminal to  C  I  Target for the assigned MSA k.
The maximum allowed power reduction is set at 30 dB by default. This means that the
power cannot be reduced by more than 30 dB from the initial to final C/(I+N), after power
control.
The power control is considered achieved when the final C/(I+N) is at less than 1 dB from
i
the  C  I  Target .
187
To summarise, at this step, each circuit-switched user is assigned a MSA, a codec mode, a corresponding number of timeslots
i
(0.5 or 1) and a corresponding minimum required power to get the  C  I  Target of the served MSA.
Then, if the user has been dropped as inactive at the beginning of the simulation, his corresponding number of timeslots is
consumed but no DL power is considered for this specific user. Inactive users only participate in the timeslot management but
do not affect DL power.
Finally, if the user has been dropped as active at the beginning of the simulation, both timeslots and powers have to be
considered to make him connected.
3.9.1.4 Coding Scheme Assignment, Throughput Evaluation and DL Power

Control
After having served the circuit traffic over one iteration, the algorithm now tries to serve packet-switched traffic.
When serving a packet-switched user, depending on the computed radio conditions at the server location, a coding scheme
is assigned to a user and a throughput per timeslot is obtained. Then some timeslots are assigned to each packet-switched
service user in order to obtain a throughput between the min and the max DL throughput demand per user defined in the
considered service properties.
As explained in "MSA (Mobile Station Allocation) Definition" on page 131, the resource element assigned to a mobile station
is an MSA. Depending on the assigned MSA, the level of quality at the mobile might be different, and consequently, its served
coding scheme so as the required number of timeslots to get a certain throughput demand. For packet-switched traffic, the
timeslot Assignment is realised in two steps. In the first step, Atoll tries to allocate the minimum throughput demand of the
service. In the second step, using remaining resources (timeslots), Atoll tries to allocate more throughput up to the maximum
throughput demand of the service. If a user cannot get its minimum throughput demand for insufficient number of available
timeslots, the user is rejected with the condition "Resource Saturation".
Assuming a server is selected for each mobile, several MSAs are candidate. For each candidate MSA, a coding scheme study
is run, using the computed C/(I+N) and based on the user terminal and mobility (See "Calculations Based on C/(I+N)" on
page 139 for more information). For each MSA, a coding scheme is obtained, from which we get a throughput per timeslot.
As explained in "Packet Throughput and Quality Analysis: Application Throughput (kbps)" on page 144, the maximum of
timeslots the user can benefit is the minimum between the number of DL timeslots defined in the selected terminal and
service. Considering the minimum DL throughput demand for the service, one can estimate how many timeslots are needed
to get that throughput on each MSA. Then, Atoll only keeps the MSAs for which this number of timeslots is lower than the
number of timeslots supported (see above) and for which there is enough remaining timeslots. Then, for each mobile, the list
of candidate coding schemes is saved.
For a given user, within his MSA list, the MSA having currently the lowest load is selected and, as a consequence, the load of
this MSA is now increased. In the same way than for circuit traffic, the effect of this mechanism results in a load balancing of
MSAs within a transmitter.
This mechanism is then reproduced for all the users requesting a packet-switched service. At this step, each packet-switched
service has a coding scheme and, ideally, is supposed to be served his DL minimum throughput demand.
The second step of resources allocation for packet-switched traffic is to share the remaining resources between connected
users in order they get their maximum throughput demand. As an example, let’s imagine than a MSA is already occupied as
follows:
• 2 TS for circuit-switched service users (3 users: 2 HR codec modes + 1 FR codec mode)
• 2.4 TS for packet-switched service users after the first step (2 users).
If this MSA is defined over a TCH subcell, its capacity is 8 TS. In other words, 4.4 TS have been used, and 3.6 TS remain. The
two packet-switched users have obtained their minimum throughput demand. In order to reach their maximum throughput
demand, the remaining TS are equally shared between the two connected users: 1.8 TS per user. If the first user can get his
maximum throughput demand with only 1.5 TS, the remaining 0.3 TS will be able to be used by the user. As a consequence,
this second user could benefit of 2.1 TS in order to get his maximum demand. If, finally, he only needs 1.3 TS to get this
demand, 0.8 TS remain unused for that MSA.
This mechanism of equally share of remaining resources is then applied for all the connected packet-switched service users
over all their MSAs.
j
For each MSA k, the assigned coding scheme j corresponds to a quality target:  C  I  Target . Due to the radio conditions, and
k
using the victim max power, a  C  I  Max is obtained.
k j
If  C  I  Max   C  I  Target , no coding scheme can be served and the mobile is rejected with the condition "No Service".
188
k j
If  C  I  Max   C  I  Target , the corresponding coding scheme is assigned to the mobile. If the MSA is on the BCCH, no power
k
control is applied. For any other TRX type, Atoll evaluates the minimum required power P Min in order to reduce the quality
j
at the user’s terminal to  C  I  Target for the assigned MSA k.
The maximum allowed power reduction is set at 30 dB by default. This means that the
power cannot be reduced by more than 30 dB from the initial to final C/(I+N), after power
control.
The power control is considered achieved when the final C/(I+N) is at less than 1 dB from
i
the  C  I  Target .
To summarise, at this step, each packet-switched user is assigned a MSA, a coding scheme, a corresponding number of
j
timeslots (which might not be an integer value) and a corresponding minimum required power to get the  C  I  Target of the
served MSA.
3.9.1.5 Subcell Traffic Loads Management

When circuit-switched and packet-switched traffic have been served or rejected, Atoll performs an update on several
parameters. The first parameter to be updated are the subcell DL and UL traffic loads. Considering that subcell loads are values
which are unique per traffic pool (e.g. BCCH and TCH subcells belong to the same traffic pool because they are in charge of
the same traffic area) in DL and in UL, the number of timeslots necessary to connect the traffic have to be summed up over
the several MSAs over a same traffic pool.
For the traffic pool TP i , the subcell DL traffic load is computed as follows:
 TS DL used
MSA TP
TL TP = -------------------------------------------------
i - where the number of DL TS available for a BCCH subcell is 7 and 8 for any other subcell.
i DL
 TS DL available
MSA
TP i
The DL traffic load value is then assigned to all the subcells of a same traffic pool.
For the traffic pool TP i , the subcell UL traffic load is computed as follows:
 TS UL used
MSA
TP
TL TP = -------------------------------------------------
i - where the number of UL TS available for a BCCH subcell is 7 and 8 for any other subcell.
i UL
 TS UL available
MSA TP
i
The UL traffic load value is then assigned to all the subcells of a same traffic pool.
3.9.1.6 Half-Rate Traffic Ratio Management

The second parameter at the end of an iteration is the Half-rate traffic ratio. This is the percentage of half-rate voice traffic in
the subcell. This value is used to calculate the number of timeslots required to respond to the voice traffic demand and is
evaluated per traffic pool. This value referring to voice traffic only, circuit-switched users only are taken into account in its
evaluation.
 HR users
MSA TP
HR RATIO TP = ------------------------------------
i
i -. 
users represents HR and FR circuit-switched service users.
users MSATP
i
MSA TP
i
The Half-Rate traffic ratio is then assigned to all the subcells of a same traffic pool.
189
3.9.1.7 DL Power Control Gain Management

At the end of each iteration, the subcell DL power control gain is evaluated by taking into account all the connected users:
• active and inactive circuit-switched service users (assuming each inactive user does not cost any DL power but only
some timeslots)
• all packet users
From the minimum required powers evaluated at the end of "Codec Mode Assignment and DL Power Control" on page 187
and "Coding Scheme Assignment, Throughput Evaluation and DL Power Control" on page 188 in order to get respectively the
appropriate codec modes and coding schemes without any excess of unneeded power, an average minimum required power
is obtained for each mobile connected to the subcell S as follows:
k
 PMin  TSi
i
S
i--------------------------------
- = P Moy where i are the mobiles connected to the subcell S, over its MSAs
S
 TS i
i
P Max
The ratio PCG S
= ---------------S (in dB), where P Max is the max power of the considered subcell, represents the mean power
P Moy S
S
control gain, due to active and inactive users, which can be assigned to the subcell.
It is essential to note that there is no power control on the BCCH and, consequently, the mean power control gain on the BCCH
is 0.
3.9.1.8 DTX DL Gain Management

A certain gain representing inactive circuit-switched service users has also to be evaluated. In "DL Power Control Gain
Management" on page 189, the mean DL power control gain concerns both active and inactive users. The DTX gain models
the fact that inactive circuit-switched users, even if they are connected to the network, do not produce the same level of
interference than active circuit-switched users.
From the minimum required powers evaluated at the end of "Codec Mode Assignment and DL Power Control" on page 187 in
order to get the appropriate codec modes without any excess of unneeded power, an average minimum required power is
obtained for each circuit-switched active mobile connected to the subcell S as follows:
ki

active
P Min  TS i
active
i active  S
------------------------------------------------------------ = P Moy where i active are the circuit-switched active mobiles connected to the subcell S, over
S active
 TS i
active
i active
its MSAs
P Moy
S
The ratio -------------------------- (in dB), where P Moy is average requested power defined in "DL Power Control Gain Management" on
P Moy S
S active
page 189 above, represents the DTX gain, due to circuit-switched active users, which can be assigned to the subcell.
3.9.1.9 GSM Simulation Results

At the end of the simulations, an active user can be connected in DL if:
• he has a serving cell assigned,
• For a circuit-switched (resp. packet-switched) service, he has a codec mode (resp. coding scheme) corresponding to
his activity status,
• he is not rejected due to resource saturation.
If a user is rejected during server determination, the cause of rejection is "No Coverage". If a user is rejected because quality
is too low to obtain any codec mode or coding scheme, the cause of rejection is "No Service". If a user is rejected because he
cannot be allocated a sufficient number of resources to obtain its codec mode or coding scheme, the cause of rejection is
"Resource Saturation," i.e., all of the cell’s resources were used up by other users.
Considering only the connected traffic at the end of the GSM part of the simulation process, the main results obtained are:
• At the subcell level
• Subcell traffic loads (UL and DL)
• DL Power control gains
• DTX gains
190
• Half-rate traffic ratios

• At the TRX level
• intra-technology UL noise rises
Subcell traffic loads, DL Power control gains and intra-technology UL noise rises can be used as input for GSM quality-based
coverage predictions.
3.10 Automatic Neighbour Allocation

The intra-technology neighbour allocation algorithm takes into account all the TBC transmitters. It means that all the TBC
transmitters of the .atl document are potential neighbours.
The transmitters to be allocated will be called TBA transmitters. They must fulfil the following conditions:
• They are active,
• They satisfy the filter criteria applied to the Transmitters folder,
• They are located inside the focus zone,
• They belong to the folder on which allocation has been executed. This folder can be either the Transmitters folder or
a group of transmitters or a single transmitter.
Only TBA transmitters may be assigned neighbours.
If no focus zone exists in the .atl document, Atoll takes into account the computation
zone.
3.10.1 Neighbour Allocation for All Transmitters

We assume a reference transmitter A and a candidate neighbour, transmitter B.
When automatic allocation starts, Atoll checks following conditions:
• The distance between both transmitters must be less than the user-definable maximum inter-site distance. If the
distance between the reference transmitter and the candidate neighbour is greater than this value, then the
candidate neighbour is discarded.
Atoll calculates the effective distance, which corresponds to the real inter-transmitter distance weighted by the
azimuths of antennas. For information on the effective distance calculation, see "Appendix: Calculation of the Inter-
Transmitter Distance" on page 195.
• The calculation options:
• Force co-site transmitters as neighbours: This option enables you to force transmitters located on the reference
transmitter site in the candidate neighbour list. This constraints can be weighted among the others and ranks the
neighbours through the importance field.
• Force adjacent transmitters as neighbours: This option enables you to force transmitters geographically adjacent
to the reference transmitter in the candidate neighbour list. This constraint can be weighted among others and
ranks the neighbours through the importance field.
191
Adjacency criterion:
Geographically adjacent transmitters are determined on the basis of their Best Server
coverages in 2G (GSM GPRS EDGE) projects. More precisely, a transmitter TXi is
considered adjacent to another transmitter TXj if there exists at least one pixel of TXi Best
Server coverage area TXj is the 2nd Best Server. The ranking of the adjacent neighbour
transmitter increases with the number of these pixels. The figure below shows the above
concept.
• When the adjacency option is checked, adjacent cells are sorted and listed from the
most adjacent to the least, depending on the above criterion. Adjacency is relative
to the number of pixels satisfying the criterion.
• Force adjacent layers as neighbours: If selected, Atoll adds all the transmitters adjacent across network layers to
the reference transmitter to the candidate neighbour list. The weight of this constraint is always the average of
the Min and Max values defined for the adjacency factor. This weight is used to calculate the rank of each
neighbour and its importance. Transmitters are considered adjacent across layers if they belong to different layers
and have a coverage overlap of at least one pixel.
• Force symmetry: This option enables user to force the reciprocity of a neighbourhood link. Therefore, if the
reference transmitter is a candidate neighbour of another transmitter, the latter will be considered as candidate
neighbour of the reference transmitter.
If the neighbours list of a transmitter is full, the reference transmitter will not be added
as a neighbour of that transmitter and that transmitter will be removed from the
reference transmitter’s neighbours list. You can force Atoll to keep that transmitter in
the reference transmitter’s neighbours list by adding the following option in the Atoll.ini
file:
[Neighbours]
DoNotDeleteSymmetrics = 1
• Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. Therefore,
you may force/forbid a transmitter to be candidate neighbour of the reference transmitter.
• Delete existing neighbours: When selecting the Delete existing neighbours option, Atoll deletes all the current
neighbours and carries out a new neighbour allocation. If not selected, the existing neighbours are kept.
• If the Use Coverage Conditions check box is selected, there must be an overlapping zone ( S A  S B ) with a given cell
edge coverage probability. Otherwise, only the distance criterion is taken into account.
The overlapping zone ( S A  S B ) is defined as follows:
• SA is the area where the received signal level from transmitter A is greater than a minimum signal level. SA is the
coverage area of reference transmitter A restricted between two boundaries; the first boundary represents the
start of the handover area (best server area of A plus handover margin named “handover start”) and the second
boundary shows the end of the handover area (best server area of A plus the margin called “handover end”).
• SB is the coverage area where the candidate transmitter B is the best server.
192
SA  SB
-  100 ) if the option “Take into account Covered Area” is
Atoll calculates either the percentage of covered area ( -----------------
SA
selected, or the percentage of traffic covered on the overlapping area S A  S B for the option “Take into account Covered
Traffic”. Then, it compares this value to the % minimum covered area (minimum percentage of covered area for the option
“Take into account Covered Area” or minimum percentage of covered traffic for the option “Take into account Covered
Traffic”). If this percentage is not exceeded, the candidate neighbour B is discarded.
The coverage condition can be weighted among the others and ranks the neighbours through the importance field (see
number 4 below).
Figure 3.14: Overlapping Zones
Atoll uses traffic map(s) selected in the default traffic analysis in order to determine the
percentage of traffic covered in the overlapping area.
• The importance values are used by the allocation algorithm to rank the neighbours according to the allocation reason,
and to quantify the neighbour importance.
Atoll lists all neighbours and ranks them by importance value so as to eliminate some of them from the neighbour list if the
maximum number of neighbours to be allocated to each transmitter is exceeded. If we consider the case for which there are
15 candidate neighbours and the maximum number of neighbours to be allocated to the reference transmitter is 8. Among
these 15 candidate neighbours, only 8 (having the highest importances) will be allocated to the reference transmitter.
In the Results part, Atoll only displays the transmitters for which it finds new neighbours. For these transmitters, it provides
the list of neighbours, the number of neighbours and the maximum number of neighbours allowed for each transmitter. In
addition, it indicates the importance (in %) of each neighbour and the allocation reason. Therefore, a neighbour may be
marked as exceptional pair, co-site, intra-HCS and inter-HCS adjacent, coverage or symmetric. For neighbours accepted for
co-site, intra-HCS or inter-HCS adjacency and coverage reasons, Atoll displays the percentage of area meeting the coverage
conditions (or the percentage of covered traffic on this area) and the corresponding surface area (km2) (or the traffic covered
on the area in Erlangs), the percentage of area meeting the adjacency conditions and the corresponding surface area (km2).
Finally, if cells have previous allocations in the list, neighbours are marked as existing.
193
• By default, the automatic neighbour allocation compares the defined Max inter-
site distance with the effective inter-transmitter distance. As a consequence,
there can be cases where the real distance between assigned neighbours is
higher than the Max inter-site distance, because the effective distance is
smaller. You can force Atoll to compare the Max inter-site distance with the real
inter-site distance by adding the following lines in Atoll.ini:
[Neighbours]
RealInterSiteDistanceCondition=1
• By default, the neighbour importance calculated with respect to distance is based
on the global Max inter-site distance setting for all neighbour candidates. As a
consequence, there can be cases where the calculated importance is different
when the global Max inter-site distance is modified. To avoid that, you can force
Atoll to prioritise the individual distances between reference transmitters and
their respective neighbour candidates by adding the following lines in Atoll.ini:
[Neighbours]
CandidatesMaxDistanceInImportanceCalculation=1
3.10.2 Neighbour Allocation for a Group of Transmitters or One

Transmitter
In this case, Atoll allocates neighbours to:
• TBA transmitters,
• Neighbours of TBA transmitters marked as exceptional pair, intra-HCS or inter-HCS adjacent and symmetric,
• Neighbours of TBA transmitters that satisfy coverage conditions.
Automatic neighbour allocation parameters are described in "Neighbour Allocation for All Transmitters" on page 191.
3.10.3 Neighbour Importance Calculation

The neighbour importance depends on the distance from the reference transmitter and on the neighbourhood cause (cf. table
below); this value varies between 0 and 100%.
Neighbourhood cause When Importance value

Only if the Delete existing neighbours option is not selected
Existing neighbour Existing importance
and in case of a new allocation
Exceptional pair Only if the Force exceptional pairs option is selected 100 %
Only if the Force co-site cells as neighbours option is
Co-site transmitter Importance Function (IF)
selected
Only if the Force intra-HCS adjacents as neighbours option is
Adjacent transmitters Importance Function (IF)
selected
Only if the Force inter-HCS adjacents as Neighbours option is
Adjacent layer Importance Function (IF)
selected
Neighbourhood relationship that
Only if the % minimum covered area is exceeded Importance Function (IF)
fulfils coverage conditions
Symmetric neighbourhood
Only if the Force neighbour symmetry option is selected Importance Function (IF)
relationship
Except the case of forced neighbours (importance = 100%), priority assigned to each neighbourhood cause is determined
using the Importance Function (IF). The IF considers the following factors for calculating the importance:
• The distance factor (Di) denoting the distance between the possible neighbour transmitter and the reference
transmitter.
d-
 Di  = 1 – ----------
d max
d is the effective distance (in m). It corresponds to the real inter-transmitter distance weighted by the azimuths of
antennas. For information on the effective distance calculation, see "Appendix: Calculation of the Inter-Transmitter
Distance" on page 195.
194
d max is the maximum distance between the reference transmitter and a possible neighbour.
• The co-site factor (C): a Boolean,

• The adjacency factor (A): the percentage of adjacency,
• The overlapping factor (O): the percentage of overlapping.
The minimum and maximum importance assigned to each of the above factors can be defined.
Factor Min importance Default value Max importance Default value

Distance factor (Di) Min(Di) 1% Max(Di) 10%
Overlapping factor (O) Min(O) 10% Max(O) 30%
Adjacency factor (A) Min(A) 30% Max(A) 60%
Co-site factor (C) Min(C) 60% Max(C) 100%
The Importance Function is evaluated as follows:

Resulting IF using the default values
Neighbourhood cause Importance Function
from the table above
Min(O)+Delta(O){Max(Di)(Di)+(100%-
Coverage 10%+20%{10%(Di)+90%(O)}+1%+9%(Di)
Max(Di))(O)}+Min(Di)+Delta(Di)(Di)
Adjacent layer (Min(A)+Max(A))/2 45%
Min(A)+Delta(A){Max(Di)(Di)+Max(O)(O)+ 30%+30%{10%(Di)+30%(O)
Adjacent transmitters
(100%-Max(Di)-Max(O))(A)}+Min(Di)+Delta(Di)(Di) +60%(A)}+1%+9%(Di)
Min(C)+Delta(C){Max(Di)(Di)+Max(O)(O)+ 60%+40%{10%(Di)+30%(O)
Co-site transmitters
Delta(X)=Max(X)-Min(X)
• Set Min(Di) and Max(Di) to 0% if you do not want to take into account the distance
factor in the importance calculation.
• If the Min and Max value ranges of the importance function factors do not overlap,
the neighbours will be ranked by neighbour cause. With the default values for
minimum and maximum importance fields, neighbours will be ranked in this order:
co-site neighbours, adjacent neighbours, and neighbours allocated based on coverage
overlapping.
• If the Min and Max value ranges of the importance function factors overlap, the
neighbours may be ranked differently. There can be a mix of the neighbourhood
causes.
• The default value of Min(O) = 1% ensures that neighbours selected for symmetry will
have an importance greater than 0%. With a value of Min(O) = 0%, neighbours
selected for symmetry will have an importance field greater than 0% only if there is
some coverage overlapping.
3.10.4 Appendix: Calculation of the Inter-Transmitter Distance

Atoll takes into account the real distance ( D in m) and azimuths of antennas in order to calculate the effective inter-
transmitter distance ( d in m).
d = D   1 + x  cos  – x  cos  
x = 0.3% so that the maximum D variation does not exceed 1%.
Figure 3.15: Inter-Transmitter Distance Computation
195
The formula above implies that two cells facing each other will have a smaller effective distance than the real physical
distance. It is this effective distance that will be taken into account rather than the real distance.
3.11 AFP Appendices

3.11.1 The AFP Cost Function
The notations listed hereafter are used to describe the cost function:
• TRG: Group of TRXs
• TRGs: Set of all the TRGs
• : If and only if
• g : Size of any group g
• ARFCN: Set of all the frequencies
ARFCN
• 2 : Set of all the subsets of frequencies
• x : The largest integer x
ARFCN
• A i g : Number of times a group g  2 is assigned to TRGi in the assignment A
For example:
• When i is NH, A i g = 1  g is a single member group containing one of the frequencies assigned at TRGi.
If |g| is not 1 or if g does not contain a frequency assigned at i, then A i g = 0 .
• When i is BBH, A i g can be either 0 or equal to the number of TRXs in TRGi.
A i g = Number of TRXs in TRGi  g is the set of frequencies assigned to TRXs of TRGi. (|g| = number of TRXs in
TRGi).
When we talk about "TRXs of i using g", and in the case of BBH, then there are |g| such virtual TRXs, each using
the entire group g and having a virtual MAIO [0, |g| - 1].
• When i is SFH, A i g must be less than or equal to the umber of TRXs in TRGi. A i g = n  g is the set of frequencies
assigned to n TRXs of TRGi.
We assume all the groups assigned to TRGi to have the same length.
• TSi: Number of timeslots available for each TRX in TRGi

• TLi: Traffic load of TRGi (calculated or user-defined)
TL i = #Erlangs of a single TRX in TRGi divided by TSi
• TSUi: Downlink timeslot use ratio (due to DTX) at TRGi

• CFi: Cost factor of TRGi (AFP Weight)
• QMINi: Minimum required quality (in C/I) at TRGi
• PMAXi: Percentage permitted to have quality lower than QMINi at TRGi
• REQi: Required number of TRXs at TRGi
A communication uses the group g in TRGi if its mobile allocation is g. The probability to be interfered is denoted by
P i i' g  A  (i’ is the TRX index). Different TRX indexes may have different MAIOs. P i i' g  A  is a function of the whole
frequency assignment. The precise definition of the term “to be interfered” is provided afterwards. The probability penalty
due to violating a separation constraint is P i i' g  A  . It is a function of the whole frequency assignment as well.
The term “Atom” will be used in the following context:

For two TRGs, i and k,
ATOM  i   ATOM  k 
i and k are synchronised, have the same HSN, the same MAL length and the same hopping mode.
NH TRGs or BBH TRGs are always in separate atoms. If two TRGs interfere but are not in the same atom, these can be taken
as unsynchronised. The quality of unsynchronised TRGs is a function of all possible frequency combinations. For synchronised
TRGs, pairs of frequencies emitted at the same time are known.
196
3.11.1.1 Cost Function

The Atoll AFP cost function is a TRX based cost and not an interference matrix entry based cost. It counts the impaired traffic
of the network TRXs in weighted Erlangs.
The cost function  is reported to the user during the AFP progress with the help of its 5 components:  mis ,  sep ,  comp ,
 corr and  dom .
 =  mis +  sep +  comp +  corr +  dom
,
 mis represents the missing TRX cost component
 sep represents the separation component
 comp represents the additional cost component (interference, cost of changing a TRX)
 corr represents the corrupted TRX cost component
 dom represents the out-of-domain frequency assignment cost component
 mis =   MIS_TRX i     TL i  CF i  TS i
i  TRGs
 corr =   CORR_TRX i     TL i  CF i  TS i
i  TRGs
 dom =   DOM_TRX i     TL i  CF i  TS i
i  TRGs
 
 
 
 sep =  
  ' i i' g  A   TL i  CF i  TS i

i  TRGs  
ARFCN
 g2 
i'  TRXs of i using g
 
 
 
 comp =  
  '' i i' g  A   TL i  CF i  TS i

i  TRGs  
ARFCN
 g2 
i'  TRXs of i using g
In the above equations,
• i’ is the TRX index belonging to  0 1 ... A i g – 1  .

• MIS_TRX i is the number of missing TRXs for the subcell i.
MIS_TRX i = MAX 0 REQi –  A i g

ARFCN
g2
•  is the cost value for a missing TRX. This value can vary between 0 and 10. The default cost value is set to 1 and can
be modified in the AFP module properties dialogue.
• CORR_TRX i is the number of corrupted TRXs for the subcell i.
•  is the cost value of a corrupted TRX. This value can vary between 0 and 10. The default cost value is set to 10 and
can be modified in the AFP module properties dialogue.
• DOM_TRX i is the number of TRXs, for the subcell i, having out-of-domain frequencies assigned.
•  is the cost value of a TRX with out-of-domain frequencies assigned. This value can vary between 0 and 1. The default
cost value is set to 0.5 and can be modified in the AFP module properties dialogue.
And, as mentioned earlier, a virtual TRX is considered in case of BBH.
If i’ is valid, the algorithm evaluates the cost of a valid TRX. This cost has two components, ' i i' g  A  and '' i i' g  A  .
197
• ' i i' g  A  is the separation violation probability penalty.
• '' i i' g  A  is complementary probability penalty due to interference and the cost of modifying a TRX.
If the option “Take into account the cost of all the TRXs” available in the AFP module properties dialogue is selected,
then,
' i i' g  A  = P' i i' g  A  and '' i i' g  A  = P'' i i' g  A 
Or if the option “Do not include the cost of TRXs having reached their quality target” available in the AFP module
properties dialogue is selected, the algorithm compares P' i i' g  A  + P'' i i' g  A  with the quality target specified for
i, P MAX :
If P' i i' g  A  + P'' i i' g  A   P MAX ,
Then ' i i' g  A  = P' i i' g  A  and '' i i' g  A  = P'' i i' g  A  .
Otherwise,
Both ' i i' g  A  and '' i i' g  A  will be equal 0.
P' i i' g  A  is the same as ' i i' g  A  (separation violation probability penalty) and P'' i i' g  A  the same as '' i i' g  A 
(complementary probability penalty due to interference and the cost of modifying a TRX) in most cases. These are explained
in detail in the next sections.
3.11.1.2 Cost Components

Separation violation and interference cost components are described hereafter. Parameters considered in the cost function
components can be fully controlled by the user. Some of these parameters are part of the general data model (quality
requirements, percentage of interference allowed per subcell), while others (such as separation costs and diversity gains) can
be managed through the properties dialogue of the Atoll AFP module.
3.11.1.2.1 Separation Violation Cost Component

The separation violation cost component is evaluated for each TRX. Estimation is based on costs specified for the required
separations.
Let SEP_CONSTR i k denote the required separation constraint between TRGi and TRGk. Let Cost s z denote the user defined
separation penalty for a required separation “s” and actual separation “z”. SEP i k v is used instead of Cost SEP_CONSTR as
i k z
abbreviation.
The AFP module properties dialogue takes probability percentages as inputs while this
document deals in probability values.
 ii'kgg'k' is considered to be the effect of a separation violation on the i' th TRX of TRGi assigned the group g, caused by the
k' th TRX of TRGk assigned the group g' .
 denotes the overall weight of the separation violation cost component. This value can be between 0 and 1, set to 1 by
default. It can be modified in the AFP module properties dialogue.
ik represents the weight of the specific separation constraint between i and k. This specific weight depends on the type of
separation violation and follows the following priority rule:
1. Exceptional pairs
2. Co-transmitters
3. Co-site
4. Neighbours
For example, if a pair of subcells are co-site and neighbours at the same time, they will be considered as co-site because higher
priority. Hence, ik of these subcells will be the weight of co-site relations. If only a neighbour relation exists between two
198
subcells, then ik will be further weighted by the neighbour relation importance. The value of ik remains between 0 and 1.
The default weights of each type of separation are available in the Separation cost tab.
If ATOM  i   ATOM  k 
 SEP i k f – f'
fg
f'  g'
Then  ii'kgg'k' =   ik  ----------------------------------------- , which is same for all values of k.
g  g'
If ATOM  i  = ATOM  k 
 SEP i k g  – g' 
     – 
Then  ii'kgg'k' =   ik  -----------------------------------------------------------------------------
f_n 0 1 ... F_N 1
F_N
In the above equations, F_N  g  is the number of frames in the MAL g. F_N  g  = g .
Since F_N  g  = F_N  g'  , we shortly denote the two as F_N .
Let f_n denote the instantaneous frame number from 0 to F_N .
th
While  =  f_n + MAIO A  modulo F_N and g  is the  frequency in g,
i g i'
th
And  =  f_n + MAIO A  modulo F_N and g'  is the  frequency in g’.
k g' k'
In addition, frequencies belonging to a MAL with a low fractional load, and breaking a separation constraint, should not be
weighted equally as in a non-hopping separation breaking case. Therefore, the cost is weighted by an interferer diversity gain.
1
Ĝ i k g g' = -----------------------------------------------------------
-
 0.1  SEP_GAIN  i k g  g'  
10
The separation gain, denoted by SEP_GAIN  i k g  g'  is basically a function of the MAL length (and, of course, of the
hopping mode). With frequency hopping, the effects of DTX and traffic load become more significant (due to the consideration
of the average case instead of the worst case). For this reason, it is possible to consider these effects in
SEP_GAIN  i k g  g'  through the relevant option available in the Advanced tab of the AFP module properties dialogue.
Without this option, the SEP_GAIN  i k g  g'  is:
SEP_GAIN  i k g  g'  = I_DIV  g 
I_DIV  g  is the user defined interferer diversity gain (dB) for a given MAL length. It is used in P i i' g  A  definition as well.
On the other hand, if this option is selected, the SEP_GAIN  i k g  g'  becomes,
 2 + ASYN_GAIN  i k g'  
SEP_GAIN  i k g  g'  = I_DIV  g  +0.5  TSU_GAIN  k   min  10 4 +   2 + I_DIV  g    -----------------------------------------------------------------  
4
TSU_GAIN  k  = log 10  -------------------------- ,

1
TL k  TSU k
And ASYN_GAIN  i k g'  = 0 if ATOM(i) = ATOM(k)

I_DIV( g'  Otherwise
More than one separation violations may exist for a TRX. Many “small” Ĝ i k g g' and ' ii'kgg' have to be combined to form
one cost element, the P' i i' g  A  . This is done through iterating over all violating assignments and by summing up an
equivalent to the probability of not being violated while considering each separation violation as an independent probability
event. This sum is naturally limited to 100% of the TRX traffic, and is given by,
199
 
 
1 –

  1 –  ii'kgg'k'  Ĝ i k g g'  

P' i i' g  A  =  
 k  TRGs 
 g'  2
ARFCN 
 k'  TRXs of k using g' 
 
In the above formula, if  k = i  , then  k'  i'  , so that interference with itself is not taken into account.
3.11.1.2.2 Interference Cost Component

The interference cost component is evaluated for each TRX. Its estimation is based on interference histograms calculated for
pairs of subcells. In addition, it takes into account frequency and interferer diversity gains and models frequency hopping and
gain due to DTX.
Interference histograms are described in User Manual (GSM GPRS EDGE project
management, GSM GPRS EDGE network optimisation, GSM GPRS EDGE generic AFP
management). Interference histograms can also be exported to files. For further
description, refer to "Interference" on page 205.
When estimating P'' i i' g  A  , the following problems are encountered:
• The QMINi C/I quality indicator corresponds to the accumulated interference level of all interferers while the C/I
interference histograms correspond to pair-wise interferences.
• Both QMINi and the histograms correspond to a single frequency. In case of a MAL containing more than one
frequencies, interferences on several different frequencies of a MAL must be combined.
This estimation, presented below, is the simplest possible as it solves the first problem by linear summation and
truncation at the value of 1 and it solves the second problem by averaging and adding the two diversity gains:
• F_DIV  g  , the frequency diversity gain, and

• I_DIV  g  , the interferer diversity gain.
Hereafter,  denotes the global weight of interference cost component. This value can vary between 0 and 1 and is set to
0.35 by default, which can be modified in the AFP module properties dialogue.
Let F_N  g  be the number of frames in the MAL g. F_N  g  = g .
Let f_n denote the instantaneous frame number from 0 to F_N .
Let MAIO A be the j’th MAIO of A k g' , j is one of the  0 1 ... A k g' – 1  TRXs.
k g' j
The value of MAIO A is one of  0 1 ... g' 

k g' j
If TRGk is NH, then MAIO A = 0.

k g' j
If TRGk is BBH, then MAIO A = j .

k g' j
As said earlier, in case of BBH, we consider g' virtual TRXs, the jth TRX has the MAIO j.
Let g i be the ith frequency in the group g.
Similar to the definition of  ii'kgg'k' , ' ii'kgg'k' is defined as an interference event. ' ii'kgg'k' is the effect interference on the i' th
TRX of TRGi assigned the group g, caused by the k' th TRX of TRGk assigned the group g' .
If ATOM  i   ATOM  k 
Probability  -----  Q_UB i k f f'

C
 I ik 
Then ' ii'kgg'k' =  --------------------------------------------------------------------------
g  g'
f  g f'  g'
Q_UB i k f f' = QMIN i – f – f'  ADJ_SUP + INTERF_GAIN  i k g  g' 
If ATOM  i  = ATOM  k 
200
Then,
Since F_N  g  = F_N  g'  , these are both represented by F_N .
 Probability  -----  Q_UB i k f f' 

C
  I 

ik
 ii'kgg'k' =  -------------------------------------------------------------------------
-
 F_N 
f_n   0 1 ... F_N – 1 
 
,
f = g ,
f' = g'  ,
 =  f_n + MAIO A  modulo F_N ,

i g i'
 =  f_n + MAIO A  modulo F_N ,

k g' k'
Q_UB i k f f' = QMIN i – f – f'  ADJ_SUP + INTERF_GAIN  i k g  g' 
 
 
 
Therefore, we have, P'' i i' g  A  = 1 –   1 – P' i i' g  A   

  1 –  ii'kgg'k'   – P' i i' g  A 

 k  TRGs

 
ARFCN
g'  2
k'  TRXs of k using g'
In the above formula, if  i = k  , then  k'  i'  , so that interference with itself is not taken into account.
The sum is limited to 100% of the TRX traffic. INTERF_GAIN  i k g  g'  is quite similar to SEP_GAIN  i k g  g'  . The
only difference is the frequency diversity gain, F_DIV  g  , added to SEP_GAIN  i k g  g'  .
3.11.1.2.3 I_DIV, F_DIV and Other Advanced Cost Parameters

When combining interference effects (or separation violation effects) on different frequencies belonging to a MAL, the
following considerations should be taken into account:
1. Non-linearity of Frame Error Rate (FER) with respect to average C/I conditions and MAL length.
2. Interference Diversity Gain. This factor represents that the effect of average negative effects over user geographic
location are directly proportional to the MAL length.
3. Frequency Diversity Gain. This factor models the gain due to diversity of multi-path effects and should be applied to
the interference cost component only.
4. The fact that long MALs with synthesized hopping permit discarding the worst case estimation and include a gain due
to DTX and low traffic load at the interferer end.
The Advanced properties tab shown in the figure below facilitates modelling these effects.
201
Figure 3.16: The Advanced tab of the AFP module Properties dialogue
The Interference Diversity Gain table lists the values of I_DIV provided as a functions of MAL length. This gain is applied to the
interference cost component and to the separation constraint violation cost component. Therefore, it provides a means to
model the non-linear FER effects and interference diversity both. The default values in this table correspond to the curve
y = 2  log 10  x  . This equation generates values somewhat lower than empirical best-found values (this is because we
prefer a slightly pessimistic cost function to be on the safe side).
The other table contains the F_DIV values, which are the same as the I_DIV values by default.
3.11.2 The AFP Blocked Traffic Cost

This section provides additional information on the AFP cost components used for the optimisation of the number of TRXs.
This optimisation is performed for each traffic pool in the network. In most cases, the traffic pool is equivalent to a transmitter
and corresponds to the BCCH and TCH subcells. In more complex cases, a traffic pool may include additional subcells, and
more than one traffic pools may exist per transmitter.
The cost component described below, and the recalculation of traffic loads, is only used when the AFP performs the
oprimisation of the number of TRXs.
The notations listed hereafter are used for the description.
• {BCCH, TCH(1), TCH(2), …, TCH(n)}: Subcells of a traffic pool.
For concentric cells, at least two traffic pools exist per transmitter.
The BCCH subcell may not always be part of the pool’s TRX types.
• {d(0), d(1), d(2), …, d(n)}: Number of required TRXs of each TRX type in the pool.
• {ts(0), ts(1), ts(2), …, ts(n)}: Numbers of traffic timeslots.
• {L(0), L(1), L(2), …, L(n)}: Traffic loads.
• {CF(0), CF(1), CF(2), …, CF(n)}: AFP cost factors.
• CS (Erlangs): Overall circuit-switched traffic demand of the traffic pool (Subcells table or
traffic analysis results).
• PS (Data Timeslots): Overall packet-switched traffic demand of the traffic pool (Subcells table or
traffic analysis results).
If CS or PS is less than 1, its value is set to 1 in order to avoid working
with transmitters carrying no traffic.
• {nb(0), nb(1), nb(2), …, nb(n)}: Number of TRXs in the frequency plan.
• {HR(0), HR(1), HR(2), …, HR(n)}: TCH HR use ratios.
202
3.11.2.1 Calculation of New Traffic Loads Including Blocked Traffic Loads

During the optimisation of the number of TRXs, traffic loads are calculated in order to determine the blocked traffic loads
BL  nb  . The blocked traffic load is then multiplied by the AFP cost weight and the number of timeslots to calculate the
blocked traffic cost.
Without the optimisation of the number of required TRXs, the network’s weighted Erlangs are calculated as follows:
n
WE =  d  i   ts  i   L  i   CF  i 
i=0
With the optimisation of the number of TRXs, the network’s weighted Erlangs are calculated as follows:
n
WE =  nb  i   ts  i    BL  nb  + L  nb    CF  i 
i=0
BL  nb  and L  nb  represent the load estimation and the blocked load estimation of the AFP. They are calculated at traffic
pool level for the vector {nb(0), nb(1), nb(2), …, nb(n)} as follows:
 HR 
PS +  CS   1 – ------------- 
 2 
BL  nb  + L  nb  = ---------------------------------------------------------------
 n

Max  1 nb  i   ts  i 

 
 i=0 
n
HR = Max i = 0  HR  i  
BL  nb  is determined from the above equation once L  nb  is known. L  nb  is obtained from the Erlang B equation
applied to the traffic pool demand and the total number of timeslots (TTS):
 n 
 nb  i   ts  i -
TTS = Max  1
 
-----------------------------
 HR  

 i = 0  1 – -------------
2 
The Max() function above gives 1 timeslot when there is no TRX.

P Blocking = ErlangB  CS TTS 
The above equations give the number of served circuit-switched timeslots (SCS):
SCS =  1 – -------------
HR   CS   1 – P
Blocking 
2 
The number of served packet-switched timeslots (SPS) is obtained as follows:

  n
 
   
SPS = Min  PS Max 1


nb  i   ts  i  – SCS 



 i=0 
 
L  nb  is given by:
SCS + SPS
L  nb  = --------------------------------------------------------------
-
 n

Max  1  nb  i   ts  i 
 
 i=0 
BL  nb  is given by:
PS + CS   1 – ------------- HR 
 2 
BL  nb  = --------------------------------------------------------------- – L  nb 
 n

 nb  i   ts  i 
Max  1  
 i=0 
203
Once L  nb  and BL  nb  are known, L  nb  replaces TLi in the cost function (See "The AFP Cost Function" on page 196),
and BL  nb  is used to generate a new cost component, the blocked Erlangs of the pool:
n
 nb  i   ts  i   BL  nb   CF  i 
i=0
3.11.2.2 Recalculation of CS and PS From Traffic Loads

In earlier versions, the detailed traffic demand information is not available. In order to guide the AFP to generate it from the
loads, the following two equations with three variables must be solved. The equations are solvable due to the monotone
nature of the Erlang B function.
Inputs for a given traffic pool:
• {d(0), d(1), d(2), …, d(n)}: Number of required TRXs of each TRX type in the pool
• L: Traffic load
 n 
 d  i   ts  i  
• TTS' : TTS' = Max  1
 

----------------------------
HR  
 i = 0  1 – -------------
2 
• MB: Maximum blocking rate (between 0 and 1).
The ratio of packet-switched demand is given by:
PS
R = ----------------------------------------------------
-
PS + CS   1 – ------------- HR 
 2 
Here, we assume that a traffic load of 1 is generated by a demand of (1+MB)*TTS’ which generates a blocking rate of MB. In
other words, the ratio is calculated so that the worst case blocking rate is BM, giving a load of 1.
The following equations are solved to find PS’, CS’, and R’, which are calculated for a traffic load of 1.
MB = ErlangB  CS' TTS' 
PS'
R' = -------------------------------------------------------
PS' + CS'   1 – ------------- HR 
 2 
PS' - + CS'
 1 + MB   TTS' = ---------------------------
 1 – -------------
HR 
 2 
When the traffic load of a pool is not 1, PS is different from PS’ and CS is different from CS’. Here, however, we assume that
R’ = R. This assumption implies that R is more or less the same as MB for big traffic pools and considerably larger than MB for
smaller pools.
The following equations are solved to find PS, CS, and R, which are calculated for the actual traffic loads.
PS
R = ----------------------------------------------------
-

PS + CS   1 – ------------- HR 
2
P Blocking = ErlangB  CS TTS' 
SCS =  1 – -------------
HR   CS   1 – P
 Blocking 
2 
  n
 
 
SPS = Min  PS Max  1 d  i   ts  i  – SCS 

  i=0  
 
n
SCS + SPS =  d  i   ts  i   L  i 
i=0
The above five equations are solved to get the values of the five variables PS, PC, P Blocking , SCS, SPS, and calculate the cost.
204
3.11.2.3 Testing the Blocked Cost Using Traffic Analysis

As long as the conditions below hold truw, the blocked cost calculation in the AFP and the effective overflow calculation in the
KPI calculation and dimensioning use the same algorithm. The conditions are:
• The AFP cost factors are 1,
• The HR ratios are the same within the subcells of a traffic pool,
• The dimensioning model is based on Erlang B,
• The timeslot configurations are the default ones,
• There exists at least one TRX in the traffic pool (and at least one Erlang of traffic),
• All transmitters belong to the same HCS Layer.
L  nb 
Effective Overflow rate = 1 – --------------------------------------
L  nb  + BL  nb 
Output: New values for CS and PS.
3.11.3 Interference
This appendix provides a high-level overview of interference taken into account by the AFP.
3.11.3.1 Using Interferences

If interferences are to be taken into account by the AFP, they must be calculated or imported beforehand. In order to do this,
the user should previously decide to take interferences into account (enabling the loading of all the potential interferers).
Otherwise, Atoll does not allow performing their computation by disabling the histogram part in the corresponding dialogue.
3.11.3.2 Cumulative Density Function of C/I Levels

For each [interfered subcell, interfering subcell] pair, Atoll calculates a C/I value on each pixel of the interfered subcell service
area (as if the two subcells share the same channel). Then, Atoll integrates these C/I values to determine a C/I distribution and
transforms this distribution function into a cumulative density function in the normal way.
In Atoll, both the IMco and IMadj are represented by this Cumulative Density function This implies that each query for the
probability to have C/I conditions worse than X dB requires a single memory access: the co-channel interference probability
at X dB. In order to deduce the adjacent interference probability value, Atoll looks up the cumulative density function at the
value corresponding to X - Y dB, Y dB being the adjacency suppression value. The following example may be helpful in further
clarifying this concept:
Example: Let [TX1, BCCH] and [TX2, BCCH] be the interfered and interfering subcells respectively. The service areas
for both have been defined by Best Server with 0 dB margin. The interference probability is stated in percentage of
interfered area.
Figure 3.17: The cumulative density of C/I levels between [TX1, BCCH] and [TX2, BCCH]
In this case, we observe that the probability for C/I (BCCH of TX2 effecting the BCCH of TX1) being greater than 0 is
100% (which is normal because TX1 is the Best Server). The probability of having a C/I value at least equal to 31 dB is
31.1%. For a required C/I level of 12 dB on the BCCH of TX1, the interference probability is 6.5% (as this requirement
is fulfilled with a probability of 93.5%).
205
The subcell power offset does not enter the calculation results in the .clc file. It is added
later by the AFP interface. On the other hand, its influence on the subcell service zone is
taken into account in the .clc file.
3.11.3.3 Precise Definition

Pci  v n C_I  is defined to be the probability of a communication (call) occupying a timeslot in subcell v (victim) to have C/
I conditions of C_I with respect to a co-channel interference from the BCCH TRX of cell n (neighbour). We assume C_I values
to be discrete and in dB. CDF(Pci) is the cumulative density function of Pci:
CDF  Pci  v n C_I   =  Pci  v n x 

xC I
3.11.3.4 Precise Interference Distribution Strategy

Why does Atoll calculate and maintain precise interference distributions, while the most common solution (used by most
other tools) is rather to compress the information into two values: the co-channel and adjacent-channel interference
probabilities?
The reason is simply that it,
• improves the AFP result,
• introduces very little (or no) overhead, and
• creates more generic interference information.
3.11.3.4.1 Direct Availability of Precise Interference Distribution to the AFP

In the presence of frequency hopping, and when one or more frequencies are common (or adjacent) in two interfering MAL
sequences, the hopping gain depends on following factors:
• the MAL length,
• the traffic load on the interferer TRX,
• DTX level, and
• the number of common (and adjacent) frequencies in the two MALs.
All these factors cannot be pre-calculated since it is the AFP that determines the MAL length and the MAL frequencies.
3.11.3.4.2 Efficient Calculation and Storage of Interference Distribution

In the innermost loop of the calculation process Atoll increments a counter each time a C/I level has a certain value. In the
case of a two-entry IM, there are only two counters for each [interfered, interferer] pair. In the case of precise distribution
information, there are about 40 counters per pair. In both cases, the number of operations is the same: one increment of an
integer value. Once Atoll finishes the counting for an [interfered, interferer] pair, it compresses the information from the
counters to a Cumulative Density Function (CDF) representation. In this way, access to interference probability at a certain
level is instantaneous. Thus, the only overheads are the read / write times to the files and the memory occupation at running
time. These two overheads are negligible and do not affect the calculations, the heaviest part of the task.
3.11.3.4.3 Robustness of the IM

By having precise C/I distributions calculated and exported, the user is free to change the following settings without the need
for recalculating their interference distributions:
1. Quality requirements of network elements (required C/I, % Probability Max, …),
2. C/I weighting (the interference levels above and below the C/I target),
3. Separation requirements and/or neighbour relations,
4. Hopping gain values, DTX activities, traffic load levels, HSNs, synchronisation information,
5. Any frequency assignment setting (MAL length directives, frequency domains, assignment strategies, number of
required TRXs, cost function parameters, …), or
6. Remove equipment
By not mixing any of the elements above, the interference information keeps its original probability units and is easier to check
and validate. Therefore, the user spends less time on interference recalculations than in the case of a two-entry matrix (
“everything” is included).
206
3.11.3.5 Traffic Load and Interference Information Discrimination

Atoll maintains the traffic load separate from the interference information. The reasons for implementing this strategy are
explained here.
Let us look at the possible alternatives to this strategy:
1. The mixed option: The interference information contains the traffic information as well. In this way, each IM entry
will contain the quantity of traffic interfered if a co-channel / adjacent channel reuse exists.
2. The separated option: The AFP has separate access to traffic load information and to interference probabilities (As in
Atoll).
Knowing the difference between the two alternative solutions explains why the second strategy has been opted for for Atoll.
However, in detail, this has been done because:
• Option 2 is a superset that contains option 1. But option 1, being a subset, does not contain option 2 (i.e. once the
information are mixed they cannot be separated).
• It does not create any overhead (the size of the additional information is negligible compared to the size of the IM).
• It helps keeping the unit definitions simpler.
• It is facilitates merging IMs with different traffic units.
• The traffic information can be used for weighting the separation violation component.
• The traffic load can be used in deciding whether a TRX can be left uncreated.
For example, if there are too many TRXs at a site and the user wishes that the AFP remove one of them, in order to be
able to not violate site constraints, the AFP must know the traffic loads in order to choose a low load TRX to be
removed.
• The gain introduced by the traffic load of the interferer depends on the hopping mode and the MAL length.
Incorporating this gain in the IM (as a result of the mixed option) means that the IMs become hopping-mode and MAL-
size dependent. This is a bad idea since the AFP should be able to change the MAL. And the user should be able to
change the hopping mode without recalculating the IM. In addition, an IM calculated externally to Atoll, with a non-
hopping BCCH can be used for the hopping TCH.
A third option also exists. Though, this option is so practically useless due to its inefficiency. It consists in mixing IM and traffic
but still keeping the traffic in its isolated form. This is again a bad idea because of the unit definition and the variety of IM
sources. It involves less benefits than the option chosen in Atoll.
207
208
Chapter 4
UMTS HSPA Networks

• "General Prediction Studies" on page 211
• "Definitions" on page 214
• "UMTS HSPA Prediction Studies" on page 284
• "Primary Scrambling Code Allocation" on page 319
• "Automatic GSM-UMTS Neighbour Allocation" on
page 329
© Forsk 2015
210
AT330_TRR_E1 Chapter 4: UMTS HSPA Networks
4 UMTS HSPA Networks

This chapter describes all the calculations performed in Atoll UMTS HSPA documents.
All the calculations are performed on TBC (to be calculated) transmitters. For the definition of TBC transmitters please refer
to "Path Loss Matrices" on page 103.
4.1 General Prediction Studies

4.1.1 Calculation Criteria
Three criteria can be studied in point analysis (Profile tab) and in common coverage studies. Study criteria are detailed in the
table below:
Study criteria Formulas
Signal level received from a transmitter on a carrier (cell)
Signal level ( P rec ) in dBm
P rec  ic  = EIRP  ic  – L path – M Shadowing – model – L Indoor + G term – L term
Path loss ( L path ) in dBm L path = L model + L ant

Tx
Total losses ( L total ) in dBm L total =  L path + L Tx + L term + L indoor + M Shadowing – model  –  G Tx + G term 
where,
EIRP is the effective isotropic radiated power of the transmitter,
ic is a carrier rank,
L model is the loss on the transmitter-receiver path (path loss) calculated by the propagation model,
L ant is the transmitter antenna attenuation (from antenna patterns),

Tx
M Shadowing – model is the shadowing margin. This parameter is taken into account when the option “Shadowing taken into
account” is selected,
L Indoor are the indoor losses, taken into account when the option “Indoor coverage” is selected,
L term are the receiver losses,
G term is the receiver antenna gain,
G Tx is the transmitter antenna gain,
L Tx is the transmitter loss ( L Tx = L total – DL ). For information on calculating transmitter loss, see "UMTS, CDMA2000, TD-
SCDMA, WiMAX, and LTE Documents" on page 30.
• EIRP  ic  = P pilot  ic  + G Tx – L Tx ( P pilot  ic  is the cell pilot power).

• When you make the prediction, you can consider the best carrier of all bands or the
best carrier of a particular frequency band (Best (All Bands/Specific Band) option). In
this case, Atoll takes the highest pilot power of carriers to calculate the signal level
received from a transmitter.
• Atoll considers that G term and L term equal zero.

4.1.2.1 Profile Tab
Atoll displays either the signal level received from the selected transmitter on a carrier ( P rec  ic  ), or the highest signal level
received from the selected transmitter on the best carrier.
211
Chapter 4: UMTS HSPA Networks ©Forsk 2015
For a selected transmitter, it is also possible to study the path loss, L path , or the total
losses, L total . Path loss and total losses are the same on any carrier.

Analysis provided in the Reception tab is based on path loss matrices. So, you can study reception from TBC transmitters for
which path loss matrices have been computed on their calculation areas.
For each transmitter, Atoll displays either the signal level received on a carrier, ( P rec  ic  ), or the highest signal level received
on the best carrier.
Reception bars are displayed in a decreasing signal level order. The maximum number of reception bars depends on the signal
level received from the best serving cell. Only reception bars of cells whose signal level is within a 30 dB margin from the best
serving cell can be displayed.
• For a selected transmitter, it is also possible to study the path loss, L path , or the total
• You can use a value other than 30 dB for the margin from the best serving cell signal
level, for example a smaller value for improving the calculation speed. For more
information on defining a different value for this margin, see the Administrator
Manual.
4.1.3 Coverage Studies

For each TBC transmitter, Txi, Atoll determines the selected criterion on each pixel inside the Txi calculation area. In fact, each
pixel within the Txi calculation area is considered as a potential (fixed or mobile) receiver.
Coverage study parameters to be set are:
• The study conditions in order to determine the service area of each TBC transmitter,
• The display settings to select how to colour service areas.

Atoll uses parameters entered in the Condition tab of the coverage study property dialogue to predetermine areas where it
will display coverage.
We can distinguish three cases:

The service area of Txi corresponds to the bins where:
Txi Txi Txi
MinimumThreshold  P rec  ic   or L total or L path   MaximumThreshold

Txi Txi Txi
And
P rec  ic   Best  P rec  ic   – M

Txi Txj
ji
M is the specified margin (dB).

Best function: considers the highest value.
212
• If the margin equals 0 dB, Atoll will consider bins where the signal level received from
Txi is the highest.
• If the margin is set to 2 dB, Atoll will consider bins where the signal level received
from Txi is either the highest or 2dB lower than the highest.
• If the margin is set to -2 dB, Atoll will consider bins where the signal level received
from Txi is 2dB higher than the signal levels from transmitters, which are 2nd best
servers.

Txi Txi Txi
And
Txi nd Txj
P rec  ic   2 Best  P rec  ic   – M
ji
2nd Best function: considers the second highest value.
Txi is the second highest.
from Txi is either the second highest or 2dB lower than the second highest.
from Txi is 2dB higher than the signal levels from transmitters, which are 3rd best
servers.

4.1.3.2.1 Plot Resolution
Prediction plot resolution is independent of the matrix resolutions and can be defined on a per study basis. Prediction plots
are generated from multi-resolution path loss matrices using bilinear interpolation method (similar to the one used to
evaluate site altitude).

It is possible to display the transmitter service area with colours depending on any transmitter attribute or other criteria such
as:
Atoll calculates signal level received from the transmitter on each pixel of each transmitter service area. A pixel of a service
area is coloured if the signal level is greater than or equal to the defined minimum thresholds (pixel colour depends on signal
level). Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as many
layers as transmitter service areas. Each layer shows the different signal levels available in the transmitter service area.
Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. Where other service
areas overlap the studied one, Atoll chooses the highest value. A pixel of a service area is coloured if the signal level is greater
than or equal to the defined thresholds (the pixel colour depends on the signal level). Coverage consists of several
layer corresponds to an area where the signal level from the best server exceeds a defined minimum threshold.
Path Loss (dB)
Atoll calculates path loss from the transmitter on each pixel of each transmitter service area. A pixel of a service area is
coloured if path loss is greater than or equal to the defined minimum thresholds (pixel colour depends on path loss). Coverage
areas. Each layer shows the different path loss levels in the transmitter service area.
213
Total Losses (dB)
Atoll calculates total losses from the transmitter on each pixel of each transmitter service area. A pixel of a service area is
coloured if total losses is greater than or equal to the defined minimum thresholds (pixel colour depends on total losses).
Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. Where other service
areas overlap the studied one, Atoll determines the best transmitter and evaluates path loss from the best transmitter. A pixel
of a service area is coloured if the path loss is greater than or equal to the defined thresholds (pixel colour depends on path
layers as defined thresholds. Each layer corresponds to an area where the path loss from the best server exceeds a defined
minimum threshold.
Atoll calculates signal levels received from transmitters on each pixel of each transmitter service area. Where service areas
overlap the studied one, Atoll determines the best transmitter and evaluates total losses from the best transmitter. A pixel of
a service area is coloured if the total losses is greater than or equal to the defined thresholds (pixel colour depends on total
layers as defined thresholds. Each layer corresponds to an area where the total losses from the best server exceed a defined
minimum threshold.
Number of Servers
Atoll evaluates how many service areas cover a pixel in order to determine the number of servers. The pixel colour depends
on the number of servers. Coverage consists of several independent layers whose visibility in the workspace can be managed.
There are as many layers as defined thresholds. Each layer corresponds to an area where the number of servers is greater than
or equal to a defined minimum threshold.
On each pixel of each transmitter service area, the coverage corresponds to the pixels where the signal level from this
transmitter fulfils signal conditions defined in Conditions tab with different Cell edge coverage probabilities. There is one
coverage area per transmitter in the explorer.
On each pixel of each transmitter service area, the coverage corresponds to the pixels where the best signal level received
fulfils signal conditions defined in Conditions tab. There is one coverage area per cell edge coverage probability in the explorer.
4.2 Definitions
This section details the terms that describe the users and the services, the input parameters as well as the formulas used in
simulations and predictions (coverage predictions and point analysis).
4.2.1 Glossary
In this chapter, we will use the following terms to describe the users and the services:
R99 users: The Circuit (R99) and Packet (R99) service users. They require an R99 bearer.
HSDPA users: The users that only support HSDPA. They have an HSDPA-capable terminal and one of these services:
• Packet (HSDPA - Best Effort),
• Packet (HSDPA - Variable Bit Rate).
HSDPA users require an R99 bearer (i.e. the A-DPCH radio bearer) and an HSDPA bearer.
HSPA users: The users that support both HSDPA and HSUPA. They have an HSPA-capable terminal and one of these services:
• Packet (HSPA - Best Effort),
• Packet (HSPA - Variable Bit Rate),
• Packet (HSPA - Constant Bit Rate).
HSPA users require an R99 bearer (i.e. the E-DPCCH/A-DPCH radio bearer), an HSDPA bearer and an HSUPA bearer.
DC-HSDPA users: The dual-cell HSDPA users. Users with dual-cell HSDPA-capable terminals that can simultaneously connect
to two HSDPA cells of the transmitter for data transfer. The R99 A-DPCH bearer is transmitted on one of the cells, which is
214
called the anchor cell. The user can be assigned an HSDPA bearer in each of the cells.
MC-HSDPA users: The multi-cell HSDPA users. Users with multi-cell HSDPA-capable terminals that can simultaneously connect
to several HSDPA cells of the transmitter for data transfer. The R99 A-DPCH bearer is transmitted on one of the cells, which is
called the anchor cell. The user can be assigned an HSDPA bearer in each of the cells.
DB-MC-HSDPA users: The dual-band multi-cell HSDPA users. Users with dual-band multi-cell HSDPA-capable terminals that
can simultaneously connect to several HSDPA cells on co-site transmitters using different frequency bands. The R99 A-DPCH
bearer is transmitted on one of the cells, which is called the anchor cell. The user can be assigned an HSDPA bearer in each of
the cells.
BE services: Best Effort services.
VBR services: Variable Bit Rate services.
CBR services: Constant Bit Rate services. CBR services do not support multi-cell HSDPA mode.
4.2.2 Inputs
This table lists simulation and prediction inputs (calculation options, quality targets, active set management conditions, etc.).
F ortho Clutter parameter None Orthogonality factor

Tx
F MUD Site equipment parameter None MUD factor
Term
F MUD Terminal parameter - HSDPA properties None MUD factor
cn first Frequency band parameter None First carrier number
cnlast Frequency band parameter None Last carrier number
cn Frequency band parameter None Carrier number step

Carrier rank of the current carrier
cn – cn first
ic =  -------------------------
- – cn lower
 cn 
ic Frequency band parameter None
Where cn lower is the number of
carrier numbers lower than cn
including excluded carriers and
carriers of other frequency bands
Threshold for macro diversity
AS_Th  Txi ic  Cell parameter None specified for a transmitter on a given
carrier ic
Handover margin for a transmitter
on a given carrier ic.
M HO  Txi ic  Cell parameter None
Used for best serving cell selection in
UMTS and HSPA specific predictions.
Cell Individual Offset for a
transmitter on a given carrier ic.
CIO  Txi ic  Cell parameter None
Used for best serving cell selection in
UMTS and HSPA specific predictions.
The minimum pilot RSCP required for
RSCP min  Txi ic  Cell parameter or Global parameter W a user to be connected to the
transmitter on a given carrier
req  E----c Ec/I0 target on downlink for the best

Q pilot  I 0  threshold Mobility parameter None
serving cell
CM – activation Pilot RSCP threshold for compressed

RSCP pilot Global parameter None
mode activation
CM – activation Ec/I0 threshold for compressed
Q pilot Global parameter None
mode activation
215
DL E b
DL
 ----
Q req - (Reception equipment, R99 bearer, Mobility) parameter None Eb/Nt target on downlink
 N t req
DL Downlink Eb/Nt target increase due

Q req Global parameter None
to compressed mode activation
UL E b
UL
 ----
-
Q req  N t req (Reception equipment, R99 bearer, Mobility) parameter None Eb/Nt target on uplink
UL Uplink Eb/Nt target increase due to

Q req Global parameter None
compressed mode activation
Max Number of channel elements
N CE – UL  N I  Site parameter None
available for a site on uplink
N CE – DL  N I  Site parameter None
available for a site on downlink
Number of channel elements of a
N CE – UL  N I  Simulation result None
site consumed by users on uplink
N CE – DL  N I  Simulation result None
site consumed by users on downlink
Number of channel elements used by
Overhead Site equipment parameter - UL overhead resources for common
N CE – UL None the cell for common channels on
channels/cell
uplink
Overhead Site equipment parameter - DL overhead resources for common
N CE – DL None the cell for common channels on
channels/cell
downlink
R99 – T CH Number of channel elements used
N CE – UL (R99 bearer, site equipment) parameter None
for R99 traffic channels on uplink
R99 – T CH Number of channel elements used
N CE – DL (R99 bearer, site equipment) parameter None
for R99 traffic channels on downlink
Number of channel elements
HSUPA
N CE (HSUPA bearer, site equipment) parameter None consumed by the HSUPA bearer on
uplink
Max Maximum Iub backhaul throughput
TP Iub – UL  N I  Site parameter kbps
for a site in the uplink
Max Maximum Iub backhaul throughput
TP Iub – DL  N I  Site parameter kbps
for a site in the downlink
Iub backhaul throughput for a site in
TP Iub – UL  N I  Simulation result kbps
the uplink
Iub backhaul throughput for a site in
TP Iub – DL  N I  Simulation result kbps
the downlink
Iub throughput required by the cell
Overhead
TP Iub – DL  N I  Site equipment parameter kbps for common channels in the
downlink
HSDPA
Overhead Iub Site equipment parameter % HSDPA Iub backhaul overhead
E1  T1  Ethernet Throughput carried by an E1/T1/

TP Site equipment parameter kbps
Ethernet link
R99 – T CH Iub backhaul throughput consumed
TPIub – UL (R99 bearer, site equipment) parameter kbps
by the R99 bearer in the uplink
R99 – T CH Iub backhaul throughput consumed
TPIub – DL (R99 bearer, site equipment) parameter kbps
by the R99 bearer in the downlink
HSUPA Iub backhaul throughput consumed
TP Iub (HSUPA bearer, site equipment) parameter kbps
by the HSUPA bearer in the uplink
Max Maximum number of 512 bit-length
N Codes  Txi ic  Simulation constraint None
OVSF codes available per cell (512)
Number of 512 bit-length OVSF
N Codes  Txi ic  Simulation result None
codes used by the cell
216

Number of 256 bit-length OVSF
Overhead Site equipment parameter - DL overhead resources for common
N Codes None codes used by the cell for common
channels/cell
channels
Maximum number of 16 bit-length
HSPDSCH – Max
N Codes  Txi ic  Cell parameter (for HSDPA only) None OVSF codes available per cell for HS-
PDSCH
Minimum number of 16 bit-length
HSPDSCH – Min
N Codes  Txi ic  Cell parameter (for HSDPA only) None OVSF codes available per cell for HS-
PDSCH
NF term Terminal parameter None Terminal Noise Figure
Transmitter parameter (user-defined or calculated from transmitter
NF Tx None Transmitter Noise Figure
equipment characteristics)
K 1.38 10-23 J/K Boltzman constant
T 293 K Ambient temperature
W 3.84 MHz Hz Spreading Bandwidth
Tx DL
NRinter – techno log y Cell parameter None Inter-technology downlink noise rise
Cell parameter
Tx UL
NRinter – techno log y Only used in uplink interference-based calculations of the Monte- None Inter-technology uplink noise rise
Carlo simulation
Network parameter Interference reduction factor

RF  ic ic adj  If not defined, it is assumed that there is no inter-carrier None between two adjacent carriers ic
interference and ic adj
Inter-technology Channel Protection

Network parameter between the signal transmitted by Tx
Tx m and received by m assuming the
ICP ic  ic If not defined, it is assumed that there is no inter-technology None
i
downlink interferences due to external transmitters frequency gap between ic i (external
network) and ic
UL Cell parameter (user-defined or simulation result) % Total uplink load factor of the cell
X
UL Uplink cell load contribution due to
X R99 Simulation result %
R99 traffic
UL Uplink cell load contribution due to
X HSUPA Cell parameter %
HSUPA traffic
UL Maximum uplink load factor of the
X max Simulation constraint (global parameter or cell parameter) %
cell
DL
%Power max Simulation constraint (global parameter or cell parameter) % Maximum percentage of used power
Tx Tx UL
N0 NF Tx  K  T  W  NR inter – techno log y W Thermal noise at transmitter
Term Tx DL
N0 NF Term  K  T  W  NR inter – techno log y W Thermal noise at terminal
Rc –3
W  10 W bps Chip rate
UL
f rake efficiency Site equipment parameter None Uplink rake receiver efficiency factor
DL Downlink rake receiver efficiency

f rake efficiency Terminal parameter None
factor
R99 R99 bearer downlink peak
TP P – DL R99 bearer parameter kbps
throughput
DL Downlink spreading factor for active
F spreading  Active user  R99 bearer parameter None
users
DL Downlink spreading factor for
F spreading  Inactive user  R99 bearer parameter None
inactive users
217

Ratio between DPCCH and DPCH
transmission duration on downlink
DL DPCCH and DPCH respectively refer
rc R99 bearer parameter None
to the Dedicated Physical Control
Channel and Dedicated Physical
Channel
Maximum connection throughput
TP max – DL Cell parameter kbps
per user on downlink
R99
TP P – UL R99 bearer parameter kbps R99 bearer uplink peak throughput
UL
f act Service parameter kbps Uplink activity factor for the service
DL Downlink activity factor for the

f act Service parameter kbps
service
UL Uplink activity factor on E-DPCCH
f act –ADPCH Service parameter kbps
channels
DL Downlink Activity factor on A-DPCH
f act –ADPCH Service parameter kbps
channel
Minimum required bit rate that the
TPD min – UL Service parameter kbps service should have in order to be
available in the uplink
TPD min – DL Service parameter kbps service should have in order to be
available in the downlink
Maximum bit rate that the service
TPD max – UL Service parameter kbps
can require in the uplink
Maximum bit rate that the service
TPD max – DL Service parameter kbps
can require in the downlink
Ratio between the DPCCH and DPCH
powers transmitted on uplink
UL DPCCH and DPCH respectively refer
rc R99 bearer parameter None
to the Dedicated Physical Control
Channel and Dedicated Physical
Channel
Maximum connection throughput
TP max – UL Cell parameter kbps
per user on uplink
W -
Gp
DL ----------------- None Service downlink processing gain
R99
TP P – DL
W -
Gp
UL ----------------- None Service uplink processing gain
R99
TP P – UL
Index of the HSDPA bearer obtained

I HSDPABearer HSDPA bearer parameter None
by the user in the cell (Txi,ic)
DL
TP P – RLC Peak RLC throughput supported by
HSDPA bearer parameter kbps
 I HSDPABearer  the HSDPA bearer
DL
Without MIMO: TP P – RLC  I HSDPABearer 
DL Peak RLC throughput provided to the
DL With MIMO (transmit diversity): TP P – RLC  I HSDPABearer 
TPP – RLC  Tx ic  kbps user in the cell (Txi,ic) in the
With MIMO (spatial multiplexing): downlink
DL Max
TP P – RLC  I HSDPABearer    1 + f SM – Gain   G SM – 1   
218

HSDPA study result
For single-carrier HSDPA users
DL
TP P – RLC  Tx ic 
DL Peak RLC throughput provided to the
TP P – RLC kbps
user in the downlink
For DC-HSDPA users

DL
TP P – RLC  Tx ic 
ic  Tx
TPD min – DL HSDPA bearer consumption for a

C HSDPABearer -----------------------------------------------------
- %
DL
TP P – RLC  I HSDPABearer  CBR service user
DL Effective RLC throughput supported

TP E – RLC HSDPA study result kbps
by the HSDPA bearer
DL Average effective RLC throughput
TP Av – E – RLC HSDPA study result kbps
supported by the HSDPA bearer
DL Peak MAC throughput supported by
TP P – MAC HSDPA study result kbps
the HSDPA bearer
DL Effective MAC throughput supported
TP E – MAC HSDPA study result kbps
by the HSDPA bearer
DL User application throughput on
TP A HSDPA study result kbps
downlink
UL User application throughput on
TP A HSUPA study result kbps
uplink
UL User average application throughput
TP Av – A HSUPA study result kbps
on uplink
Index of the HSUPA bearer obtained
I HSUPABearer HSUPA Bearer parameter None
in the cell (Txi,ic)
Maximum number of
retransmissions a HARQ process will
N Rtx
HSUPA bearer selection parameter kbps perform for a block of data before
 I HSUPABearer  moving on to a new block of data, for
the HSUPA bearer index
UL
TP P – RLC Peak RLC throughput supported by
HSUPA bearer parameter kbps
 I HSUPABearer  the HSUPA bearer
HSUPA study result Peak RLC throughput provided to the

UL
TP P – RLC UL kbps
TP P – RLC  I HSUPABearer  user in the cell (Txi,ic) in the uplink
TPD min – UL HSUPA bearer consumption for a

C HSUPABearer -----------------------------------------------------
- %
UL
TP P – RLC  I HSUPABearer  CBR service user
UL Minimum effective RLC throughput

TP Min –E – R LC HSUPA study result kbps
supported by the HSUPA bearer
UL Average effective RLC throughput
TP Av – E – R LC HSUPA study result kbps
supported by the HSUPA bearer
UL Peak MAC throughput supported by
TP P – M AC HSUPA study result kbps
the HSUPA bearer
TP Offset Service parameter (for HSDPA only) kbps Throughput offset
f TP – Scaling Service parameter (for HSDPA only) % Scaling factor

Maximum shared power
P max  Txi  Transmitter parameter W Available only if the inter-carrier
power sharing option is activated
P SCH  Txi ic  Cell parameter W Cell synchronisation channel power
219

Cell other common channels (except
P OtherCCH  Txi ic  Cell parameter W
CPICH and SCH) power
P pilot  Txi ic  Cell parameter W Cell pilot power
Cell parameter (user-defined or simulation result) (for HSDPA only) Available cell HSDPA power
P HSDPA  Txi ic  W HSDPA: High Speed Downlink Packet
P HS – PDSCH  Txi ic  + n HS – SCCH  P HS – SCCH  Txi ic 
Access
Cell HS-PDSCH power
P HS – PDSCH  Txi ic  Simulation result (for HSDPA only) W HS-PDSCH: High Speed Physical
Downlink Shared Channel
Cell HS-SCCH power
P HS – SCCH  Txi ic  Cell parameter (for HSDPA only) W HS-SCCH: High Speed Shared Control
Channel
number of HS-SCCH channels
n HS – SCCH Cell parameter (for HSDPA only)
managed by the cell
P Headroom  Txi ic  Cell parameter (for HSDPA only) W Cell headroom power
P max  Txi ic  Cell parameter W Maximum Cell power

R99 traffic channel power
P tch  Txi ic  Simulation result W
transmitted on carrier ic
min Minimum power allowed on R99
P tch R99 bearer parameter W
traffic data channel
max Maximum power allowed on R99
P tch R99 bearer parameter W
traffic data channel
Cell HSUPA power
P HSUPA  Txi ic  Cell parameter W HSUPA: High Speed Uplink Packet
Access
Transmitter HSDPA power
P tx –H SDPA  Txi ic  Simulation result W
transmitted on carrier ic
Simulation result
P pilot  Txi ic  + P SCH  Txi ic  + P OtherCCH  Txi ic  +

Transmitter R99 power transmitted
P tx – R99  Txi ic  W
 
DL
P tch  Txi ic  + P tch  Txi ic   f act –ADPCH on carrier ic
tch(ic) used for tch(ic) used for
R99 users HSUPA users
Cell parameter (user-defined or simulation result) Transmitter total power transmitted

P tx  Txi ic  W
P tx – R99  Txi ic  + P tx – H SDPA  Txi ic  + P HSUPA  Txi ic  on carrier ic
Terminal power transmitted to
P term – R99 Calculated in the simulation but not displayed W
obtain the R99 radio bearer
Terminal power transmitted to
P term – HSUPA Calculated in the simulation but not displayed W
obtain the HSUPA radio bearer
Simulation result
P term – R99 
UL
f act – ADPCH + P term – HSUPA for HSPA users Total power transmitted by the
P term W
terminal
P term – R99 for R99 and HSDPA users
min
P term Terminal parameter W Minimum terminal power allowed
max
P term Terminal parameter W Maximum terminal power allowed
Percentage of BTS signal correctly

 BTS BTS parameter %
transmitted
Percentage of terminal signal
 term Terminal parameter %
correctly transmitted
Percentage of pilot finger -
 Clutter parameter % percentage of signal received by the
terminal pilot finger
220
G Tx Antenna parameter None Transmitter antenna gain
G Term Terminal parameter None Terminal gain

DL
G Div R99 bearer parameter - Depends on the transmitter Tx diversity None Gain due to transmit diversity
UL
G Div R99 bearer parameter - Depends on the transmitter Rx diversity None Gain due to receive diversity
Maximum spatial multiplexing gain

Max
G SM MIMO configuration parameter dB for a given number of transmission
and reception antennas
Downlink Transmit Diversity gain for
DL
G TD MIMO configuration parameter dB a given number of transmission and
reception antenna ports
f SM – Gain Clutter parameter None Spatial multiplexing gain factor
DL
G TD Clutter parameter dB Additional diversity gain in downlink

L Tx
None Transmitter lossa
L body Service parameter None Body loss
L Term Terminal parameter None Terminal loss
L indoor Clutter parameter Indoor loss
L path Propagation model result None Path loss

Result calculated from cell edge coverage probability and model Model Shadowing margin
M Shadowing – model None
standard deviation Only used in prediction studies
Result calculated from cell edge coverage probability and Ec/I0 Ec/I0 Shadowing margin
M Shadowing – Ec  Io None
DL npaths DL gain due to availability of several
DL G macro – diversity = M Shadowing – Ec  Io – M Shadowing – Ec  Io
G macro – diversity None
n=2 or 3 pilot signals at the mobile b.
M Shadowing –  Eb  Nt  Result calculated from cell edge coverage probability and DL Eb/Nt DL Eb/Nt Shadowing margin
DL
None
M Shadowing –  Eb  Nt  Result calculated from cell edge coverage probability and UL Eb/Nt UL Eb/Nt Shadowing margin
UL
None
UL npaths
G macro – diversity = M Shadowing –  Eb  Nt  – M Shadowing – Eb  Nt 
UL UL UL UL quality gain due to signal diversity
n=2 or 3 in soft handoffc.
Global parameter (default value)
Random shadowing error drawn
E Shadowing Simulation result None during Monte-Carlo simulation
Only used in simulations
221
In prediction studiesd
For Ec/I0 calculation

L path  L Tx  L term  L body  L indoor  M Shadowing – Ec  Io
-------------------------------------------------------------------------------------------------------------------------------------
G Tx  G term
For DL Eb/Nt calculation
L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt 
DL
LT ----------------------------------------------------------------------------------------------------------------------------------------------
- None Transmitter-terminal total loss
G Tx  G term
For UL Eb/Nt calculation
UL
----------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
In simulations
L path  L Tx  L term  L body  L indoor  E Shadowing
-------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
P pilot  Txi ic 
P c  Txi ic  -------------------------------- W Chip power received at terminal
LT
DL P tch  Txi ic  Bit power received at terminal on

P b  Txi ic  ----------------------------- W
LT carrier ic
DL P tx  Txi ic  Total power received at terminal

P tot  Txi ic  --------------------------- W
LT from a transmitter on carrier ic
P tch  Txi ic  Total power received at terminal

DL
P traf  Txi ic   -----------------------------
LT W from traffic channels of a transmitter
tch  ic  on carrier ic
UL P term Bit power received at transmitter on

P b  ic  ------------ W
LT carrier ic used by terminal
UL P term – R99 Bit power received at transmitter on

P b – R99  ic  ------------------------ W
LT carrier ic used by terminal
UL UL UL Bit power received at transmitter on

P b – DPDCH  ic  P b – R99  ic    1 – r c  W
DPDCH from a terminal on carrier ic
a. L Tx = L total – UL on uplink and L Tx = L total – DL on downlink. For information on calculating transmitter losses on uplink and downlink,
see "UMTS, CDMA2000, TD-SCDMA, WiMAX, and LTE Documents" on page 30.
npaths
b. M Shadowing –Ec  Io corresponds to the shadowing margin evaluated from the shadowing error probability density function (n paths) in case
of downlink Ec/I0 modelling.
npaths
c. M Shadowing –  Eb  Nt  corresponds to the shadowing margin evaluated from the shadowing error probability density function (n paths) in
UL
case of uplink soft handoff modelling.
d. In uplink prediction studies, only carrier power level is downgraded by the shadowing margin ( M Shadowing –  Eb  Nt  ). In downlink
UL
prediction studies, carrier power level and intra-cell interference are downgraded by the shadowing model ( M Shadowing –  Eb  Nt  or
DL
M Shadowing – Ec  Io ) while extra-cell interference level is not. Therefore, M Shadowing –  Eb  Nt  or M Shadowing – Ec  Io is set to 1 in downlink
DL
extra-cell interference calculation.
4.2.3 Ec/I0 Calculation

This table details the pilot quality ( Q pilot or Ec  Io ) calculations.
 DL P SCH  txi ic  Downlink intra-cell interference at

P tot  txi ic  –  BTS     P tot  txi ic  – -----------------------------
DL
I intra  txi ic  DL - W
L  T terminal on carrier ic
222
Downlink extra-cell interference at


DL
DL
I extra  ic  P tot  txj ic  W
terminal on carrier ic
txj j  i
 Ptot  txj icadj 

DL
DL Downlink inter-carrier interference
I inter – carrier  ic  txj j
W
---------------------------------------------
- at terminal on carrier ic
RF  ic ic adj 
Tx Downlink inter-technology
DL
I inter – techno log y  ic   --------------------------------------
Tx
L total  ICP ic  ic
Tx m W interference at terminal on carrier ic
ni i a
Without Pilot:
DL DL DL DL
I intra  txi ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic 
Term
DL + N0 –  1 –     BTS  P c  txi ic  Total received noise at terminal on
I 0  ic  W
carrier ic
DL DL DL
P tot  txi ic  + I extra  ic  + I inter – carrier  ic 
Total noise:
DL Term
+ I inter – techno log y  ic  + N 0
E  BTS    P c  txi ic  Quality level at terminal on pilot for

Q pilot  txi ic    ----c --------------------------------------------------
DL
- None
I0 I 0  ic  carrier ic
a. In the case of an interfering GSM external network in frequency hopping, the ICP value is weighted according to the fractional load.
4.2.4 DL Eb/Nt Calculation

This table details calculations of downlink traffic channel quality ( Q tch or  ------ ).
DL Eb
Nt DL
 DL P SCH  txi ic  Downlink intra-cell interference at

P tot  txi ic  –  BTS  F ortho   P tot  txi ic  – -----------------------------
DL
I intra  txi ic  DL - W
L T terminal on carrier ic


DL
DL
txj j  i

DL
W
---------------------------------------------
DL
I inter – techno log y  ic   -------------------------------------
L
Tx
 ICP
Tx m
- W interference at terminal on carrier ic
ni total ic i ic a
DL DL DL DL DL Term Total received noise at terminal on

N tot  ic  I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0 W
carrier ic
Without useful signal:
DL
 BTS  P b  txi ic 
E -  G DL
------------------------------------------------------------------------------------------------------ DL
Div  G p
Quality level at terminal on a traffic
Q tch  txi ic    ----b-
DL DL
DL
N tot  ic  –  1 – F ortho    BTS  P b  txi ic  None channel from one transmitter on
 N t DL
DL carrier ic b
 BTS  P b  txi ic 
-  G DL
Total Noise: -------------------------------------------
DL
Div  G p
DL
N tot  ic 
Quality level at terminal using carrier


DL DL
DL f rake efficiency  Q tch  tx k ic  ic due to combination of all
Q  ic  None
tx k  ActiveSet transmitters of the active set
(Macro-diversity conditions).
223

DL
Q  ic 
DL
G SHO --------------------------------------------------- None Soft handover gain on downlink
DL
Q tch  BestServer ic 
DL Required transmitter traffic channel

req Q req
P tch  txi ic  -  P tch  txi ic 
----------------- W power to achieve Eb/Nt target at
DL
Q  ic  terminal on carrier ic
b. Calculation option may be selected in the Global parameters tab. The chosen option will be taken into account only in simulations. In point
analysis and coverage studies, Atoll uses the option “Total noise” to evaluate DL and UL Eb/Nt.
4.2.5 UL Eb/Nt Calculation

This table details calculations of uplink traffic channel quality ( Q tch or  ------ ).
UL Eb
Nt UL
Total power received at transmitter

 Pb
UL
UL intra  ic 
I tot  txi ic  W from intra-cell terminals using carrier
term
txi
ic


UL
UL extra P b  ic 
I tot  txi ic  W from extra-cell terminals using
term
txj j  i
carrier ic
 Pb
UL
 ic adj 
UL Uplink inter-carrier interference at
I inter – carrier  txi ic  term W
txj j
-----------------------------------
-
UL UL extra UL intra Total received interference at

I tot  txi ic  I tot
Tx
 txi ic  +  1 – F MUD   term  I tot  txi ic + I inter – carrier  txi W
UL
ic  transmitter on carrier ic
Total noise at transmitter on carrier

UL UL tx
N tot  txi ic  I tot  txi ic  + N0 W ic
(Uplink interference)
UL
 term  P b – DPDCH  ic 
-  G UL
-------------------------------------------------------------------------------------------------------- UL
Div  G p
E UL Tx UL Quality level at transmitter on a
Q tch  txi ic    ----b- N tot  txi ic  –  1 – F MUD    term  P b  ic 
UL
 N t UL None
traffic channel for carrier ic a
UL
 term  P b – DPDCH  ic 
-  G UL
Total noise: ---------------------------------------------------
UL
Div  G p
UL
N tot  txi ic 
224

UL
No HO: Q tch  txi ic 

UL UL
Softer HO: f rake efficiency  Q tch  txk ic 
tx k  ActiveSet
 samesite 
Soft, softer/soft HO (No MRC): Quality level at site using carrier ic

due to combination of all
Max  Q UL tch  tx k
UL
ic    G macro – diversity transmitters of the active set located
tx k  ActiveSet at the same site and taking into
UL
Q  ic  None account increasing of the quality due
Softer/soft HO (MRC): to macro-diversity (macro-diversity
gain).
Max  
 UL  UL

UL UL
txk ,txl  ActiveSet  f rake efficiency  Q tch  tx k ic  Q tch  tx l ic  In simulations G macro – diversity = 1 .
 
tx k  samesite  tx k 
tx  othersite
l
UL
 G macro – diversity
UL
Q  ic 
UL
G SHO --------------------------------------------------- None Soft handover gain on uplink
UL
Q tch  BestServer ic 
UL Required terminal power to achieve

req Q req
P term  ic  -  P term
----------------- W Eb/Nt target at transmitter on carrier
UL
Q  ic  ic
a. Calculation option may be selected in the Global parameters tab. The chosen option will be taken into account only in simulations. In point
4.3 Simulations
The simulation process consists of two steps:
1. Obtaining a realistic user distribution
Atoll generates a user distribution using a Monte-Carlo algorithm, which requires traffic maps and data as input. The
resulting user distribution complies with the traffic database and maps provided to the algorithm.
the simulation and on the network interferences. A user may be either active or inactive. Both active and inactive
users consume radio resources and create interference.
Then, Atoll randomly assigns a shadowing error to each user using the probability distribution that describes the
shadowing effect.
Finally, another random trial determines user positions in their respective traffic zone and whether they are indoors
or outdoors (according to the clutter weighting and the indoor ratio per clutter class defined for the traffic maps).
2. Power control simulation
4.3.1 Generating a Realistic User Distribution

During the simulation, a first random trial is performed to determine the number of users and their activity status. Four
activity status are modelled:
• Active UL: the user is active on UL and inactive on DL
• Active DL: the user is active on DL and inactive on UL
• Active UL+DL: the user is active on UL and on DL
• Inactive: the user is inactive on UL and on DL
225
The determination of the number of users and the activity status allocation depend on the type of traffic cartography used.
Atoll determines the total number of users attempting connection in each simulation
based on the Poisson distribution. This may lead to slight variations in the total numbers
of users in different simulations. To have the same total number of users in each
simulation of a group, add the following lines in the Atoll.ini file:
[Simulation]
RandomTotalUsers=0
4.3.1.1 Simulations Based on User Profile Traffic Maps

User profile environment based traffic maps: Each pixel of the map is assigned an environment class which contains a list of
user profiles with an associated mobility type and a given density (number of subscribers with the same profile per km²).
User profile traffic maps: Each polygon and line of the map is assigned a density of subscribers with given user profile and
mobility type. If the map is composed of points, each point is assigned a number of subscribers with given user profile and
mobility type.
The user profile models the behaviour of the different subscriber categories. Each user profile contains a list of services and
their associated parameters describing how these services are accessed by the subscriber.
From environment (or polygon) surface (S) and user profile density (D), a number of subscribers (X) per user profile is inferred.
X = SD
• When user profile traffic maps are composed of lines, the number of subscribers (X)
per user profile is calculated from the line length (L) and the user profile density (D)
(nb of subscribers per km) as follows: X = L  D
• The number of subscribers (X) is an input when a user profile traffic map is composed
of points.
For each behaviour described in a user profile, according to the service, frequency use and exchange volume, Atoll calculates
the probability for the user being active in uplink and in downlink at an instant t.
4.3.1.1.1 Circuit Switched Service (i)

User profile parameters for circuit switched services are:
• The used terminal (equipment used for the service (from the Terminals table)),
• The average number of calls per hour N call ,
• The average duration of a call (seconds) d .
The number of users and their distribution per activity status is determined as follows:
• Calculation of the service usage duration per hour ( p 0 : probability of a connection):
N call  d
p o = -------------------
-
3600
• Calculation of the number of users trying to access the service i ( n i ):
ni = X  p0
Next, we can take into account activity periods during the connection in order to determine the activity status of each user.
• Calculation of activity probabilities:
UL DL
Probability of being inactive on UL and DL: p inactive =  1 – f act    1 – f act 
UL DL
Probability of being active on UL only: p UL = f act   1 – f act 
DL UL
Probability of being active on DL only: p DL = f act   1 – f act 
UL DL
Probability of being active both on UL and DL: p UL + DL = f act  f act
UL DL
Where, f act and f act are respectively the UL and DL activity factors defined for the circuit switched service i.
• Calculation of number of users per activity status:
226
inactive
Number of inactive users on UL and DL: n i = n i  p inactive
Number of users active on UL and inactive on DL: n i  UL  = n i  p UL
Number of users active on DL and inactive on UL: n i  DL  = n i  p DL
Number of users active on UL and DL both: n i  UL + DL  = n i  p UL + DL
Therefore, a user when he is connected can have four different activity status: either active on both links, or inactive on both
links, or active on UL only, or active on DL only.
4.3.1.1.2 Packet Switched Service (j)

User profile parameters for packet switched services are:
• The used terminal (equipment used for the service (from the Terminals table)),
• The average number of packet sessions per hour N sess ,
• The volume (in kbytes) which is transferred on the downlink V DL and the uplink V UL during a session.
A packet session consists of several packet calls separated by a reading time. Each packet call is defined by its size and may be
divided in packets of fixed size (1500 Bytes) separated by an inter arrival time.
In Atoll, a packet session is described by following parameters:
UL
N packet –c all : Average number of packet calls on the uplink during a session,
DL
N packet –c all : Average number of packet calls on the downlink during a session,
UL
T packet – call : Average time (millisecond) between two packets calls on the uplink ,
DL
T packet – call : Average time (millisecond) between two packets calls on the downlink ,
UL
T packet : Average time (millisecond) between two packets on the uplink ,
DL
T packet : Average time (millisecond) between two packets on the downlink ,
UL
S packet : Packet size (Bytes) on uplink,
DL
S packet : Packet size (Bytes) on downlink.
Figure 4.1: Description of a Packet Session
• Calculation of the average packet call size (kBytes):
UL V UL DL V DL
S packet –c all = ---------------------------------------- and S packet –c all = ----------------------------------------
UL UL DL DL
N packet –c all  f eff N packet –c all  f eff
UL DL
Where f eff and f eff are the UL and DL efficiency factors defined for the packet switched service j.
227
UL DL
For packet (HSDPA) and packet (HSPA) services, f eff and f eff are set to 1.
• Calculation of the average number of packets per packet call:

UL DL
UL  S packet –c all   S packet – c all 
- + 1 and N DL
N packet = int  ------------------------------- packet = int  -------------------------------
- + 1
UL
 S packet  1024  S DL
packet  1024
1kBytes = 1024Bytes.
• Calculation of the average duration of inactivity within a packet call (s):

UL UL DL DL
UL  N packet – 1   T packet DL  N packet – 1   T packet
 D Inactivity  packet – call = ---------------------------------------------------------
- and  D Inactivity  packet – call = ---------------------------------------------------------
-
1000 1000
• Calculation of the average duration of inactivity in a session (s):
UL UL UL DL DL DL
 D Inactivity  session = N packet –c all   D Inactivity  packet – call and  D Inactivity  session = N packet –c all   D Inactivity  packet – call
• Calculation of the average duration of activity in a session (s):

UL UL DL DL
UL UL N packet  S packet  8 N packet  S packet  8
- and  D DL
 D Activity  session = N packet –c all  ------------------------------------------------
DL
Activity  session = N packet – c all  ------------------------------------------------
-
UL DL
TP Av  1000 TP Av  1000
UL DL
Where TP Av and TPAv are the uplink and downlink average requested throughputs defined for the service j.
Therefore, the average duration of a connection (in s) is:

UL UL UL DL DL DL
D Connection =  D Activity  session +  D Inactivity  session and D Connection =  D Activity  session +  D Inactivity  session
• Calculation of the service usage duration per hour (probability of a connection):
UL N sess UL DL N sess DL
p Connection = ------------  D Connection and p Connection = ------------  D Connection
3600 3600
• Calculation of the probability of being connected:
UL DL
p Connected = 1 –  1 – p Connection    1 – p Connection 
Therefore, the number of users who want to get the service j is:
n j = X  p Connected
As you can see on the picture above, we have to consider three possible cases when a user is connected:
• 1st case: At a given time, packets are downloaded and uploaded.

In this case, the probability of being connected is:
UL DL
UL + DL p Connection  p Connection
p Connected = ---------------------------------------------------------
-
p Connected
• 2nd case: At a given time, packet are uploaded (no packet is downloaded).
Here, the probability of being connected is:
UL DL
UL p Connection   1 – p Connection 
p Connected = -----------------------------------------------------------------------
-
p Connected
• 3rd case: At a given time, packet are downloaded (no packet is uploaded).
In this case, the probability of being connected is:
DL UL
DL p Connection   1 – p Connection 
p Connected = -----------------------------------------------------------------------
-
p Connected
228
Now, we have to take into account activity periods during the connection in order to determine the activity status of each
user.
• Calculation of the probability of being active:
UL DL
UL  D Activity  session DL  D Activity  session
f = -------------------------------------------------------------------------------------------- and f = --------------------------------------------------------------------------------------------
UL UL DL DL
  D Inactivity  session +  D Activity  session    D Inactivity  session +  D Activity  session 
Therefore, we have:
• 1st case: At a given time, packets are downloaded and uploaded.

The user can be active on UL and inactive on DL; this probability is:
1 UL DL UL + DL
p UL = f   1 – f   p Connected
The user can be active on DL and inactive on UL; this probability is:
1 DL UL UL + DL
p DL = f   1 – f   p Connected
The user can be active on both links; this probability is:

1 UL DL UL + DL
p UL + DL = f f  p Connected
The user can be inactive on both links; this probability is:

1 UL DL UL + DL
p inactive =  1 – f    1 – f   p Connected
• 2nd case: At a given time, packet are uploaded (no packet is downloaded).
The user can be active on UL and inactive on DL; this probability is:
2 UL UL
p UL = f  p Connected

2 UL UL
p inactive =  1 – f   p Connected
• 3rd case: At a given time, packet are downloaded (no packet is uploaded).
The user can be active on DL and inactive on UL; this probability is:
3 DL DL
p DL = f  p Connected

3 DL DL
p inactive =  1 – f   p Connected
• Calculation of number of users per activity status

inactive 1 2 3
Number of inactive users on UL and DL: n j = n j   p inactive + p inactive + p inactive 
1 2
Number of users active on UL and inactive on DL: n j  UL  = n j   p UL + p UL 
1 3
Number of users active on DL and inactive on UL: n j  DL  = n j   p DL + p DL 
1
Number of users active on UL and DL: n j  UL + DL  = n j  p UL + DL
Therefore, a user when he is connected can have four different activity status: either active on both links, or inactive on both
links, or active on UL only, or active on DL only.
The user distribution per service and the activity status distribution between the users
are average distributions. And the service and the activity status of each user are
randomly drawn in each simulation. Therefore, if you compute several simulations at
once, the average number of users per service and average numbers of inactive, active
on UL, active on DL and active on UL and DL users, respectively, will correspond to
calculated distributions. But if you check each simulation, the user distribution between
services as well as the activity status distribution between users is different in each of
them.
229
4.3.1.2 Simulations Based on Sector Traffic Maps

Sector traffic maps can be based on live traffic data from OMC (Operation and Maintenance Centre). Traffic is spread over the
best server coverage area of each transmitter and each coverage area is assigned either the throughputs in the uplink and in
the downlink or the number of users per activity status or the total number of users (including all activity statuses).
4.3.1.2.1 Throughputs in Uplink and Downlink

When selecting Throughputs in Uplink and Downlink, you can input the throughput demands in the uplink and downlink for
each sector and for each listed service.
Atoll calculates the number of users active in uplink and in downlink in the Txi cell using the service (NUL and NDL) as follows:
UL DL
Rt Rt
N UL = ----------
- and N DL = ----------
-
UL DL
TP Av TP Av
UL
Rt is the kbits per second transmitted in UL in the Txi cell to supply the service.
DL
Rt is the kbits per second transmitted in DL in the Txi cell to supply the service.
DL
TP Av is the downlink average requested throughput defined for the service,
UL
TP Av is the uplink average requested throughput defined for the service.
NUL and NDL values include:

• Users active in uplink and inactive in downlink (ni(UL)),
• Users active in downlink and inactive in uplink (ni(DL)),
• And users active in both links (ni(UL+DL)).
Atoll takes into account activity periods during the connection in order to determine the activity status of each user.
Activity probabilities are calculated as follows:
UL DL
Probability of being inactive in UL and DL: p inactive =  1 – f act    1 – f act 
UL DL
Probability of being active in UL only: p UL = f act   1 – f act 
DL UL
Probability of being active in DL only: p DL = f act   1 – f act 
UL DL
Probability of being active both in UL and DL: p UL + DL = f act  f act
UL DL
Where, f act and f act are respectively the UL and DL activity factors defined for the service i.
Then, Atoll calculates the number of users per activity status:

We have:
 p UL + p UL + DL    n j  UL  + n j  DL  + n j  UL + DL   = N UL
 p DL + p UL + DL    n j  UL  + n j  DL  + n j  UL + DL   = N DL
Therefore, we have:
N UL  p UL + DL N DL  p UL + DL
Number of users active in UL and DL both: n i  UL + DL  = min  --------------------------------- --------------------------------
p UL + p UL + DL p DL + p UL + DL
Number of users active in UL and inactive in DL: n i  UL  = N UL – n i  UL + DL 
Number of users active in DL and inactive in UL: n i  DL  = N DL – n i  UL + DL 
inactive  n j  UL  + n j  DL  + n j  UL + DL  
Number of inactive users in UL and DL: n i -  p inactive
= ------------------------------------------------------------------------------
1 – p inactive
Therefore, a connected user can have four different activity status: either active in both links, or inactive in both links, or active
in UL only, or active in DL only.
230
4.3.1.2.2 Total Number of Users (All Activity Statuses)

When selecting Total Number of Users (All Activity Statuses), you can input the number of connected users for each sector
and for each listed service ( n i ).
UL DL
Probability of being inactive in UL and DL: p inactive =  1 – f act    1 – f act 
UL DL
Probability of being active in UL only: p UL = f act   1 – f act 
DL UL
Probability of being active in DL only: p DL = f act   1 – f act 
UL DL
Probability of being active both in UL and DL: p UL + DL = f act  f act
UL DL
Where, f act and f act are respectively the UL and DL activity factors defined for the service i.

inactive
Number of inactive users in UL and DL: n i = n i  p inactive
Number of users active in UL and inactive in DL: n i  UL  = n i  p UL
Number of users active in DL and inactive in UL: n i  DL  = n i  p DL
Number of users active in UL and DL both: n i  UL + DL  = n i  p UL + DL
4.3.1.2.3 Number of Users per Activity Status

inactive
When selecting Number of Users per Activity Status, you can directly input the number of inactive users ( n i ), the
number of users active in the uplink ( n i  UL  ), in the downlink ( n i  DL  ) and in the uplink and downlink ( n i  UL + DL  ), for
each sector and for each service.
The activity status distribution between users is an average distribution. In fact, in each
simulation, the activity status of each user is randomly drawn. Therefore, if you compute
several simulations at once, average numbers of inactive, active on UL, active on DL and
active on UL and DL users correspond to the calculated distribution. But if you check each
simulation, the activity status distribution between users is different in each of them.
4.3.2 Power Control Simulation

The power control algorithm simulates the way a UMTS network regulates itself by using uplink and downlink power controls
in order to minimize interference and maximize capacity.
HSDPA users are linked to the A-DPCH radio bearer (an R99 radio bearer). Therefore, the network uses a A-DPCH power
control on UL and DL and then it performs fast link adaptation on DL in order to select an HSDPA radio bearer. For HSPA users,
the network first uses a E-DPCCH/A-DPCH power control on UL and DL, checks that there is an HSDPA connection on downlink
and then carries out noise rise scheduling in order to select an HSUPA radio bearer on uplink. Atoll simulates these network
regulation mechanisms with an iterative algorithm and calculates, for each user distribution, network parameters such as cell
power, mobile terminal power, active set and handoff status for each terminal. During each iteration of the algorithm, all the
users selected during the user distribution generation (1st step) attempt to connect one by one to network transmitters. The
process is repeated until the network is balanced, i.e., until the convergence criteria (on UL and DL) are satisfied.
231
Figure 4.2: UMTS HSPA Power Control Algorithm
As shown in Figure 4.2 on page 232, the simulation algorithm is divided in three parts. All users are evaluated by the R99 part
of the algorithm. HSDPA and HSPA users, unless they have been rejected during the R99 part of the algorithm, are then
evaluated by the HSDPA part of the algorithm. Finally, HSPA users, unless they have been rejected during the R99 or HSDPA
parts of the algorithm, are then evaluated by the HSUPA part of the algorithm.
The steps of this algorithm are detailed below.
4.3.2.1 Algorithm Initialization

The total power transmitted by the base station txi on the carrier ic m , P Tx  txi ic m  , is initialised to
P pilot  txi ic m  + P SCH  txi ic m  + P otherCCH  txi ic m  + P HSDPA  txi ic m  + P HSUPA  txi ic  .
UL intra UL extra
UL
Uplink powers received by the base station txi on carrier ic m , I tot  txi ic m  , I tot  txi ic m  and I inter – carrier  txi ic m 
are initialised to 0 W (i.e. no connected mobile).
232
UL
UL I tot  txi ic m 
Therefore, we have:  X R99  txi ic m   k = -------------------------------
- = 0
UL
N tot  txi ic m 
4.3.2.2 R99 Part of the Algorithm

The algorithm is detailed for any iteration k. Xk is the value of the X (variable) at the iteration k. In the algorithm, the
UL DL
thresholds, Q req and Q req , depend on the user mobility type and are defined in the R99 bearer selection table. All variables
are described in Definitions and formulas part. The bearer downgrading is not dealt with.
The algorithm applies to single frequency band networks and to multi-band networks.
For each mobile (Mb), Atoll only considers the cells (txi,ic) for which the pilot RSCP exceeds the minimum pilot RSCP:
P c  txi M b ic   RSCP min  txi ic  .
For each mobile Mb, we have the following steps:
Determination of Mb’s Best Serving Cell
For each transmitter txi containing Mb in its calculation area and working on a frequency band supported by the Mb’s terminal
).
   BTS  P c  txi M b ic 
Calculation of Q pilot  txi ic Mb  = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
k DL DL DL DL Term
P tot  txi ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0
If user selects “without Pilot”

   BTS  P c  txi M b ic 
Q pilot  txi ic Mb  = ------------------------------------------------------------------------------------------------------------------------------------------------------------
-
k
 DL DL DL DL 
 I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  
 Term 
 + N0 –  1 –     BTS  P c  txi M b ic  
Determination of the candidate cells, (txBS,ic).
For each carrier ic, selection of the transmitter with the highest Q pilot  txi M b ic  ,  tx BS ic   M i  .
k
Analysis of candidate cells, (txBS,ic).

For each pair (txBS,ic), calculation of the uplink load factor:
UL
UL I tot  tx BS ic  UL
 X R99  tx BS ic  k = -------------------------------
- + X
UL
N tot  tx BS ic 
UL
X corresponds to the load rise due to the mobile. For information on how this parameter is calculated, see "Admission
Control in the R99 Part" on page 276.
Rejection of bad candidate cells if the pilot is not received or if the uplink load factor is exceeded during the admission load
control (if simulation respects a loading factor constraint and Mb was not connected in previous iteration)
pilot
If Q pilot  tx BS M b ic   Q req  Mobility  M b   then (txBS,ic) is rejected by Mb
k
UL UL
If  X R99  tx BS ic   k  X max , then (txBS,ic) is rejected by Mb
Else
Keep (txBS,ic) as good candidate cell
If no good candidate cell has been selected, Mb has failed to be connected to the network and is rejected.
For each NodeB having candidate cells, determination of the best carrier, icBS, within the set of candidate cells of the
NodeB.
For MC-HSDPA and DB-MC-HSDPA users, this carrier is referred to as the "anchor" carrier.
If a given carrier is specified for the service requested by Mb
ic BS  M b  is the carrier specified for the service
Else the carrier selection mode defined for the site equipment is considered.
233
If carrier selection mode is “Min. UL Load Factor”

UL
ic BS  M b  is the carrier where we obtain the lowest  X R99  tx BS ic   k
Else if carrier selection mode is “Min. DL Total Power”

ic BS  M b  is the carrier where we obtain the lowest P tx  tx BS ic  k
Else if carrier selection mode is “Random”

ic BS  M b  is randomly selected
Else if carrier selection mode is "Sequential"

UL UL
ic BS  M b  is the first carrier where  X R99  tx BS ic   k  X max
Endif
Determination of the best serving cell, (txBS,icBS)
max
(tx BS,ic BS) k  M b  is the best serving cell ( BestCell k  M b  ) and its pilot quality is Q pilot  M b 
k
In the following lines, we will consider ic as the carrier used by the best serving cell
Selection of the second serving cell for DC-HSDPA users
MC-HSDPA and DB-MC-HSDPA users are processed as DC-HSDPA users.
If txBS supports multi-cell HSDPA and if it has several carriers, selection of the second carrier, ic2, among the adjacent carriers.
For each carrier adjacent to the best serving carrier, icp, calculation of Q pilot  tx BS ic p M b 
k
Selection of the carrier, ic2, with the highest Q pilot  tx BS ic p M b 

k
pilot
If Q pilot  tx BS ic 2 M b   Q req  Mobility  M b   then (txBS,ic2) is rejected by Mb
k
Else
Keep (txBS,ic2) as second serving cell
Active Set Determination
For each station txi containing Mb in its calculation area, using ic , and, if neighbours are used, neighbour of BestCell k  M b 
   BTS  P c  txi M b ic 
Calculation of Q pilot  txi M b ic  = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
DL DL DL DL Term
k
P tot  txi ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0

   BTS  P c  txi M b ic 
Q pilot  txi M b ic  = ----------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
k
 DL DL DL DL 
I
 intra  txi  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  
 Term 
 + N0 –  1 –     BTS  P c  txi M b ic  
Rejection of txi from the active set if difference with the best server is too high
max
If Q pilot  M b  – Q pilot  txi M b ic   AS_Th  BestCell k  M b   then txi is rejected
k k
Else txi is included in the Mb active set

Rejection of a station if the mobile active set is full
Station with the lowest Q pilot in the active set is rejected
k
EndFor
Uplink Power Control
R99 – req
Calculation of the terminal power required by Mb to obtain the R99 radio bearer: P term  M b ic  k
For each cell (txi,ic) of the Mb active set
234
Calculation of quality level on Mb traffic channel at (txi,ic), with the minimum power allowed on traffic channel for the Mb
service
req
UL P term – R99  M b ic  k – 1
P b – R99  txi M b ic  = ----------------------------------------------------
L T  txi M b 
UL UL UL
P b – DPDCH  txi M b ic  = P b – R99  txi M b ic    1 – r c 
UL UL UL
P b – DPCCH  txi M b ic  = P b – R99  txi M b ic   r c
UL UL UL
P b – R99  txi M b ic  = P b – DPCCH  txi M b ic  + P b – DPDCH  txi M b ic  if the user is active,
UL UL
P b – R99  txi M b ic  = P b – DPCCH  txi M b ic  if the user is inactive,
UL
UL  term  P b – DPDCH  txi M b ic  k
-  G UL
Q tch  txi M b ic  k = ---------------------------------------------------------------------------------------------------------------------------------------------
UL
p  Service  M b    G div
UL Tx UL
N tot  txi ic  –  1 – F MUD    term  P b – R99  txi M b ic  k – 1
If user selects "Total noise",

UL
UL  term  P b – DPDCH  txi M b ic  k UL UL
Q tch  txi M b ic  k = --------------------------------------------------------------------------
UL
 G p  Service  Mb    G div
End For
If (Mb is in not in handoff)
UL UL
Q k  M b  = Q tch  txi M b ic  k
Else if (Mi is in softer handoff)

UL UL UL
Q k  M b  = f rake efficiency  Q tch  txi M b ic  k
txi  ActiveSet
Else if (Mb is in soft, or softer/soft without MRC)
UL UL UL
Qk  Mb  = Max  Q tch  txi M b ic  k    G macro – diversity  2 links
txi  ActiveSet
Else if (Mb is in soft/soft)
UL UL UL
Qk  Mb  = Max  Q tch  txi M b ic  k    G macro – diversity  3 links
txi  ActiveSet
Else if (Mb is in softer/soft with MRC)
 
 UL 

UL UL UL UL
Qk  Mb  = Max  f rake efficiency  Q tch  ic  Q tch  ic    G macro – diversity  2 links
 other site 
 txi  ActiveSet 
 samesite 
End If
UL
req Q req  Service  M b  Mobility  M b  
-  P req
P term – R99  M b ic  k = ------------------------------------------------------------------------------------- term – R99  M b ic  k – 1
UL
Qk  Mb 
If compressed mode is operated,
Compressed mode is operated if Mi and Sj support compressed mode, and

Resulting CM – activation
• Either Q pilot  txi M b ic   Q pilot if the Ec/I0 Active option is selected,
k
CM – activation
• Or P c  txi M b ic   RSCP pilot if the RSCP Active option is selected.
235
UL UL
req Q req  Service  M b  Mobility  M b    Q req   Service  M b  Mobility  M b   
-  P req
P term – R99  M b ic  k = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ term – R99  M b ic  k – 1
UL
Qk  Mb 
req min req min

If P term – R99  M b ic  k  P term  M b  then P term – R99  M b ic  k = P term  txi M b 
req max
If P term – R99  M b ic  k  P term  M b  then Mb cannot select any cell and its active set is cleared
R99
If TP P – UL  M b   TP Max – UL  txi ic  then Mb cannot be connected
Endif
Downlink Power Control
If (mobile does not use a packet switched service that is inactive on the downlink)
For each cell (txi,ic) in Mb active set
Calculation of quality level on (txi,ic) traffic channel at Mb with the minimum power allowed on traffic channel for the Mb
service
min
DL P tch  Service  M b  
P b  txi M b ic  = -----------------------------------------------
DL
DL  BTS  P b  txi M b ic  k
-  G DL
Q tch  txi M b ic  k = --------------------------------------------------------------------------------------------------------------------------
DL
DL DL
N tot  ic  –  1 – F ortho    BTS  P b  txi M b ic  k – 1
If the user selects the option "Total noise"

DL
DL  BTS  P b  txi M b ic  k
-  G DL
Q tch  txi M b ic  k = --------------------------------------------------------
DL
DL
N tot  ic 
End For

DL DL DL
txi  ActiveSet
Do
For each cell (txi,ic) in Mb active set
Calculation of the required power for DL traffic channel between (txi,ic) and Mb:
DL
req Q req  Service  M b  Mobility  M b  
-  P min
P tch  txi M b ic  k = ------------------------------------------------------------------------------------- tch  Service  M b  
DL
Qk  Mb 
If compressed mode is operated.

DL DL
req Q req  Service  M b  Mobility  M b    Q req   Service  Mb  Mobility  M b   
-  P min
P tch  txi M b ic  k = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ tch  Service  M b  
DL
Qk  Mb 
Compressed mode is operated if Mi and Sj support compressed mode, and

• Either Q pilot  txi M b ic   Q pilot if the Ec/I0 Active option is selected,
k
CM – activation
• Or P c  txi M b ic   RSCP pilot if the RSCP Active option is selected.
req max max

If P tch  txi M b ic  k  P tch  Service  M b   then  txi ic  is set to P tch
DL max
Recalculation of a decreased Q req (a part of the required quality is managed by the cells set to P tch )
req
DL P tch  Service  M b  
P b  txi M b ic  = ----------------------------------------------
236
DL
DL  BTS  P b  txi M b ic 
-  G DL
Q tch  txi M b ic  k = -----------------------------------------------------------------------------------------------------------------
DL
DL DL
N tot  ic  –  1 – F ortho    BTS  P b  txi M b ic 
DL DL DL
If the user is inactive, then his contribution to interference in the calculation of N tot  ic  is P b  txi M b ic   r c .
EndFor

DL DL DL
txi  ActiveSet
DL DL
While Q k  M b   Q req  Service  M b  Mobility  M b   and Mb active set is not empty
R99
If TP P – DL  M b   TP Max – DL  txi ic  then Mb cannot be connected
Endif
Uplink and Downlink Interference Update
Update of interference on active mobiles only (old contributions of mobiles and stations are replaced by the new ones).
For each cell (txi,ic)
UL
Update of N tot  txi ic 
EndFor
For each mobile Mi
DL
Update of N tot  ic 
EndFor
EndFor
Control of Radio Resource Limits (OVSF Codes, Cell Power, Channel Elements, Iub Backhaul Throughput)

P tx  txi ic  DL
While ----------------------------k  %Powermax
P max
Rejection of the mobile with the lowest service priority starting from the last admitted
EndFor
Codes Codes
While N  txi ic  k  N max  txi ic 
EndFor
For each NodeB, Ni
CE – DL CE – DL
While N  N i  k  N max  Ni 
CE – UL CE – UL
While N  N i  k  N max  Ni 
EndFor
For each NodeB, Ni
Max
While TP Iub – DL  N I  k  TP Iub – DL  N I 
Max
While TP Iub – UL  N I  k  TPIub – UL  N I 
237
EndFor
Uplink Load Factor Control
UL UL
For each cell (txi,ic) with X R99  txi ic   X max
EndFor
UL UL
While at least one cell with X R99  txi ic   X max exists.
4.3.2.3 HSDPA Part of the Algorithm

HSDPA BE, HSDPA VBR, HSPA BE and HSPA VBR service users active on DL as well as all HSPA CBR service users (i.e., active and
inactive), unless they have been rejected during the R99 part of the algorithm, are then evaluated by the HSDPA part of the
algorithm.
4.3.2.3.1 HSDPA Power Allocation

The total transmitted power of the cell ( P tx  ic  ) is the sum of the transmitted R99 power, the HSUPA power and the
transmitted HSDPA power.
P tx  ic  = P tx – R99  ic  + P tx –H SDPA  ic  + P HSUPA  ic 
• In case of a static HSDPA power allocation strategy, Atoll checks in the simulation that:
DL
P tx  ic   P max  ic   %Power max
where:
DL
%Power max is the maximum DL load allowed.
Therefore, if the maximum DL load is set to 100%, we have:

P tx  ic   P max  ic 
• In case of dynamic HSDPA power allocation strategy, Atoll checks in the simulation that:
DL
P tx – R99  ic  + P HSUPA  ic   P max  ic   %Power max
And it calculates the available HSDPA power as follows:

P HSDPA  ic  = P max  ic  – P Headroom  ic  – P tx – R99  ic  – P HSUPA  ic 
4.3.2.3.2 Number of HS-SCCH Channels and Maximum Number of HSDPA Bearer Users
The number of HS-SCCH channels ( n HS – SCCH ) is the maximum number of HS-SCCH channels that the cell can manage. This
parameter is used to manage the number of BE and VBR service users simultaneously connected to an HSDPA bearer. This
parameter is not taken into account for CBR service users as HS-SCCH-less operation (i.e., HS-DSCH transmissions without any
accompanying HS-SCCH) is performed.
Each HSDPA BE, HSDPA VBR, HSPA BE and HSPA VBR service user consumes one HS-SCCH channel. Therefore, at a time (over
a transmission time interval), the number of these users connected to an HSDPA bearer cannot exceed the number of HS-
SCCH channels per cell.
The maximum number of HSDPA bearer users ( n max ) corresponds to the maximum number of HSDPA bearer users that the
cell can support. Here, all HSDPA bearer users, i.e., HSDPA BE, HSDPA VBR, HSPA BE and HSPA VBR and HSPA CBR service users,
are taken into consideration.
Let us assume there are 30 users in the cell:
• 10 HSPA CBR service users with any activity status.
• 2 HSDPA VBR service users active on DL.
• 18 HSDPA BE and HSPA BE service users active on DL.
All users are connected to the A-DCH R99 bearer. Finally, the number of HS-SCCH channels and the maximum number of
HSDPA bearer users respectively equal 4 and 25.
The scheduler manages the maximum number of users within each cell. CBR service users have the highest priority and are
processed first, in the order established during the generation of the user distribution. After processing the CBR service users,
Atoll processes the remaining HSDPA bearer users (i.e., HSDPA VBR, HSPA VBR, HSDPA BE and HSPA BE service users). VBR
238
service users have the highest priority and are managed before BE service users. For each type of service, the scheduler ranks
the users according to the selected scheduling technique. Users are treated as described in the figure below.
Figure 4.3: Connection status of HSDPA bearer users
• All CBR service users may be served if there are enough HSDPA power, Iub backhaul throughput and OVSF codes
available in order for them to obtain the lowest HSDPA bearer that provides a peak RLC throughput higher or equal to
the minimum throughput demand defined for the service. In this case, they will be connected. Else, they will be
rejected.
• The two VBR service users may be simultaneously served if there are enough HSDPA power, Iub backhaul throughput
and OVSF codes available in order for them to obtain an HSDPA bearer that provides a peak RLC throughput higher or
equal to the minimum throughput demand defined for the service. In this case, they will be connected. Else, they will
be rejected.
• Then, among the BE service users:
• The first two users may be simultaneously served if there are enough HSDPA power, Iub backhaul throughput and
OVSF codes available in order for them to obtain an HSDPA bearer. In this case, they will be connected. Else, they
will be delayed.
• The next eleven ones will be delayed since there are no longer HS-SCCH channels available. Their connection
status will be "HS-SCCH Channels Saturation".
• Finally, the last five users will be rejected because the maximum number of HSDPA bearer users has been fixed to
25. Their connection status will be "HSDPA Scheduler Saturation".
4.3.2.3.3 HSDPA Bearer Allocation Process

The HSDPA bearer allocation process depends on the type of service requested by the user. As explained before, CBR service
users have the highest priority and are processed first, in the order established during the generation of the user distribution.
After processing the CBR service users, the scheduler ranks the remaining users (i.e., VBR and BE service users) and shares the
cell radio resources between them. VBR service users have the highest priority and are managed before BE service users.
CBR Service Users
Let us focus on the ten CBR service users mentioned in the example of the previous paragraph "Number of HS-SCCH Channels
and Maximum Number of HSDPA Bearer Users" on page 238. Fast link adaptation is carried out on these users in order to
determine if they can obtain an HSDPA bearer that provides a peak RLC throughput higher or equal to the service minimum
throughput demand. As HS-SCCH less operation is performed, only HSDPA bearers using the QPSK modulation and two HS-
PDSCH channels at the maximum can be selected and allocated to the users. The users are processed in the order established
during the generation of the user distribution and the cell’s available HSDPA power is shared between them as explained
below. Several CBR service users can share the same HSDPA bearer. Then, Atoll calculates the HSDPA bearer consumption ( C
in %) for each user and takes into account this parameter when it determines the resources consumed by the user (i.e., the
HSDPA power used, the number of OVSF codes and the Iub backhaul throughput).
In the bearer allocation process shown below, the 10 CBR service users are represented by Mj, with j = 1 to 10. And, the initial
values of their respective HSDPA powers is 0, i.e. PHSDPA(B(MX)) = 0, where X = 0 to 10. These power values are assigned one
by one by the scheduler, so that with their allocated values, looped back to the starting point, are used in successive steps.
239
For the user, Mj, with j varying from 1 to 10:
j 1 Sufficient HS-SCCH No
PHSDPA   (PHSDPA (M X )) served power to reach the Mj is rejected
X 0 minimum quality
threshold?
Yes
Enough
16-bit OVSF codes
No
available to support the Mj is rejected
lowest HSDPA bearer
allocated?
Yes
Sufficient Iub
backhaul throughput
No
to support the lowest Mj is rejected
HSDPA bearer
allocated?
Yes
Sufficient HSDPA
power to obtain No
Mj is rejected
the lowest HSDPA
bearer allocated?
Yes
Determination of the Best HSDPA Bearer BB(Mj)
Cell and UE both capable Yes BB(Mj) selected

of supporting BB(Mj)? B(Mj) = BB(Mj)
No
Bearer Downgrading
B(Mj)
RLC Peak Rate of No

B(Mj) > Mj Min Throughput Mj is rejected
Demand?
Yes
Allocation of Min Throughput Demand to Mj

Mj connected with B(Mj)
(PHSDPA(Mj))served=PHS-PDSCH(B(Mj)) x C(B(Mj))
Update of Available Radio Resources
No Mj = M10?
Yes
Resource allocation for Variable Bit Rate and

Best Effort service users
Figure 4.4: HSDPA Bearer Allocation Process for CBR Service Users
VBR Service Users
After processing the CBR service users, the scheduler shares the cell’s remaining resources between HSDPA and HSPA VBR
service users. Let us focus on the two HSDPA - VBR service users mentioned in the example of the previous paragraph,
"Number of HS-SCCH Channels and Maximum Number of HSDPA Bearer Users" on page 238. A new fast link adaptation is
carried out on these users in order to determine if they can obtain an HSDPA bearer that provides a peak RLC throughput
higher or equal to the service minimum throughput demand. They are processed in the order defined by the scheduler and
the cell’s HSDPA power available after all CBR service users have been served is shared between them as explained below.
In the bearer allocation process shown below, the 2 VBR service users are represented by Mj, with j = 1 to 2. And, the initial
240
j 1 Sufficient HS-SCCH No
PHSDPA   (PHSDPA (M X )) served power to reach the Mj is rejected
X 0 minimum quality
threshold?
Yes
Enough
16-bit OVSF codes
No
available to support the Mj is rejected
lowest HSDPA bearer
allocated?
Yes
Sufficient Iub
backhaul throughput
No
to support the lowest Mj is rejected
HSDPA bearer
allocated?
Yes
Sufficient HSDPA
power to obtain No
Mj is rejected
the lowest HSDPA
bearer allocated?
Yes
Determination of the Best HSDPA Bearer BB(Mj)
Bearer Downgrading B(Mj) until:

1. Cell and UE both capable of supporting B(Mj)
And
2. RLC Peak Rate of B(Mj) > Mj Min Throughput
Demand
RLC Peak Rate of No

B(Mj) > Mj Min Throughput Mj is rejected
Demand?
Yes
Mj connected with B(Mj)

(PHSDPA(Mj))served=PHS-PDSCH(B(Mj)) +nHS-SCCHxPHS-SCCH(Mj)
No Mj = M2?
Yes
Resource allocation for Best Effort service users

Figure 4.5: HSDPA Bearer Allocation Process for VBR Service Users
BE Service Users
After processing the VBR service users, the scheduler shares the cell’s remaining resources between BE service users. Let us
focus on the HSDPA and HSPA BE service users, especially on the first four users mentioned in the example of the previous
paragraph, "Number of HS-SCCH Channels and Maximum Number of HSDPA Bearer Users" on page 238. A new fast link
adaptation is carried out on these users in order to determine if they can obtain an HSDPA bearer. They are processed in the
order defined by the scheduler and the cell’s HSDPA power available after all CBR and VBR service users have been served is
shared between them as explained below.
In the bearer allocation process shown below, the 4 BE service users are represented by Mj, with j = 1 to 4. And, the initial
241
Figure 4.6: HSDPA Bearer Allocation Process for BE Service Users
4.3.2.3.4 Fast Link Adaptation Modelling

Fast link adaptation (or Adaptive Modulation and Coding) is used in HSDPA. The power on the HS-DSCH channel is transmitted
at a constant power while the modulation, the coding and the number of codes are changed to adapt to the radio conditions
variations. Based on the reported channel quality indicator (CQI), the node-B may change every 2ms the modulation (QPSK,
16QAM, 64QAM), the coding and the number of codes during a communication.
Atoll calculates for each user either the best pilot quality (CPICH Ec/Nt) or the best HS-PDSCH quality (HS-PDSCH Ec/Nt); this
depends on the option selected in Global parameters (HSDPA part): CQI based on CPICH quality or CQI based on HS-PDSCH
quality (CQI means channel quality indicator). Then, it determines the HS-PDSCH CQI, calculates the best bearer that can be
used and selects the suitable bearer so as to comply with cell and terminal user equipment HSDPA capabilities. Once the
bearer selected, Atoll finds the highest downlink throughput that can be provided to the user and may calculate the
application throughput.
242
CQI Based on CPICH Quality
When the option “CQI based on CPICH quality” is selected, Atoll proceeds as follows.
1. CPICH Quality Calculation
Let us assume the following notation:  ------  ic 

Ec
corresponds to the CPICH quality.
Nt pilot
Two options, available in Global parameters, may be used to calculate Nt: option Without useful signal or option Total noise.
Therefore, we have:
 BTS    P c  ic 
 -----
Ec-
 ic  i
= -----------------------------------------
- for the total noise option,
 Nt  pilot DL
N tot  ic 
And
 BTS    P c  ic 
 -----
Ec-
 ic  i
= --------------------------------------------------------------------------------- for the without useful signal option.
 Nt  pilot DL
N tot  ic  –  1 –     BTS  P c  ic 
i
With
DL DL DL DL DL term
N tot  ic  = I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0
DL  DL P SCH  ic   DL P SCH  ic 

I intra  ic  = P tot  ic  +  BTS   1 – F MUD    1 –     P tot  ic  – -------------------
- –  BTS   P tot  ic  – -------------------
DL term
-
LT LT 
txi  txi   txi 

DL DL
I extra  ic  = P tot  ic 
txj j  i
 Ptot  icadj 
DL
DL  j
I inter – carrier  ic  = txj
------------------------------------
icadj is a carrier adjacent to ic.
RF  ic ic adj  is the interference reduction factor, defined between ic and icadj and set to a value different from 0.
DL
I inter – techno log y  ic  is the inter-technology interference at the receiver on ic.
Tx

DL
I inter – techno log y  ic  = --------------------------------------
Tx Tx m
ni
i
th
ic i is the i interfering carrier of an external transmitter
Tx m
ICPic  ic is the inter-technology Channel Protection between the signal transmitted by Tx and received by m assuming the
i
frequency gap between ic i (external network) and ic .
P pilot  ic 
P c  ic  = ---------------------
-
i LT
i
L path  L Tx  L term  L body  L indoor  E Shadowing 3

L T = -------------------------------------------------------------------------------------------------------------------
- ( )
G Tx  G term
term
 BTS ,  and N 0 are defined in "Inputs" on page 215.
3. In the HSDPA coverage prediction, L T is calculated as follows:

L T = ------------------------------------------------------------------------------------------------------------------------------------
-)
G Tx  G term
243
Atoll performs intra-cell interference computations based on the total power. You can
instruct Atoll to use maximum power by adding the following lines in the Atoll.ini file:
[CDMA]
PmaxInIntraItf = 1
In this case, Atoll considers the following formula:
P max  ic  P max  ic  – P SCH  ic 

+  BTS  1 – F MUD  1 –    ----------------------------------------------
-
DL term
I intra  ic  = -------------------- 
LT LT
P max  ic  – P SCH  ic 
– BTS   ----------------------------------------------
-
 LT 
2. CPICH CQI Determination

Let us assume the following notation:  CQI  pilot corresponds to the CPICH CQI.  CQI  pilot is read in the table
 CQI  pilot = f   ------  ic   . This table is defined for the terminal reception equipment and the selected mobility.
Ec
Nt pilot

3. HS-PDSCH Quality Calculation

Atoll proceeds as follows:
1st step: Atoll calculates the HS-SCCH power ( P HS – SCCH ).
P HS – SCCH  ic  is the HS-SCCH power on carrier ic. It is either fixed by the user (when the option “HS-SCCH Power Dynamic
Allocation”in the cell property dialogue is unchecked) or dynamically calculated (when the option “HS-SCCH Power Dynamic
Allocation” is selected).
req
In this case, the HS-SCCH power is controlled so as to reach the required HS-SCCH Ec/Nt (noted  ------  ic 
Ec
). It is
Nt HS – SCCH
specified in mobility properties.
We have:
 BTS  P c  ic 
 -----
Ec-
 ic  i
= -------------------------------
 Nt  HS – SCCH DL
N tot  ic 
And
 BTS  P c  ic 
 -----
Ec-
 ic  = i
----------------------------------------------------------------------------------------------------------------------------
- for the without useful signal option.
 Nt  HS – SCCH DL term
N tot  ic  –  1 – F ortho    1 – F MUD    BTS  P c  ic 
i
With
DL DL DL DL DL term
 DL   DL 
DL P SCH  ic  P SCH  ic 
I intra  ic  = P tot  ic  +  BTS   1 – F MUD    1 – F ortho    P tot  ic  – -------------------
- –  BTS   P tot  ic  – -------------------
DL term
-
LT LT

DL DL
txj j  i
DL
DL  j
------------------------------------
DL
244
Tx

DL
Tx Tx m
n
i i
th
Tx m
i
P HS – SCCH  ic 
P c  ic  = -------------------------------
-
i LT
i
and
L T = -------------------------------------------------------------------------------------------------------------------
- ( )
G Tx  G term
term term
 BTS , F ortho , F MUD and N 0 are defined in "Inputs" on page 215.
Therefore,
req
  -----
Ec-
 ic   N tot  ic 
DL
  Nt HS – SCCH 
P HS – SCCH  ic  =  ------------------------------------------------------------------  L T for the total noise option,
  BTS  i
 
And
req
  -----
Ec-
   DL 
 HS – SCCH  N tot  ic 
ic
  Nt 
P HS – SCCH  ic  =  -------------------------------------------------------------------------------------------------------------------------------------------
req
-   L T for the without useful signal option.
  1 + 1 – F    1 – F
term
   -----
Ec-
 ic    i
 BTS  ortho MUD  Nt  HS – SCCH 
2nd step: Atoll calculates the HS-PDSCH power ( P HS – PDSCH ).
P HSDPA  ic  is the power available for HSDPA on the carrier ic. This parameter is either a simulation output, or a user-defined
cell input.
P HSDPA  ic  = P HS – PDSCH  ic  + n HS – SCCH  P HS – SCCH  ic 
Therefore, we have:
P HS – PDSCH  ic  = P HSDPA  ic  – n HS – SCCH  P HS – SCCH  ic 
n HS – SCCH is the number of HS-SCCH channels.
3rd step: Then, Atoll evaluates the HS-PDSCH quality
Let us assume the following notation:  ------  ic 

Ec
corresponds to the HS-PDSCH quality.
 Nt  HS – PDSCH
We have:
 BTS  P c  ic 
 -----
Ec-
 ic  i
= -------------------------------
 Nt  HS – PDSCH DL
N tot  ic 
And
 BTS  P c  ic 
 -----
Ec- 
 Nt  ic  HS – PDSCH = ----------------------------------------------------------------------------------------------------------------------------
i
P c  ic 
DL term i
N tot  ic  –  1 – F ortho    1 – F MUD    BTS  ---------------
n

L T = ------------------------------------------------------------------------------------------------------------------------------------
-)
G Tx  G term
245
Here, Atoll works on the assumption that five HS-PDSCH channels are used (n=5).
With
DL DL DL DL DL term
 DL   DL 
I intra  ic  = P tot  ic  +  BTS   1 – F MUD    1 – F ortho    P tot  ic  – -------------------
- –  BTS   P tot  ic  – -------------------
DL term
-
 LT   LT 

DL DL
I extra  ic   = P tot  ic 
txj j  i
DL
DL  j
------------------------------------
DL
Tx

DL
Tx Tx m
ni
i
th
Tx m
ICP ic  ic is the inter-technology Channel Protection between the signal transmitted by Tx and received by m assuming the
i
P HS – PDSCH  ic 
P c  ic  = ----------------------------------
-
i LT
i
And
L T = -------------------------------------------------------------------------------------------------------------------
- ( )
G Tx  G term
term term
[CDMA]
PmaxInIntraItf = 1
P max  ic  P max  ic  – P SCH  ic  P max  ic  – P SCH  ic 

+  BTS  1 – F MUD  1 –    ----------------------------------------------- –  BTS   -----------------------------------------------
DL term
I intra  ic  = --------------------
LT LT LT
4. HS-PDSCH CQI Determination

The best bearer that can be used depends on the HS-PDSCH CQI. Let us assume the following notation:  CQI  HS – PDSCH
corresponds to the HS-PDSCH CQI. Atoll calculates  CQI  HS – PDSCH as follows:
 CQI  HS – PDSCH =  CQI  pilot – P pilot + P HS – PDSCH

L T = ------------------------------------------------------------------------------------------------------------------------------------
-)
G Tx  G term
246
5. HSDPA Bearer Selection

Atoll selects the HSDPA bearer associated to this CQI (in the table Best Bearer=f(HS-PDSCH CQI) defined for the terminal
reception equipment and the user mobility) and compatible with the user equipment and cell capabilities.
HSDPA bearers can be classified into two categories:
• HSDPA bearers using QPSK and 16QAM modulations: They can be selected for all users connected to HSPA and HSPA+
capable cells. The number of HS-PDSCH channels required by the bearer must not exceed the maximum number of
HS-PDSCH codes available for the cell.
For VBR service users, the selected HSDPA bearer must provide a peak RLC throughput between the minimum and the
maximum throughput demands defined for the service.
For CBR service users, HS-SCCH-less operation (i.e., HS-DSCH transmissions without any accompanying HS-SCCH) is
performed. In this case, the UE is not informed about the transmission format and has to revert to blind decoding of
the transport format used on the HS-DSCH. Complexity of blind detections in the UE is decreased by limiting the
transmission formats that can be used (i.e., the HSDPA bearers available). Therefore, only HSDPA bearers using the
QPSK modulation and two HS-PDSCH channels at the maximum can be selected and allocated to these users.
Additionally, the selected HSDPA bearer must provide a peak RLC throughput higher or equal to the minimum
throughput demand defined for the service.
• HSDPA bearers using 64QAM modulation (improvement introduced by the release 7 of the 3GPP UTRA specifications,
referred to as HSPA+): These HSDPA bearers can be allocated to VBR and BE service users connected to cells with
HSPA+ capabilities only. The number of HS-PDSCH channels required by the bearer must not exceed the maximum
number of HS-PDSCH codes available for the cell. For VBR service users, the selected HSDPA must provide a peak RLC
throughput between the minimum and the maximum throughput demands defined for the service. These HSDPA
bearers cannot be allocated to CBR service users.
Atoll considers an HSDPA bearer as compatible with the user equipment if:
• The transport block size does not exceed the maximum transport block size supported by the user equipment.
• The number of HS-PDSCH channels required by the bearer does not exceed the maximum number of HS-PDSCH
channels that the terminal can use.
• The modulation is supported by the user equipment.
When there are several HSDPA bearers compatible, Atoll selects the HSDPA bearer that provides the highest RLC peak
throughput. When several HSDPA bearers can supply the same RLC peak throughput, Atoll chooses the HSDPA bearer with
the highest modulation scheme. Finally, if no HSDPA bearer is compatible, Atoll allocates a lower HSDPA bearer compatible
with the user equipment and cell capabilities which needs fewer resources.
Let’s consider the following examples.
Example1: One HSDPA BE service user with category 13 user equipment and a 50km/h mobility.
The user equipment capabilities are:
• Maximum transport block size: 35280 bits
• Maximum number of HS-PDSCH channels: 15
• Highest modulation supported: 64QAM
• MIMO Support: No
Figure 4.7: HSDPA UE Categories Table
The cell to which the user is connected supports HSPA+ functionalities (i.e. 64QAM modulation in the DL and MIMO systems)
and the maximum number of HS-PDSCH channels is 15.
247
1st case: The CQI experienced by the user equals 26. Therefore, Atoll can choose between two HSDPA bearers, the bearer
indexes 26 and 31.
Characteristics of the bearer index 26 are:
• Transport block size: 17237 bits
• Number of HS-PDSCH channels used: 12
• 16QAM modulation is used
• Peak RLC Throughput: 8.32 Mb/s
Both HSDPA bearers are compatible with the user equipment and cell capabilities. Atoll selects the HSDPA bearer that
provides the highest RLC peak throughput, i.e. the bearer index 26.
Figure 4.8: HSDPA Radio Bearers Table
2nd case: The CQI experienced by the user equals 27. Therefore, Atoll can choose between two HSDPA bearers, the bearer
indexes 27 and 32.
Both HSDPA bearers are compatible with the user equipment and cell capabilities and the peak RLC throughput they provide
is the same. Atoll selects the HSDPA bearer using the highest modulation scheme, i.e. the bearer index 32.
Example 2: One HSDPA BE user experiencing a CQI of 26.
Therefore, Atoll can choose between two HSDPA bearers, the bearer indexes 26 and 31.
248

1st case: The user equipment category is 9. The cell to which the user is connected supports HSPA+ functionalities (i.e. 64QAM
modulation in the DL and MIMO systems) and the maximum number of HS-PDSCH channels is 15.
The user equipment characteristics are the following:
The bearer index 31 cannot be selected because it requires a modulation scheme not supported by the terminal. Only the
bearer index 26 is compatible with the user equipment capabilities. Atoll selects it.
2nd case: The user equipment category is 8. The cell to which the user is connected supports HSPA+ functionalities (i.e. 64QAM
modulation in the DL and MIMO systems) and the maximum number of HS-PDSCH channels is 15.
The user equipment characteristics are the following:
Here, none of HSDPA bearers are compatible with the user equipment capabilities.
The bearer index 31 cannot be selected because it requires a modulation scheme not supported by the terminal. With the
bearer index 26, the number of HS-PDSCH channels (12) exceeds the maximum number of HS-PDSCH channels the terminal
can use (10), and the transport block size (17237 bits) exceeds the maximum transport block size (14411 bits) the terminal can
carried.
In the HSDPA Radio Bearer table, Atoll selects a lower HSDPA bearer compatible with cell and UE category capabilities. It
selects the bearer index 25.
• The number of HS-PDSCH channels (10) does not exceed the maximum number of HS-PDSCH channels the terminal
can use (10) and the maximum number of HS-PDSCH channels available at the cell level (15),
• The transport block size (14411 bits) does not exceed the maximum transport block size (14411 bits) the terminal can
carried.
• 16QAM modulation is supported by the terminal and the cell.
3rd case: The user equipment category is 13. The cell to which the user is connected supports HSPA functionalities and the
maximum number of HS-PDSCH channels is 15.
The user equipment capabilities are:
• Maximum number of HS-PDSCH channels:15
The bearer index 31 cannot be selected because it requires a modulation scheme not supported by the cell. On the other hand,
the bearer index 26 is compatible with cell and UE category capabilities. Therefore, it is allocated.
6. HS-PDSCH Quality Update
Once the bearer selected, Atoll exactly knows the number of HS-PDSCH channels. Therefore, when the method “Without
useful signal” is used, it may recalculate the HS-PDSCH quality with the real number of HS-PDSCH channels (A default value
(5) was taken into account in the first HS-PDSCH quality calculation).
CQI Based on HS-PDSCH Quality
When the option “CQI based on HS-PDSCH quality” is selected, Atoll proceeds as follows.
1. HS-PDSCH Quality Calculation
1st step: Atoll calculates the HS-SCCH power ( P HS – SCCH ).
249
P HS – SCCH  ic  is the HS-SCCH power on carrier ic. It is either fixed by the user (when the option “HS-SCCH Power Dynamic
Allocation”in the cell property dialogue is unchecked) or dynamically calculated (when the option “HS-SCCH Power Dynamic
Allocation” is selected).
req
In this case, the HS-SCCH power is controlled so as to reach the required HS-SCCH Ec/Nt (noted  ------  ic 
Ec
). It is
Nt HS – SCCH
specified in mobility properties.
We have:
 BTS  P c  ic 
 -----
Ec-
 ic  i
= -------------------------------
N tot  ic 
And
 BTS  P c  ic 
 -----
Ec-
 ic  = i
----------------------------------------------------------------------------------------------------------------------------
 Nt  HS – SCCH DL term
i
With
DL DL DL DL DL term
 DL   DL 
- –  BTS   P tot  ic  – -------------------
DL term
-
 LT   LT 

DL DL
txj j  i
DL
DL  j
------------------------------------
DL
Tx

DL
Tx Tx m
ni
i
th
Tx m
i
P c  ic  = -------------------------------
-
i LT
i
And
L T = -------------------------------------------------------------------------------------------------------------------
- ( )
G Tx  G term
term term
Therefore,

L T = ------------------------------------------------------------------------------------------------------------------------------------
-)
G Tx  G term
250
req
  -----
Ec-
  DL 
 HS – SCCH  N tot  ic 
ic
  Nt
P HS – SCCH  ic  =  ------------------------------------------------------------------  L T for the total noise option,
  BTS  i
 
And
req
  -----
Ec-  DL 
  Nt  ic  HS – SCCH  N tot  ic  
P HS – SCCH  ic  =  -------------------------------------------------------------------------------------------------------------------------------------------
req
-  L T for the without useful signal option.
  1 + 1 – F term  Ec   i
 BTS  ortho    1 – F MUD    ------  ic   
Nt HS – SCCH
2nd step: Atoll calculates the HS-PDSCH power ( P HS – PDSCH )
P HSDPA  ic  is the power available for HSDPA on the carrier ic. This parameter is either a simulation output, or a user-defined
cell input.
Therefore, we have:
n HS – SCCH is the number of HS-SCCH channels.
3rd step: Then, Atoll evaluates the HS-PDSCH quality

Ec
We have:
 BTS  P c  ic 
 -----
Ec- 
 Nt  ic  HS – PDSCH = -------------------------------
i
DL
N tot  ic 
And
 BTS  P c  ic 
 -----
Ec-
 ic  i
= ----------------------------------------------------------------------------------------------------------------------------
 Nt  HS – PDSCH P c  ic 
DL term i
N tot  ic  –  1 – F ortho    1 – F MUD    BTS  ---------------
n
Here, Atoll works on the assumption that five HS-PDSCH channels are used (n=5). Then, it calculates the HS-PDSCH CQI and
the bearer to be used. Once the bearer selected, Atoll exactly knows the number of HS-PDSCH channels and recalculates the
HS-PDSCH quality with the real number of HS-PDSCH channels.
With
DL DL DL DL DL term
 DL   DL 
I intra  ic  = P tot  ic  +  BTS   1 – F MUD    1 – F ortho    P tot  ic  – -------------------
- –  BTS   P tot  ic  – -------------------
DL term
-
LT LT

DL DL
I extra  ic   = P tot  ic 
txj j  i
DL
DL  j
------------------------------------
DL
251
Tx

DL
Tx Tx m
n
i i
th
Tx m
i
P HS – PDSCH  ic 
P c  ic  = ----------------------------------
-
i LT
i
And
L T = -------------------------------------------------------------------------------------------------------------------
- ( )
G Tx  G term
term term
[CDMA]
PmaxInIntraItf = 1
P max  ic  P max  ic  – P SCH  ic  P max  ic  – P SCH  ic 

+  BTS  1 – F MUD  1 –    ----------------------------------------------
- –  BTS   ----------------------------------------------
-
DL term
I intra  ic  = --------------------
LT  LT   LT 
2. HS-PDSCH CQI Determination

Let us assume the following notation:  CQI  HS – PDSCH corresponds to the HS-PDSCH CQI.  CQI  HS – PDSCH is read in the table
 CQI  HS – PDSCH = f   ------  ic   . This table is defined for the terminal reception equipment and the specified
Ec
  Nt  HS – PDSCH
mobility.
3. HSDPA Bearer Selection
The bearer is selected as described in "HSDPA Bearer Selection" on page 246.
4.3.2.3.5 MIMO Modelling

MIMO - Transmit Diversity
If the user is connected to a cell that supports HSPA+ with transmit diversity and if he has a MIMO-capable terminal (i.e., a
terminal with an HSDPA UE category supporting MIMO), he will benefit from downlink diversity gain on the HS-PDSCH Ec/Nt.
 -----
Ec-
 ic  =  ------  ic 
Ec DL DL
+ G TD + G TD in dB
 Nt  HS – PDSCH  Nt  HS – PDSCH
Where
DL
G TD is the downlink transmit diversity gain (in dB) corresponding to the numbers of transmission and reception antenna ports
(respectively defined in the transmitter and terminal properties).
DL
G TD is the additional diversity gain in downlink (in dB). It is defined for the clutter class of the user.

L T = ------------------------------------------------------------------------------------------------------------------------------------
-)
G Tx  G term
252
MIMO - Spatial Multiplexing
If the user is connected to a cell that supports HSPA+ with spatial multiplexing and if he has a MIMO-capable terminal (i.e., a
terminal with an HSDPA UE category supporting MIMO), he will benefit from the spatial multiplexing gain in its peak RLC
throughput.
In this case, the peak RLC throughput obtained by the user is the following:
DL DL Max
TP P – R LC = TP P –R LC  Index HSDPABearer    1 + f SM – Gain   G SM – 1  
Where
DL
TP P – R LC  Index HSDPABearer  is the peak RLC throughput that the selected HSDPA bearer ( Index HSDPABearer ) can provide in the
cell (Txi, ic). It is read in the HSDPA Radio Bearer table.
Max
G SM is the maximum spatial multiplexing gain (in dB) for a given number of transmission and reception antennas
(respectively defined in the transmitter and terminal properties).
f SM – Gain is the spatial multiplexing gain factor defined for the clutter
4.3.2.3.6 Scheduling Algorithms

The scheduler manages the maximum number of users within each cell. CBR service users have the highest priority and are
processed first, in the order established during the generation of the user distribution. After processing the CBR service users,
the scheduler processes the remaining users (i.e., VBR and BE service users). VBR service users have the highest priority and
are managed before BE service users. For each type of service, the scheduler ranks the users according the scheduling
technique. Three scheduling algorithms are available, Max C/I, Round Robin and Proportional Fair. Impact they have on the
simulation result is described in the tables below.
Let us consider a cell with 16 HSDPA and HSPA BE service users. All of them are active on DL and connected to the A-DCH R99
bearer. There is neither CBR service user, nor VBR service user in the cell and the number of HS-SCCH channels and the
maximum number of HSDPA bearer users have been respectively set to 4 and 15.
Max C/I
15 users (where 15 corresponds to the maximum number of HSDPA bearer users defined) enters the scheduler in the same
order as in the simulation. Then, they are sorted in descending order by the channel quality indicator (CQI), i.e. in a best bearer
descending order.
DL Obtained
Mobiles Simulation Rank Best Bearer (kbps) Connection Status
Throughput (kbps)
M1 2 2400 2400+3.4 Connected
M2 15 2400 1440+3.4 Connected
M3 8 2080 160+3.4 Connected
M4 9 2080 3.4 Delayed
M5 10 2080 3.4 Delayed
M6 12 2080 3.4 Delayed
M7 13 2080 3.4 Delayed
M8 14 2080 3.4 Delayed
M9 7 1920 3.4 Delayed
M10 1 1600 3.4 Delayed
M11 3 1600 3.4 Delayed
M12 4 1600 3.4 Delayed
M13 5 1600 3.4 Delayed
M14 6 1600 3.4 Delayed
M15 11 1440 3.4 Delayed
M16 16 2080 0 Scheduler Saturation
Round Robin
Users are taken into account in the same order than the one in the simulation (random order).
253
DL Obtained
Mobiles Simulation Rank Best Bearer (kbps) Connection Status
Throughput (kbps)
M1 1 1600 1600+3.4 Connected
M2 2 2400 960+3.4 Connected
M3 3 1600 3.4 Delayed
M4 4 1600 3.4 Delayed
M5 5 1600 3.4 Delayed
M6 6 1600 3.4 Delayed
M7 7 1920 3.4 Delayed
M8 8 2080 3.4 Delayed
M9 9 2080 3.4 Delayed
M10 10 2080 3.4 Delayed
M11 11 1440 3.4 Delayed
M12 12 2080 3.4 Delayed
M13 13 2080 3.4 Delayed
M14 14 2080 3.4 Delayed
M15 15 2400 3.4 Delayed
M16 16 2080 0 Scheduler Saturation
Proportional Fair
15 users (where 15 corresponds to the maximum number of HSDPA bearer users defined) enters the scheduler in the same
order as in the simulation. Then, they are sorted in an ascending order according to a new random parameter which
corresponds to a combination of the user rank in the simulation and the channel quality indicator (CQI).
For a user i, the random parameter RP i is calculated as follows:
Simu CQI
RP i = 50  R i + 50  R i
Where,
Simu
Ri is the user rank in the simulation.
CQI
Ri is the user rank according to the CQI.
You can change the default weights by editing the atoll.ini file. For more information, see
the Administrator Manual.
DL Obtained
Simulation Best Bearer Connection
Mobiles CQI Rank RP Throughput
Rank (kbps) Status
(kbps)
M1 2 1 150 2400 2400 Connected
M2 1 10 550 1600 960 Connected
M3 8 3 550 2080 160 Connected
M4 9 4 650 2080 3.4 Delayed
M5 3 11 700 1600 3.4 Delayed
M6 10 5 750 2080 3.4 Delayed
M7 4 12 800 1600 3.4 Delayed
M8 7 9 800 1920 3.4 Delayed
M9 15 2 850 2400 3.4 Delayed
M10 5 13 900 1600 3.4 Delayed
M11 12 6 900 2080 3.4 Delayed
M12 6 14 1000 1600 3.4 Delayed
M13 13 7 1000 2080 3.4 Delayed
254
DL Obtained
Mobiles CQI Rank RP Throughput
Rank (kbps) Status
(kbps)
M14 14 8 1100 2080 3.4 Delayed
M15 11 15 1300 1440 3.4 Delayed
Scheduler
M16 16 - - 2080 0
Saturation
4.3.2.3.7 Dual-Cell HSDPA

For transmitters that support multi-cell HSDPA mode, the scheduler manages a single queue of users at the Node B.
MC-HSDPA and DB-MC-HSDPA users are processed as DC-HSDPA users if they are connected to two carriers. Otherwise, they
are considered as single-cell HSDPA users. All users belonging to the transmitter, i.e., DC-HSDPA and single-carrier HSDPA
users, are ranked together in a unique list. DC-HSDPA users are considered twice in the list as they may be assigned two
different HSDPA bearers in the two cells.
CBR service users have the highest priority and are processed first, in the order established during the generation of the user
distribution. After processing the CBR service users, the scheduler processes the remaining users (i.e., VBR and BE service
users). VBR service users have the highest priority and are managed before BE service users. For each type of service, the
scheduler ranks the users according the scheduling technique (Max C/I, Round Robin and Proportional Fair). After the users
have been ranked, the scheduler allocates HSDPA resources to each user following the calculated order as long as there are
resources available. Even if there is a unique list of users at the transmitter level, the resources of each cell are not shared and
each carrier has its own pool of resources (number of HS-SCCH channels, maximum number of HSDPA bearer users, HSDPA
power, number of OVSF codes). Only site-level resources (such as the Iub throughput and the channel elements) are shared
between the users of the two cells.
Let us consider a transmitter with 16 BE service users. The transmitter supports the multi-cell HSDPA mode. There is neither
CBR service user, nor VBR service users. All users are active in DL and connected to the A-DCH R99 bearer. Among the users,
there are 6 DC-HSDPA users (i.e., terminal with UE categories 21 to 24).
Simulation Rank DC-HSDPA Support Carriers Comments
1 Yes 1 and 2 Anchor carrier: 2
2 No 2
3 No 1
5 No 1
6 No 2
7 No 1
8 No 2
10 No 1
11 No 2
13 No 2
15 No 1
In each cell, the number of HS-SCCH channels and the maximum number of HSDPA bearer users have been respectively set
to 4 and 7.
The scheduling algorithms defined for the two cells are the same as the one selected for the transmitter.
Each DC-HSDPA user is counted twice, once in each cell, as he may be assigned two different HSDPA bearers in the two cells.
Therefore, the scheduler manages the users ranked 1st to 11th (i.e. 4 single-carrier users connected to the first carrier, 4 single-
carrier users connected to the second carrier and 3 DC-HSDPA users). Users ranked 12th to 16th are rejected because the
maximum number of HSDPA bearer users that the scheduler can manage in a cell is exceeded.
Impact the scheduling algorithms have on the simulation results is described in the tables below.
255
Max C/I
7 users from each cell (where 7 corresponds to the maximum number of HSDPA bearer users defined for each cell), i.e., a total
of 14 users enter the scheduler in the same order as in the simulation. Then, they are sorted in the order of decreasing channel
quality indicator (CQI), i.e. in a best bearer descending order.
DL Obtained
Mobiles Carrier CQI Throughput
Rank (kbps) Status
(kbps)
M1 1 5 21 3040 3040+3.4 Connected
M2
2 4 19 2400 2400+3.4 Connected
(DC-HSDPA)
M3 2 8 18 2080 1440+3.4 Connected
M2
1 4 17 1920 1920 Connected
(DC-HSDPA)
M4
1 9 17 1920 960+3.4 Connected
(DC-HSDPA)
M5 1 3 16 1600 3.4 Delayed
M4
2 9 16 1600 1120 Connected
(DC-HSDPA)
M6 2 2 15 1440 3.4 Delayed
M7 1 7 14 1120 3.4 Delayed
M8 1 10 14 1120 3.4 Delayed
M9
2 1 13 960 3.4 Delayed
(DC-HSDPA)
M10 2 6 13 960 3.4 Delayed
M9
1 1 12 800 0 Delayed
(DC-HSDPA)
M11 2 11 12 800 3.4 Delayed
M12 1 14 1120 Scheduler
12 0
(DC-HSDPA) 2 15 1440 Saturation
Scheduler
M13 2 13 17 1920 0
Saturation
14 0
Scheduler
M15 1 15 17 1920 0
Saturation
16 0
The scheduled DC-HSDPA users have the following status:
• The user ranked 4th (here M2) is connected to an HSDPA bearer in each cell. He obtains a total DL throughput of 4323.4
kbps (2403.4+1920).
kbps (963.4+1120).
• The first user (here M9) is delayed in the two cells. He obtains a total DL throughput of 3.4 kbps.
Round Robin
of 14 users enter the scheduler in the same order as in the simulation.
DL Obtained
Rank (kbps) Status
(kbps)
M1
1 1 12 800 800 Connected
(DC-HSDPA)
256
DL Obtained
Rank (kbps) Status
(kbps)
M1
2 1 13 960 960+3.4 Connected
(DC-HSDPA)
M2 2 2 15 1440 1440+3.4 Connected
M3 1 3 16 1600 1600+3.4 Connected
M4
2 4 19 2400 1600+3.4 Connected
(DC-HSDPA)
M4
1 4 17 1920 960 Connected
(DC-HSDPA)
M5 1 5 21 3040 480+3.4 Connected
M6 2 6 13 960 160+3.4 Connected
M7 1 7 14 1120 3.4 Delayed
M8 2 8 18 2080 3.4 Delayed
M9
2 9 16 1600 0 Delayed
(DC-HSDPA)
M9
1 9 17 1920 3.4 Delayed
(DC-HSDPA)
M10 1 10 14 1120 3.4 Delayed
M11 2 11 12 800 3.4 Delayed
12 0
Scheduler
M13 2 13 17 1920 0
Saturation
14 0
Scheduler
M15 1 15 17 1920 0
Saturation
16 0

• The first user (here M1) is connected to an HSDPA bearer in each cell. He obtains a total DL throughput of 1763.4 kbps
(800+963.4).
kbps (1603.4+960).
• The user ranked 9th (here M9) is delayed in the two cells. He obtains a total DL throughput of 3.4 kbps.
Proportional Fair
of 14 users enter the scheduler in the same order as in the simulation. Then, they are sorted in an ascending order according
to a new random parameter which corresponds to a combination of the user rank in the simulation and the channel quality
indicator (CQI).
For a user i, the random parameter RPi is calculated as follows:
Simu CQI
RPi = 50  R i + 50  R i
Where,
Simu
Ri is the user rank in the simulation.
CQI
Ri is the user rank according to the CQI.
257
You can change the default weights by editing the atoll.ini file. For more information, see
DL
Simulation Best Bearer Obtained Connection
Mobiles Carrier CQI CQI Rank RP
Rank (kbps) Throughput Status
(kbps)
M1
2 4 19 2 300 2400 2400+3.4 Connected
DC-HSDPA
M2 1 5 21 1 300 3040 3040+3.4 Connected
M1
1 4 17 4 400 1920 1440 Connected
DC-HSDPA
M3 1 3 16 6 450 1600 800+3.4 Connected
M4 2 2 15 8 500 1440 1120+3.4 Connected
M5 2 8 18 3 550 2080 800+3.4 Connected
M6
2 1 13 11 600 960 480+3.4 Connected
DC-HSDPA
M6
1 1 12 13 700 800 0 Delayed
DC-HSDPA
M7
1 9 17 5 700 1920 3.4 Delayed
DC-HSDPA
M8 1 7 14 9 800 1120 3.4 Delayed
M7
2 9 16 7 800 1600 0 Delayed
DC-HSDPA
M9 2 6 13 12 900 960 3.4 Delayed
M10 1 10 14 10 1000 1120 3.4 Delayed
M11 2 11 12 14 1250 800 3.4 Delayed
M12 1 14 1120 Scheduler Scheduler
12 0 0
(DC-HSDPA) 2 15 1440 Saturation Saturation
Scheduler Scheduler
M13 2 13 17 1920 0 0
Saturation Saturation
14 0 0
Scheduler Scheduler
M15 1 15 17 1920 0 0
Saturation Saturation
16 0 0
kbps (2403.4+1440).
• The first user (here M6) is connected to an HSDPA bearer in his anchor cell and delayed in the other cell. He obtains a
total DL throughput of 483.4 kbps (483.4+0).
• The user ranked 9th (here M7) is delayed in the two cells. He obtains a total DL throughput of 3.4 kbps.
4.3.2.4 HSUPA Part of the Algorithm

HSPA VBR and BE service users active in the UL as well as all HSPA CBR service users (i.e., active and inactive), unless they have
been rejected during the R99 or HSDPA parts of the algorithm, are then evaluated by the HSUPA part of the algorithm. Atoll
manages the maximum number of users within each cell. CBR service users have the highest priority and are processed first,
in the order established during the generation of the user distribution. Then, Atoll considers VBR service users in the order
established during the generation of the user distribution and lastly, it processes BE service users in the order established
during the generation of the user distribution.
Let us assume there are 12 HSPA users in the cell:
258
• 3 CBR service users with any activity status. All of them have been connected to an HSDPA bearer.
• 2 packet VBR service users. They have been connected to an HSDPA bearer.
• 7 packet BE service users active on UL. The first two users have been connected to an HSDPA bearer, the last one has
been rejected and the remaining four have been delayed in the HSDPA part.
Finally, the maximum number of HSUPA bearer users equals 10.
In this case, Atoll will consider the first ten HSPA users only and will reject the last two users in order not to exceed the
maximum number of HSUPA bearer users allowed in the cell (their connection status is "HSUPA scheduler saturation").
Evaluation by the
HSDPA
Mobiles Service Simulation Rank HSUPA part of the
Connection Status
algorithm
M1 CBR 4 Connected Yes
M4 VBR 3 Connected Yes
M5 VBR 5 Connected Yes
M6 BE 1 Connected Yes
M7 BE 2 Connected Yes
M8 BE 6 Delayed Yes
M9 BE 8 Delayed Yes
M10 BE 10 Delayed Yes
M11 BE 11 Delayed No
M12 BE 12 Rejected No
4.3.2.4.1 Admission Control

During admission control, Atoll selects a list of HSUPA bearers for each user. The selected HSUPA bearers have to be
compatible with the user equipment and capabilities of each HSUPA cell of the active set.
For CBR service users, the list is restricted to HSUPA bearers that provide a peak RLC throughput higher than the minimum
throughput demand.
For VBR service users, the list of compatible bearers is restricted to HSUPA bearers that provide a peak RLC throughput
between the maximum and the minimum throughput demands.
Let us focus on one HSPA-BE service user with category 3 user equipment and a 50km/h mobility. This user is connected to
one cell only. The cell supports HSPA+ functionalities, i.e the cell supports QPSK and 16QAM modulations in the UL.
HSUPA user equipment categories are provided in the HSUPA User Equipment Categories table. The capabilities of the
category 3 user equipment are:
• Maximum Number of E-DPDCH codes: 2
• TTI 2 ms: No so it supports 10 ms TTI
• Minimum Spreading Factor: 4
• Maximum Block Size for a 2ms TTI: no value
• Maximum Block Size for a 10ms TTI: 14484 bits
• Highest Modulation Supported: QPSK
Figure 4.9: HSUPA UE Categories Table
HSUPA bearer characteristics are provided in the HSUPA Bearer table. An HSUPA bearer is described with following
characteristics:
• Radio Bearer Index: The bearer index number.
259
• TTI Duration (ms): The TTI duration in ms. The TTI can be 2 or 10 ms.
• Transport Block Size (Bits): The transport block size in bits.
• Number of E-DPDCH Codes: The number of E-DPDCH channels used.
• Minimum Spreading Factor: The smallest spreading factor used.
• Modulation: the modulation used (QPSK or 16QAM)
• Peak RLC Throughput (bps): The RLC peak throughput represents the peak throughput without coding (redundancy,
overhead, addressing, etc.).
HSUPA bearers can be classified into two categories:
• HSUPA bearers using QPSK modulation: They can be selected for users connected to HSPA and HSPA+ capable cells.
• HSUPA bearers using 16QAM modulation (improvement introduced by the release 7 of the 3GPP UTRA specifications,
referred to as HSPA+). These HSUPA bearers can be allocated to users connected to cells with HSPA+ capabilities only.
Atoll considers an HSUPA bearer as compatible with the category 3 user equipment if:
• The TTI duration used by the bearer is supported by the user equipment (10 ms).
• The transport block size does not exceed the maximum transport block size supported by the user equipment (14484
bits):
• The number of E-DPDCH channels required by the bearer does not exceed the maximum number of E-DPDCH channels
that the terminal can use (2).
• The minimum spreading factor used by the bearer is not less than the smallest spreading factor supported by the
terminal (4).
• The modulation required by the bearer is supported by the terminal.
The HSUPA bearers compatible with category 3 user equipment are framed in red:
Figure 4.10: HSUPA Radio Bearers Table
Then, during admission control, Atoll checks that the lowest compatible bearer in terms of the required E-DPDCH Ec⁄Nt does
not require a terminal power higher than the maximum terminal power allowed.
Atoll uses the HSUPA Bearer Selection table. Among the compatible HSUPA bearers, Atoll chooses the one with the lowest
required Ec/Nt threshold.
Here, this is the index 1 HSUPA bearer; the required Ec/Nt threshold to obtain this bearer is -21.7dB.
Ec req
Then, from the required Ec/Nt threshold,  ------
req
, Atoll calculates the required terminal power, P term – HSUPA .
 Nt E – DPDCH
Ec req
P term – HSUPA =  ------
req UL
 L T  N tot
Nt E – DPDCH
With
UL intra UL extra
UL tx UL tx
N tot  ic  =  1 – F MUD   term   I tot  ic  + I tot  ic  + I inter – carrier  ic  + N 0

L T = -------------------------------------------------------------------------------------------------------------------
- ( )
G Tx  G term
260
UL UL
tx intra extra UL tx
 term , F MUD , I tot , I tot , I inter – carrier and N 0 are defined in "Inputs" on page 215.
Figure 4.11: HSUPA Bearer Selection Table
req
Atoll rejects the user if the terminal power required to obtain the lowest compatible HSUPA bearer ( P term – HSUPA ) exceeds
the maximum terminal power (his connection status is "HSUPA Admission Rejection").
At the end of this step, the number of non-rejected HSUPA bearer users is n HSUPA . All of them will be connected to an HSUPA
bearer at the end.
4.3.2.4.2 HSUPA Bearer Allocation Process

The HSUPA bearer allocation process depends on the type of service requested by the user. As explained before, CBR service
users have the highest priority and are processed first, in the order established during the generation of the user distribution.
After the admission control on CBR service users, Atoll performs a noise rise scheduling, followed by a radio resource control.
Then, it repeats the same steps on VBR service users first, and lastly on BE service users, in the order established during the
generation of the user distribution.
CBR Service Users
Let us focus on the three CBR service users mentioned in the example of the previous paragraph "HSUPA Part of the
Algorithm" on page 258. We assume that all of them have been admitted. Noise rise scheduling and radio resource control
are carried out on each user in order to determine the best HSUPA bearer that the user can obtain. Several CBR service users
can share the same HSUPA bearer. Then, Atoll calculates the HSUPA bearer consumption ( C in %) for each user and takes into
account this parameter when it determines the resources consumed by the user (i.e., the terminal power used, the number
of channel elements and the Iub backhaul throughput).
In the bearer allocation process shown below, the 3 CBR service users are represented by Mj, with j = 1 to 3.
8. In the HSUPA coverage prediction, L T is calculated as follows:

UL
L T = ----------------------------------------------------------------------------------------------------------------------------------------------
-)
G Tx  G term
261
Determination of the best HSUPA bearer B(Mj)
Allocation of the minimum throughput demand to

Mj
Calculation of C(B(Mj))
Sufficient Iub
backhaul throughput No Is there a lower HSUPA
to support the HSUPA bearer available?
bearer? Yes
No
Downgrading to lower
Yes Mj is rejected HSUPA bearer
Yes
Enough
channel elements No Is there a lower HSUPA
available to support the bearer available?
HSUPA bearer?
No
Yes
Mj is rejected
Pterm-HSUPA recalculation and interference update
No
Mj = M3?
Yes
Resource allocation for packet (HSPA –

Variable Bit Rate) service users
Figure 4.12: HSUPA Bearer Allocation Process for CBR Service Users
VBR Service Users
Let us focus on the two VBR service users mentioned in the example of the previous paragraph "HSUPA Part of the Algorithm"
on page 258. We assume that all of them have been admitted. Noise rise scheduling and radio resource control are carried
out on each user in order to determine the best HSUPA bearer that the user can obtain.
In the bearer allocation process shown below, the 2 VBR service users are represented by Mj, with j = 1 to 2.
Determination of the best HSUPA bearer
Sufficient Iub
bearer? Yes
No
Downgrading to lower HSUPA
Yes Mj is rejected bearer
Yes
Enough
HSUPA bearer?
No
Yes
Mj is rejected
No
Mj = M2?
Yes
Resource allocation for packet (HSPA – Best

Effort) service users
Figure 4.13: HSUPA Bearer Allocation Process for VBR Service Users
BE Service Users
Let us focus on the five BE service users mentioned in the example of the previous paragraph "HSUPA Part of the Algorithm"
on page 258. We assume that all of them have been admitted. Noise rise scheduling and radio resource control are carried
out on each user in order to determine the best HSUPA bearer that the user can obtain.
In the bearer allocation process shown below, the 5 BE service users are represented by Mj, with j = 1 to 5.
262
Determination of the best HSUPA bearer
Sufficient Iub
bearer? Yes
No
Downgrading to lower HSUPA
Yes Mj is rejected bearer
Yes
Enough
HSUPA bearer?
No
Yes
Mj is rejected
No
Mj = M5?
Figure 4.14: HSUPA Bearer Allocation Process for BE Service Users
4.3.2.4.3 Noise Rise Scheduling

Determination of the Obtained HSUPA Bearer
The obtained HSUPA radio bearer is the bearer that the user obtains after noise rise scheduling and radio resource control.
CBR service users have the highest priority and are processed first. Therefore, after the admission control, the noise rise
scheduling algorithm attempts to evenly share the remaining cell load between the CBR service users admitted in admission
control; in terms of HSUPA, each user is allocated a right to produce interference. The remaining cell load factor on uplink
UL
( X HSPA – CBR  txi ic  ) depends on the maximum load factor allowed on uplink and how much uplink load is produced by the
served R99 traffic. It can be expressed as follows:
UL UL UL
X HSPA – CBR  txi ic  = X max  txi ic  – X R99  txi ic 
Then, Atoll evenly shares the remaining cell load factor between the CBR service users admitted during the previous step
( n HSPA – CBR ).
UL
UL X HSPA – CBR  txi ic 
X user  txi ic  = ------------------------------------------------
n HSPA – CBR
Ec max
From this value, Atoll calculates the maximum E-DPDCH Ec⁄Nt allowed (  ------ ) for each CBR service user. For further
information on the calculation, see "Uplink Load Factor Due to One User" on page 281.
Ec- max
 ----- 1
 Nt E – DPDCH = -------------------------------------------
UL
- for the Without useful signal option
F  txi ic 
---------------------------------- – 1
UL
X user  txi ic 
UL
Ec- max
 ----- X user
= --------------- for the Total noise option
 Nt E – DPDCH UL
F
Then, it selects an HSUPA bearer. The allocation depends on the maximum E-DPDCH Ec⁄Nt allowed and on UE and cell capa-
bilities. Atoll selects the best HSUPA bearer from the HSUPA compatible bearers. This is the HSUPA bearer ( Index HSUPABearer )
UL
TP P – R LC  Index HSUPABearer 
with the highest potential throughput ( ----------------------------------------------------------------- ) where:
N Rtx  Index HSUPABearer 
Ec- req
 ----- Ec max
•   ------
 Nt E – DPDCH  Nt E – DPDCH
req max
• And P term – HSUPA  P term
req
When several HSUPA bearers are available, Atoll selects the one with the lowest  ------
Ec
.
Nt E – DPDCH
263
After the noise rise scheduling, Atoll carries out radio resource control, verifying if enough channel elements and Iub backhaul
throughput are available for the HSUPA bearer assigned to the user. For information on radio resource control, see "Radio
Resource Control" on page 267.
After processing all CBR service users, Atoll carries out noise rise scheduling and radio resource control on VBR service users.
During the noise rise scheduling, Atoll distributes the remaining cell load factor available after all CBR service users have been
served. It can be expressed as follows:
UL UL UL UL
X HSPA – VBR  txi ic  = X max  txi ic  – X R99  txi ic  – X HSPA – CBR  txi ic 
The remaining cell load factor is shared equally between the admitted VBR service users ( n HSPA – VBR ).
UL
UL X HSPA – VBR  txi ic 
X user  txi ic  = ------------------------------------------------
-
n HSPA – VBR
Ec max
From this value, Atoll calculates the maximum E-DPDCH Ec⁄Nt allowed (  ------ ) as explained above and selects an
HSUPA bearer for each VBR service user. After the noise rise scheduling, Atoll carries out radio resource control on VBR service
users. For information on radio resource control, see "Radio Resource Control" on page 267.
After processing VBR service users, Atoll carries out noise rise scheduling and radio resource control on BE service users.
During the noise rise scheduling, Atoll distributes the remaining cell load factor available after all CBR and VBR service users
have been served. It can be expressed as follows:
UL UL UL UL UL
X HSPA  txi ic  = X max  txi ic  – X R99  txi ic  – X HSPA – CBR  txi ic  – X HSPA – VBR  txi ic 
The remaining cell load factor is shared equally between the admitted BE service users ( n HSPA ).
UL
UL X HSPA  txi ic 
X user  txi ic  = ------------------------------------
n HSPA
Ec max
From this value, Atoll calculates the maximum E-DPDCH Ec⁄Nt allowed (  ------ ) as explained above and selects an
HSUPA bearer for each BE service user. After the noise rise scheduling, Atoll carries out radio resource control on BE service
users. For information on radio resource control, see "Radio Resource Control" on page 267.
Example: We have a cell with six BE service users, and neither CBR user nor VBR user. All BE service users have been admitted.
The remaining cell load factor equal to 0.6 is shared between the BE service users. Therefore, the UL load factor allotted to
each user is 0.1. Let’s take the cell UL reuse factor equal to 1.5. Atoll calculates the maximum E-DPDCH Ec⁄Nt allowed (the
Without useful signal option is selected).
Ec max
We have:  ------ = -11.5 dB
Here, the obtained HSUPA bearer is the index 5 HSUPA bearer. It provides a potential throughput of 128 kbps and requires
E-DPDCH Ec⁄Nt of -13 dB (lower than -11.5 dB) and a terminal power lower than the maximum terminal power allowed.
. Required Ec/Nt Peak RLC Throughput Potential Throughput
HSUPA Bearers Index Nb of Retransmissions
Threshold (dB) (kbps) (kbps)
1 -21.7 2 32 16
2 -19 2 64 32
3 -16.1 2 128 64
4 -13.9 2 192 96
5 -13 2 256 128
6 -10.1 2 512 256
7 -8 2 768 384
8 -7 2 1024 512
AtollAtollNoise Rise Scheduling in Soft Handover
With HSUPA, uplink soft handover impacts the scheduling operation. While HSDPA sends data from one cell only, with HSUPA
all cells in the active set receive the transmission from the terminal. Therefore, all the cells are impacted by the transmission
in terms of noise rise.
264
For each HSPA-capable cell of the active set  tx k ic  , Atoll calculates the maximum E-DPDCH Ec⁄Nt allowed
max
(  ------
Ec
 tx  ic  ) as explained in "HSUPA Bearer Allocation Process" on page 261.
Nt E – DPDCH k
For each cell of the active set  tx k ic  , Atoll calculates the maximum terminal power allowed to obtain an HSUPA radio bearer
max
( P term – HSUPA  tx k ic  ).
max
P term – HSUPA  tx k ic  = min    ------  tx  ic   L T  N tot  P term
max Ec UL max
 Nt E – DPDCH k
With
UL UL extra
UL tx intra UL tx

L T = -------------------------------------------------------------------------------------------------------------------
- ( )
G Tx  G term
UL intra UL extra
tx UL tx
As HSUPA bearer users in soft handover use the lowest granted noise rise, Atoll chooses the lowest of maximum terminal
power allowed for each cell of the active set  tx k ic  .
max max
P term – HSUPA = min  P term – HSUPA  tx k ic  
tx  AS
k
max
Once Atoll knows the selected maximum terminal power ( P term – HSUPA ), it recalculates the maximum E-DPDCH Ec⁄Nt allowed
Ec max
(  ------  tx  ic  ) for each HSUPA-capable cell of the active set.
max
Ec- max
 ----- P term – HSUPA
 Nt E – DPDCH  tx k ic  = -----------------------------
-
UL
L T  N tot
max
Then, Atoll calculates the maximum E-DPDCH Ec⁄Nt allowed (  ------
Ec
) after signal recombination of all HSUPA capable
cells of the active set 10.
For softer (1/2) and softer-softer (1/3) handovers, we have:
max max
 -----
Ec-  -----
Ec-

UL
= f rake efficiency   tx  ic 
 Nt E – DPDCH  Nt E – DPDCH k
txk  ActiveSet
 samesite 
Ec max Ec- max

For soft (2/2) and soft-soft (3/3) handovers, we have:  ------ = Max   -----  tx  ic 
 Nt E – DPDCH   Nt E – DPDCH k 
txk  ActiveSet
For softer-soft handover (2/3), it depends on if the MRC option is selected (option available in Global parameters). If selected,
we have:
9. In the HSUPA coverage prediction, L T is calculated as follows:

UL
L T = ----------------------------------------------------------------------------------------------------------------------------------------------
-)
G Tx  G term
265
Max  
 -----
Ec- max  UL  Ec- max  Ec- max 
 Nt E – DPDCH = tx ,tx  ActiveSet  f rake efficiency 
k l   Nt 
-----
 E – DPDCH
 tx k ic   
-----
Nt E – DPDCH
 tx l ic 

tx  samesite  tx 
k k
tx  othersite
l
max
Ec- max
Else, we have:  ------ Max   -----  tx  ic 
Ec
=
txk  ActiveSet
Then, Atoll selects an HSUPA bearer as previously explained in "HSUPA Bearer Allocation Process" on page 261. The allocation
depends on the maximum E-DPDCH Ec⁄Nt allowed and on UE and cell capabilities. Atoll selects the best HSUPA bearer from
the HSUPA compatible bearers. This is the HSUPA bearer ( Index HSUPABearer ) with the highest potential throughput
UL
TP P – R LC  Index HSUPABearer 
( ----------------------------------------------------------------- ) where:
Ec- req
 ----- Ec- max
 -----
•  Nt E – DPDCH   Nt E – DPDCH
Ec req
When several HSUPA bearers are available, Atoll selects the one with the lowest  ------ .
Nt E – DPDCH
10. In HSUPA coverage predictions, we have the following:
Ec max max
For softer (1/2) and softer-softer (1/3) handovers:  ------  -----
Ec-

UL
= f rake efficiency   Nt E – DPDCH  tx k ic 
Nt E – DPDCH
txk  ActiveSet
 samesite 
Ec max Ec- max

For soft handover (2/2):  ------ Max   -----  tx  ic    G macro – diversity  2links
UL
=
txk  ActiveSet
Ec max Ec- max

For soft-soft handover (3/3):  ------ Max   -----  tx  ic    G macro – diversity  3links
UL
=
tx k  ActiveSet
For softer-soft handover (2/3), it depends on if the MRC option is selected (option available in Global parameters). If selected,
we have:
Max  
 -----
Ec- max  UL  Ec- max  Ec- max 
= txk ,txl  ActiveSet  rake efficiency

f  
----- 
tx  ic   ----- 
 Nt E – DPDCH l
tx  ic 

tx k  samesite  tx k 
tx  othersite
l
UL
  G macro – diversity  2links
max
Ec- max
Else, we have:  ------ Max   ----- 
Ec UL
  Nt E – DPDCH  tx k ic    G macro – diversity  2links
=
Nt E – DPDCH txk  ActiveSet
266
Determination of the Requested HSUPA Bearer
The requested HSUPA radio bearer is selected from the HSUPA bearers compatible with the user equipment. Atoll determines
the HSUPA bearer the user would obtain by considering the entire remaining load of the cell. The user is treated as if he is the
only user in the cell. Therefore, if we go on with the previous example, the maximum E-DPDCH Ec⁄Nt allowed is equal to -1.8
dB and the requested HSUPA bearer is the index 7 HSUPA bearer. It requires E-DPDCH Ec⁄Nt of -8 dB (lower than -1.8 dB) and
a terminal power lower than the maximum terminal power allowed.
4.3.2.4.4 Radio Resource Control

Atoll checks to see if enough channel elements are available and if the Iub backhaul throughput is sufficient for the HSUPA
bearer assigned to the user (taking into account the maximum number of channel elements defined for the site and the
maximum Iub backhaul throughput allowed on the site in the uplink). If not, Atoll allocates a lower HSUPA bearer
("downgrading") which needs fewer channel elements and consumes lower Iub backhaul throughput. If no channel elements
are available, the user is rejected. On the same hand, if the maximum Iub backhaul throughput allowed on the site in the
uplink is still exceeded even by using the lowest HSDPA bearer, the user is rejected.
4.3.2.5 Convergence Criteria

The convergence criteria are evaluated for each iteration, and can be written as follow:
DL DL
  max P tx  ic  k – P tx  ic  k – 1   max N user  ic  k – N user  ic  k – 1 
 DL  Stations  Stations
-  100 
= max int  -----------------------------------------------------------------------------------  100  int ---------------------------------------------------------------------------------------------
  P tx  ic  k   DL
N user  ic  k 
  
UL UL UL UL
  max I tot  ic  k – I tot  ic  k – 1   max N user  ic  k – N user  ic  k – 1 
 UL = max  int  -----------------------------------------------------------------------------------
-  100  int  ---------------------------------------------------------------------------------------------
-  100 
Stations Stations
UL   UL 
  I tot  ic  k   N user  ic  k 
Atoll stops the algorithm if:
1st case: Between two successive iterations,  UL and  DL are lower than their respective thresholds (defined when creating
a simulation).
The simulation has reached convergence.
Example: Let us assume that the maximum number of iterations is 100, UL and DL convergence thresholds are set to 5. If
 UL  5 and  DL  5 between the 4th and the 5th iteration, Atoll stops the algorithm after the 5th iteration. Convergence has
been reached.
2nd case: After 30 iterations,  UL and/or  DL are still higher than their respective thresholds and from the 30th iteration,  UL
and/or  DL do not decrease during the next 15 successive iterations.
The simulation has not reached convergence (specific divergence symbol).

Examples: Let us assume that the maximum number of iterations is 100, UL and DL convergence thresholds are set to 5.
1. After the 30th iteration,  UL and/or  DL equal 100 and do not decrease during the next 15 successive iterations: Atoll
stops the algorithm at the 46th iteration. Convergence has not been reached.
2. After the 30th iteration,  UL and/or  DL equal 80, they start decreasing slowly until the 40th iteration (without going
under the thresholds) and then, do not change during 15 successive iterations: Atoll stops the algorithm at the 56th
iteration without reaching convergence.
3rd case: After the last iteration.

If  UL and/or  DL are still strictly higher than their respective thresholds, the simulation has not reached convergence
(specific divergence symbol).
If  UL and  DL are lower than their respective thresholds, the simulation has reached convergence.
4.3.3 Results
4.3.3.1 R99 Related Results
This table contains some R99 specific simulation results provided in the Cells and Mobiles tabs of the simulation property
dialogue.
267
  E1  T1  Ethernet 
 TPIub – DL  N I    TP
RoundUp  Max   
Number of E1/T1/Ethernet links
Nb E1  T1  Ethernet None
  E1  T1  Ethernet  required by the site
  TP Iub – UL  N I   TP 
 DL 
 P  txi ic  – P SCH  txi ic 
P tot  txi ic  – F ortho   BTS   tot -----------------------------
-
DL Downlink intra-cell interference at
DL
I intra  txi ic  LT  None
 txi  terminal on carrier ic
DL
–  1 – F ortho    BTS  P b  txi ic 


DL
DL
txj j  i

DL
I inter – carrier  ic   j
W
---------------------------------------------
txj - at terminal on carrier ic
DL
Tx
L total  ICP n  ic
ni i a
Total effective interference at

DL DL DL DL DL
I tot  ic  I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  W terminal on carrier ic (after
unscrambling)
DL DL Term Total received noise at terminal on
N tot  ic  I tot  ic  + N 0 W
carrier ic

 Pb
UL
UL intra  ic 
term
txi
ic


UL
UL extra P b  ic 
term
txj j  i
carrier ic
 Pb
UL
 ic adj 
txj j
-----------------------------------
-

Tx
 txi ic +  1 – F MUD   term  I tot  txi ic  +I inter – carrier  txi icW
UL
transmitter on carrier ic

UL UL tx
N tot  txi ic  I tot  txi ic  + N0 W ic
UL
I tot  txi ic 
UL
X  txi ic  ----------------------------
- None Cell uplink load factor on carrier ic
UL
UL
UL
F  txi ic  --------------------------------------------------------------------------------------- None Cell uplink reuse factor on carrier ic
UL
intra Tx
I tot  txi ic    1 – F MUD   term 
1 Cell uplink reuse efficiency factor on

UL
E  txi ic  --------------------------- None
UL
F  txi ic  carrier ic
268

Simulation result available per cell
DL DL
 I extra  ic  + I inter – carrier  ic    L T
---------------------------------------------------------------------------------- + 1 – F ortho   BTS
DL
P Tx  txi ic 
 ---------------------------------------------------------------------------------------------------------------------------------
1 - + 1 – F
----------  
tch DL ortho BTS
CI req
DL
X  txi ic  DL None Downlink load factor on carrier ic
DL Q req
with CI req = ---------
-
DL
Gp
Simulation result available per mobile
DL
I tot  ic 
------------------
-
DL
N tot  ic 
DL
I tot  ic 
DL
F  txi ic  -----------------------------
- None Downlink reuse factor on a carrier ic
DL
I intra  txi ic 
DL DL
NR  txi ic  – 10 log  1 – X  txi ic   dB Noise rise on downlink
UL UL
NR  txi ic  – 10 log  1 – X  txi ic   dB Noise rise on uplink
4.3.3.2 HSPA Related Results

At the end of the R99 part, the users can be:
• Either connected and in this case, they obtain the requested R99 bearer,
• Or rejected exactly for the same reasons as R99 users.
Only connected HSDPA and HSPA users are considered in the HSDPA part. At the end of the HSDPA part, BE service users can
be:
• Either connected if they obtain an HSDPA bearer,
• Or rejected if the maximum number of HSDPA bearer users per cell is exceeded,
• Or delayed in case of lack of resources (HSDPA power, HS-SCCH power, HS-SCCH channels, OVSF codes).
VBR service users can be:
• Or rejected for the following reasons: the maximum number of HSDPA bearer users per cell is exceeded, the lowest
HSDPA bearer the user can obtain does not provide a peak RLC throughput higher than the minimum throughput
demand, the HS-SCCH signal quality is not sufficient, there are no more OVSF codes available, the maximum Iub
backhaul throughput allowed on the site in the downlink is exceeded.
CBR service users can be:
• Or rejected for the following reasons: the maximum number of HSDPA bearer users per cell is exceeded, the lowest
HSDPA bearer the user can obtain does not provide a peak RLC throughput higher than the minimum throughput
demand, the HS-SCCH signal quality is not sufficient, there are no more OVSF codes available, the maximum Iub
backhaul throughput allowed on the site in the downlink is exceeded.
In the HSUPA part, Atoll processes HSPA service users who are connected to an HSDPA bearer or were delayed in the previous
step. At the end, they can be:
• Either connected if they obtain an HSUPA bearer,
• Or rejected for the following reasons: the maximum number of HSUPA bearer users per cell is exceeded, the terminal
power required to obtain the lowest compatible HSUPA bearer exceeds the maximum terminal power, there are no
more channel elements available, the maximum Iub backhaul throughput allowed on the site in the uplink is
exceeded, the lowest compatible HSUPA bearer they can obtain does not provide a peak RLC throughput higher than
the minimum throughput demand (only for CBR and VBR service users).
4.3.3.2.1 Statistics Tab

In the Statistics tab, Atoll displays as results:
• The number of rejected users.
269
• The number of delayed users.

• The number of R99 bearer users connected to a cell (result of the R99 part). This figure includes R99 users as well as
HSDPA and HSPA users since all of them request an R99 bearer.
• The number of R99 bearer users per frequency band.
• The number of R99 bearer users per activity status.
R99 R99
• The downlink and uplink peak throughputs ( TP P – D L and TP P – U L ) generated by their connection to R99 bearers.
Only active users are considered.
 
R99 R99 R99 R99
TP P –D L = TP P – DL  R99 Bearer  and TP P – U L = TP P – UL  R99 Bearer 
Active Active
users users
R99 R99
TP P – DL  R99 Bearer  is the downlink peak throughput of the user R99 radio bearer and TP P – UL  R99 Bearer  is the uplink
peak throughput of the user R99 radio bearer.
• The number of connected users with an HSDPA bearer (result of the HSDPA part) and the downlink peak RLC
throughput they generate. HSDPA and HSPA service users are considered since they all request an HSDPA bearer. On
DL
the other hand, only active users are taken into consideration in the downlink throughput calculation ( TP HSDPA ).

DL DL
TP HSDPA = TP P – RLC
Active
users
DL
TP P – RLC is the peak RLC throughput provided in the downlink.
• The number of connected users with an HSUPA bearer (result of the HSUPA part). Only HSPA service users are
considered.
In addition, Atoll indicates the uplink peak RLC throughput generated by active users connected with an HSUPA bearer
UL
( TP HSUPA ):

UL UL
TP HSUPA = TP P – RLC
Active
users
UL
TP P – RLC is the peak RLC throughput provided in the uplink.
4.3.3.2.2Mobiles Tab
In the Mobiles tab, Atoll indicates for each user:
UL DL
• The uplink and downlink total requested throughputs in kbps (respectively, TP requested  M b  and TP requested  M b  )
For R99 users, the DL and UL total requested throughputs correspond to the DL and UL peak throughputs of the R99 bearer
associated to the service.
DL R99
TP requested  M b  = TP P – DL  R99 Bearer 
UL R99
TP requested  M b  = TP P – UL  R99 Bearer 
For HSDPA users, the uplink requested throughput corresponds to the peak throughput of ADPCH R99 radio bearer and the
downlink requested throughput is the sum of the ADPCH radio bearer peak throughput and the peak RLC throughput(s) that
the selected HSDPA radio bearer(s) can provide. Here, the user is treated as if he is the only user in the cell and then, Atoll
determines the HSDPA bearer the user would obtain by considering the entire HSDPA power available of the cell.
DL R99 DL
TP requested  M b  = TP P – DL  ADPCH R99 Bearer  + TP P – RLC for single-carrier users

DL R99 DL
TP requested  M b  = TP P – DL  ADPCH R99 Bearer  AnchorCell + TP P – RLC  c  for dual-carrier users
c  Serving Cells
UL R99
TP requested  M b  = TP P – UL  ADPCH R99 Bearer 
For HSPA users, the uplink requested throughput is equal to the sum of the ADPCH-EDPCCH radio bearer peak throughput and
the peak RLC throughput of the requested HSUPA radio bearer. The requested HSUPA radio bearer is selected from the HSUPA
270
bearers compatible with the user equipment. Here, the user is treated as if he is the only user in the cell and then, Atoll
determines the HSUPA bearer the user would obtain by considering the entire remaining load of the cell. The downlink
requested throughput is the sum of the ADPCH-EDPCCH radio bearer peak throughput and the peak RLC throughput(s) that
the requested HSDPA radio bearer(s) can provide. The requested HSDPA radio bearer is determined as explained in the
previous paragraph.
DL R99 DL
TP requested  M b  = TP P – DL  ADPCH – EDPCCH R99 Bearer  + TP P – RLC for single-carrier users

DL R99 DL
TP requested  M b  = TP P – DL  ADPCH – EDPCCH R99 Bearer  AnchorCell + TP P – RLC  c  for dual-carrier users
c  Serving cells
UL R99 UL
TP requested  M b  = TP P – UL  ADPCH – EDPCCH R99 Bearer  + TP P – RLC
UL DL
• The uplink and downlink total obtained throughputs in kbps (respectively, TP obtained  M b  and TP obtained  M b  )
For R99 service users, the obtained throughput is the same as the requested throughput if he is connected without being
downgraded. Otherwise, the obtained throughput is lower (it corresponds to the peak throughput of the selected R99 bearer).
If the user is rejected, the obtained throughput is zero.
In the downlink, HSDPA bearer users can be connected to a single cell or to two adjacent cells of the same transmitter when
the user has a DC-HSDPA-capable terminal and when the transmitter supports the multi-cell HSDPA mode.
For a single-carrier HSDPA service user connected to an HSDPA bearer, the downlink obtained throughput corresponds to the
instantaneous throughput; this is the sum of the A-DPCH radio bearer peak throughput and the peak RLC throughput provided
by the selected HSDPA radio bearer after scheduling and radio resource control. If the user is delayed (he is only connected
to an R99 radio bearer), downlink obtained throughput corresponds to the downlink peak throughput of the ADPCH radio
bearer. Finally, if the user is rejected either in the R99 part or in the HSDPA part (i.e., because the HSDPA scheduler is
saturated), the downlink obtained throughput is zero.
For a dual-carrier HSDPA service user connected to two HSDPA bearers, the downlink obtained throughput corresponds to
the instantaneous throughput; this is the sum of the peak throughput provided by the A-DPCH radio bearer in the anchor cell
and the peak RLC throughputs provided by the selected HSDPA radio bearers after scheduling and radio resource control. If
the user is connected to one cell and delayed in the other cell, the downlink obtained throughput is the sum of the peak
throughput provided by the A-DPCH radio bearer in the anchor cell and the peak RLC throughput provided by the selected
HSDPA radio bearer after scheduling and radio resource control. If the user is delayed in the two cells (he is only connected
to an R99 radio bearer in the anchor cell), the downlink obtained throughput corresponds to the downlink peak throughput
of the ADPCH radio bearer in the anchor cell. Finally, if the user is rejected either in the R99 part or in the HSDPA part (i.e.,
because the HSDPA scheduler is saturated), the downlink obtained throughput is zero.
In the uplink, HSDPA service users can only have a single-carrier connection. When the user is either connected or delayed,
the uplink obtained throughput corresponds to the uplink peak throughput of the ADPCH radio bearer. If the user is rejected
either in the R99 part or in the HSDPA part (i.e., because the HSDPA scheduler is saturated), the uplink obtained throughput
is zero.
For single-carrier HSPA VBR and BE service users, on downlink, if the user is connected to an HSDPA bearer, the downlink
obtained throughput corresponds to the instantaneous throughput. The instantaneous throughput is the sum of the ADPCH-
EDPCCH radio bearer peak throughput and the peak RLC throughput provided by the selected HSDPA radio bearer after
scheduling and radio resource control. If the user is delayed, the downlink obtained throughput corresponds to the downlink
peak throughput of ADPCH-EDPCCH radio bearer. If the user is rejected, the downlink obtained throughput is "0".
For dual-carrier HSPA VBR and BE service users connected to two HSDPA bearers, the downlink obtained throughput
corresponds to the instantaneous throughput; this is the sum of the peak throughput provided by the ADPCH-EDPCCH radio
bearer in the anchor cell and the peak RLC throughputs provided by the selected HSDPA radio bearers after scheduling and
radio resource control. If the user is connected to one cell and delayed in the other cell, the downlink obtained throughput is
the sum of the peak throughput provided by the ADPCH-EDPCCH radio bearer in the anchor cell and the peak RLC throughput
provided by the selected HSDPA radio bearer after scheduling and radio resource control. If the user is delayed in the two cells
(he is only connected to an R99 radio bearer in the anchor cell), the downlink obtained throughput corresponds to the
downlink peak throughput of the ADPCH-EDPCCH radio bearer in the anchor cell. Finally, if the user is rejected, the downlink
obtained throughput is zero.
In uplink, HSPA VBR and BE service users can only have a single-carrier connection. When the user is connected to an HSUPA
bearer, the uplink obtained throughput is the sum of the ADPCH-EDPCCH radio bearer peak throughput and the peak RLC
throughput provided by the selected HSUPA radio bearer after noise rise scheduling. If the user is rejected, the uplink obtained
throughput is zero.
For a connected HSPA CBR service user, the uplink and downlink total obtained throughputs are the sum of the ADPCH-
EDPCCH radio bearer peak throughput and the minimum throughput demand defined for the service. If the user is rejected,
the uplink and downlink total obtained throughputs are "0".
• The mobile total power ( P term )
271
UL
P term = P term – R99  f act –EDPCCH + P term – HSUPA for HSPA VBR and BE service users.
UL
P term = P term – R99  f act –EDPCCH + P term – HSUPA  C HSDPABearer for HSPA CBR service users.
UL
For HSPA CBR service users, f act –EDPCCH = 0.1 .
And
P term = P term – R99 for R99 and HSDPA users.
DL
• The HSDPA application throughput in kbps ( TP A  M b  )
This is the net HSDPA throughput without coding (redundancy, overhead, addressing, etc.).

DL
TP P – RLC  c    1 – BLER HSDPA 
DL c  Serving cells
TP A  M b  = --------------------------------------------------------------------------------------------------------  f TP – Scaling – TP Offset
TTI
Where:
DL
TP P – RLC is the peak RLC throughput provided to the user by the selected HSDPA radio bearer after scheduling and radio
resource control.
BLER HSDPA is read in the quality graph defined for the triplet “reception equipment-selected bearer-mobility” (HSDPA Quality
Graphs tab in the Reception equipment properties). This graph describes the variation of BLER as a function of the measured
quality (HS-PDSCH Ec/Nt). Knowing the HS-PDSCH Ec/Nt, Atoll calculates the corresponding BLER.
TP Offset and f TP – Scaling represent the scaling factor between the application throughput and the RLC (Radio Link Control)
throughput and the throughput offset respectively. These two parameters model the header information and other
supplementary data that does not appear at the application level. They are defined in the service properties.
TTI is the minimum number of TTI (Transmission Time Interval) between two TTI used; it is defined in the terminal user
equipment category properties.
• The number of OVSF codes
This is the number of 512-bit length OVSF codes consumed by the user.
• The required HSDPA power in dBm (  P HSDPA  required )
It corresponds to the HSDPA power required to provide the HSDPA bearer user with the downlink requested throughput. The
downlink requested throughput is the throughput the user would obtain if he was the only user in the cell. In this case, Atoll
determines the HSDPA bearer the user would obtain by considering the entire HSDPA power available of the cell.
 P HSDPA  required =  P HS – PDSCH  used + n HS – SCCH  P HS – SCCH
 P HS – PDSCH  used is the HS-PDSCH power required to obtain the selected HSDPA bearer (in dBm). If the HSDPA bearer
allocated to the user is the best one,  P HS – PDSCH used corresponds to the available HS-PDSCH power of the cell. On the other
hand, if the HSDPA bearer has been downgraded in order to be compliant with cell and UE capabilities or for another reason,
 P HS – PDSCH  used will be lower than the available HS-PDSCH power of the cell.
• The served HSDPA power in dBm (  P HSDPA  served )
This is the HSDPA power required to provide the HSDPA bearer user with the downlink obtained throughput. The downlink
obtained rate is the throughput experienced by the user after scheduling and radio resource control.
 P HSDPA  served =  P HS – PDSCH  used + n HS – SCCH  P HS – SCCH for HSDPA users, HSPA BE and VBR service users.
And
 P HSDPA  served =  P HS – PDSCH  used  C HSDPABearer for HSPA CBR service users
Where
 P HS – PDSCH  used is the HS-PDSCH power required to obtain the selected HSDPA bearer.
• The No. of HSUPA Retransmissions (Required)
272
The maximum number of retransmissions in order to have the requested HSUPA radio bearer with a given BLER.
• The No. of HSUPA Retransmissions (Obtained)
The maximum number of retransmissions in order to have the obtained HSUPA radio bearer with a given BLER.
UL
• The HSUPA application throughput in kbps ( TP A  M b  )
This is the net HSUPA throughput without coding (redundancy, overhead, addressing, etc.).
UL
UL TP P – RLC  M b    1 – BLER HSUPA   f TP – Scaling – TP Offset
TP A  M b  = ----------------------------------------------------------------------------------------------------------------------------------------
N Rtx
Where:
UL
TP P – RLC is the peak RLC throughput provided by the selected HSUPA radio bearer after noise rise scheduling.
BLER HSUPA is the residual BLER after N Rtx retransmissions. It is read in the quality graph defined for the quartet “reception
equipment-selected bearer-number of retransmissions-mobility” (HSUPA Quality Graphs tab in the Reception equipment
properties). This graph describes the variation of BLER as a function of the measured quality (E-DPDCH Ec/Nt). Knowing the E-
DPDCH Ec/Nt, Atoll calculates the corresponding BLER.
TP Offset and f TP – Scaling respectively represent the scaling factor between the application throughput and the RLC (Radio Link
Control) throughput and the throughput offset. These two parameters model the header information and other
N Rtx is the maximum number of retransmissions for the obtained HSUPA bearer. This figure is read in the HSUPA Bearer
Selection table.
The following columns appear if, when creating the simulation, you select "Detailed information about mobiles":
• The uplink and downlink requested peak RLC throughputs (kbps)
Downlink and uplink requested peak RLC throughputs are not calculated for R99 users.
For HSDPA users, the uplink peak RLC throughput is not calculated and the downlink requested peak RLC throughput is the
throughput that the selected HSDPA radio bearer(s) can provide. Here, the user is treated as if he is the only user in the cell
and then, Atoll determines the HSDPA bearer he would obtain by considering the entire HSDPA power available of the cell.
For HSPA users, the requested uplink peak RLC throughput is the throughput of the requested HSUPA radio bearer. The
requested HSUPA radio bearer is selected from the HSUPA bearers compatible with the user equipment. Here, the user is
treated as if he is the only user in the cell and then, Atoll determines the HSUPA bearer the user would obtain by considering
the entire remaining load of the cell. If the user is connected to one or two HSDPA bearers in the downlink, the downlink
requested peak RLC throughput is the throughput that the requested HSDPA radio bearer(s) can provide. The requested
HSDPA radio bearer is determined as explained in the previous paragraph.
• The uplink and downlink obtained peak RLC throughput (kbps)
Downlink and uplink obtained peak RLC throughputs are not calculated for R99 users.
For HSDPA users connected to one or two HSDPA bearers, the uplink obtained peak RLC throughput is not calculated, and the
downlink obtained peak RLC throughput is the throughput provided by the selected HSDPA radio bearer(s) after scheduling
and radio resource control.
For connected HSPA BE and VBR service users, on uplink, if the user is connected to an HSUPA bearer, the obtained uplink
peak RLC throughput is the throughput provided by the selected HSUPA radio bearer after noise rise scheduling. On downlink,
if the user is connected to one or two HSDPA bearers, the downlink obtained peak RLC throughput is the throughput provided
by the selected HSDPA radio bearer(s) after scheduling and radio resource control.
For a connected HSPA CBR service user, the uplink and downlink obtained peak RLC throughputs are the uplink and downlink
minimum throughput demands defined for the service.
4.3.3.2.3 Cells Tab

In the Cells tab, Atoll gives:
• The available HSDPA power in the cell, c, in dBm ( P HSDPA  c  ):
This is:
• Either a fixed value in case of a static HSDPA power allocation strategy,
• Or a simulation result when the option "HSDPA Power Dynamic Allocation" is selected. We have:
P HSDPA  c  = P max  c  – P Headroom  c  – P tx – R99  c  – P HSUPA  c 
273
 
DL
with P tx – R99  c  = P pilot  c  + P SCH  c  + P OtherCCH  c  + P tch  c  + P tch  c   f act –ADPCH
tch used for tch used for
R99 users HSPA users
• The transmitted HSDPA power in the cell, c, in dBm ( P tx –H SDPA  c  ):
It corresponds to the HSDPA power used to serve HSDPA bearer users.
P tx –H SDPA  c  =   P HSDPA  M b   served

Mb  c
• The number of HSDPA users in the cell

They are the connected and delayed HSDPA bearer users. HSDPA and HSPA users are considered since they all request an
HSDPA bearer. DC-HSDPA users are accounted for once in each serving cell.
• The number of simultaneous HSDPA users in the cell ( n M )
b
It corresponds to the number of connected HSDPA bearer users that the cell supports at a time, i.e. within one transmission
time interval. All these users are connected to the cell at the end of the HSDPA part of the simulation; they have a connection
with the R99 bearer and an HSDPA bearer. DC-HSDPA users are accounted for once in each serving cell.
DL
• The instantaneous HSDPA throughput in the cell, c, in kbps ( TP Inst  c  )
This is the number of kilobits per second that the cell supports on downlink to provide simultaneous connected HSDPA bearer
users with an HSDPA bearer. We will differentiate single-carrier users (Ms) from DC-HSDPA users (Md-HSDPA stands for HSDPA
BE and VBR users, and Md-HSPA refers to HSPA BE and VBR service users).
DL R99 DL
TP requested  M b  = TP P – DL  ADPCH R99 Bearer  + TP P – RLC
 
DL R99 DL
TP obtained  M s  +  TP P – DL  R99 Bearer  + TP P – RLC  M d – HSDPA  
Ms  c
M d – HSDPA  c
c is the anchor cell

DL
+ TP P – RLC  M d – HSDPA  +
M c
d – HSDPA
DL
TP Inst  cell  = c is the secondary cell

R99 DL
 TP P – DL  R99 Bearer  + TP P – RLC  M d – HSPA   +
M d – HSPA  c
c is the anchor cell

DL
TP P – RLC  M d – HSPA 
M d – HSPA  c
c is the secondary cell
DL
TP P – RLC is the peak RLC throughput provided by the selected HSDPA radio bearer after scheduling and radio resource
control.
R99
TP P – DL  R99 Bearer  is the peak throughput of the ADPCH radio bearer if the user is an HSDPA user. For HSPA users, it
corresponds to the ADPCH-EDPCCH radio bearer.
DL
• The instantaneous HSDPA Effective MAC Throughput in the cell, c, in kbps ( TPE –M AC  c  )
S block  M b 

DL
TP E – M AC  c  = ---------------------------------------
-
T TTI   TTI  M b 
Mb  c
Where,
S block  M b  is the transport block size (in kbits) of the HSDPA bearer selected by the user; it is defined for each HSDPA bearer
in the HSDPA Radio Bearers table.
TTI  M b  is the minimum number of TTI (Transmission Time Interval) between two TTI used; it is defined in the terminal user
274
–3
T TTI is the TTI duration, i.e. 2 10 s (2000 TTI in one second). This value is specified by the 3GPP.
DL
• The average instantaneous HSDPA throughput in the cell, c, in kbps ( TP Av – Inst  c  )
DL
DL TP Inst  c 
TP Av – Inst  c  = --------------------
-
nM
b
DL
• The HSDPA application throughput in the cell, c, in kbps ( TP A  c  )
DL
TP P – RLC  M b    1 – BLER HSDPA   f TP – Scaling – TP Offset

DL
Either TP A  c  = -------------------------------------------------------------------------------------------------------------------------------------- if the scheduling algorithm is Round Robin or
TTI
M c
b
Proportional Fair,
DL
DL TP P – RLC  M b  maxC  I     1 – BLER HSDPA   f TP – Scaling – TP Offset
Or TPA  c  = --------------------------------------------------------------------------------------------------------------------------------------------------------------
- if the scheduling algorithm is Max C/I.
TTI
M b  maxC  I  is the user with the highest C  I in the cell.
DL
TP P – RLC is the peak RLC throughput provided by the selected HSDPA radio bearer after scheduling and radio resource
control.
quality (HS-PDSCH Ec/Nt). Knowing the HS-PDSCH Ec/Nt, Atoll calculates the corresponding BLER.
f TP – Scaling and TP Offset respectively represent the scaling factor between the application throughput and the RLC (Radio Link
DL
• The minimum HSDPA RLC peak throughput in kbps ( min  TP P – RLC  M b   )
M b  cell
It corresponds to the lowest of RLC peak throughputs obtained by HSDPA bearer users connected to the cell.
DL
• The maximum HSDPA RLC peak throughput in kbps ( max  TP P – RLC  M b   )
M b  cell
It corresponds to the highest of RLC peak throughputs obtained by HSDPA bearer users connected to the cell.
• The number of HSUPA users in the cell ( n M ):
c
They are the HSUPA bearer users connected to the cell.

UL
• The HSUPA application throughput in the cell, c, in kbps ( TP A  c  )

UL UL
TP A  c  = TP A  M b 
Mb  c
UL
• The uplink cell load factor due to HSUPA traffic ( X HSUPA  c  ):
UL
UL  I tot  c   HSUPA
X HSUPA  c  = ---------------------------------
UL
N tot  c 
Where
UL
 I tot  c   HSUPA is the total interference at transmitter received from HSUPA bearer users.
4.3.3.2.4 Sites Tab

In the Sites tab, Atoll displays:
DL
• The instantaneous HSDPA throughput carried by the site in kbps ( TP Inst  site  )
275

DL DL
TP Inst  site  = TP Inst  c 
c  site
DL
• The instantaneous HSDPA Effective MAC Throughput carried by the site in kbps ( T MAC  site  in kbps)

DL DL
TP E – M AC  site  = TP E –M AC  c 
c  site
UL
• The HSUPA throughput carried by the site in kbps ( TP  site  )

UL UL
TP  site  = TP obtained  M c 
M c  site
4.3.4 Appendices
4.3.4.1 Admission Control in the R99 Part
During admission control in the R99 part of the simulation, Atoll calculates the uplink load factor of a considered cell assuming
the mobile concerned is connected to it. Here, activity status assigned to users is not taken into account. So even if the mobile
is not active on UL, it can be rejected due to cell load saturation. To calculate the cell UL load factor, either Atoll takes into
account the mobile power determined during power control if mobile was connected in previous iteration, or it estimates a
UL
load rise due to the mobile and adds it to the current load. The load rise ( X ) is calculated as follows:
UL 1
X = ----------------------------------------------
W
1 + ------------------------------------
UL UL
Q req  R nominal
4.3.4.2 Resources Management

4.3.4.2.1 OVSF Codes Management
OVSF codes are managed in the downlink during the simulation since this resource is downlink limited only. Atoll checks the
availability of this resource during the simulation, first in the R99 part and then in the HSDPA part. It determines the number
of codes that will be consumed by each cell.
OVSF codes form a binary tree. Codes of longer lengths are generated from codes of a shorter length. Length-k OVSF codes
are generated from length-k/2 OVSF codes. Therefore, if one channel needs 1 length-k/2 OVSF code, it is equivalent to use 2
length-k OVSF codes, or 4 length-2k OVSF codes and so on.
512 512-bit-length codes per cell are available in UMTS HSPA projects.
In the R99 part, during the resource control, Atoll determines the number of 512 bit-length codes that will be consumed for
each cell.
If the cell supports HSPA, Atoll allocates codes for the DL channels used for HSUPA:
• A 128 bit-length code for the E-HICH and E-RGCH channels (i.e. four 512 bit-length OVSF codes), for each cell.
Therefore, Atoll will take four 512-bit-length codes,
• A 256 bit-length code for the E-AGCH channel (i.e. two 512 bit-length OVSF codes), for each cell. Therefore, Atoll will
take two 512-bit-length codes,
If the cell supports HSDPA, Atoll reserves for potential HSDPA bearer users:
HS – PDSCH – Min
• The minimum number of HS-PDSCH codes defined for the cell, N Codes . They are 16-bit length OVSF codes
(i.e. thirty-two 512 bit-length OVSF codes). Therefore, Atoll will take 32  N Codes 512-bit-length codes,
• A 128 bit-length code per HS-SCCH channel (i.e. four 512 bit-length OVSF codes), for each cell. Therefore, Atoll will
take 4  n HS – SCCH 512-bit-length codes,
Then, it allocates to the cell OVSF codes to support R99 bearers required by users:
• A 256 bit-length code per common channel (i.e. two 512 bit-length OVSF codes), for each cell. Therefore, Atoll will
Overhead
take 2  N Codes 512-bit-length codes,
• A code per cell-receiver link, for TCH (traffic channels). The length of code to be allocated, Code_Length, depends on
the user activity. We have:
276
DL
Either Code_Length = F spreading  Active user  when the user is active,
DL
Or Code_Length = F spreading  Inactive user  if the user is inactive.
TCH
The number of 512 bit-length OVSF codes needed N Codes is calculated from the length of the code to be allocated as follows:
TCH 512
N Codes = -------------------------------
Code_Length
Figure 4.15: OVSF Code Tree Indices (Not OVSF Code Numbers)
The OVSF code allocation follows the “Buddy” algorithm, which guarantees that:
• If a k-length OVSF code is used, all of its children with lengths 2k, 4k, …, cannot be used as they will not be orthogonal.
• If a k-length OVSF code is used, all of its ancestors with lengths k/2, k/4, …, cannot be used as they will not be
orthogonal.
Example: We consider a user with a service requiring the UDD64 R99 radio bearer. This user is active on DL while connected
to a cell (which does not support HSDPA). The spreading factor for active users has been set to 64 and site equipment requires
four overhead downlink channel elements per cell. Atoll will consume four 256 bit-length OVSF codes for common channels
(i.e. eight 512 bit-length OVSF codes) and a 64 bit-length OVSF code for traffic channels (i.e. eight additional 512 bit-length
OVSF codes).
• In the R99 part, the OVSF code allocation follows the mobile connection order (mobile
order in the Mobiles tab).
• In DC-HSDPA, A-DPCH is only transmitted in the anchor carrier. Therefore, a DC-
HSDPA user requires R99 resources in the best serving cell only and consumes the
same amount of R99 resources as a single-cell HSDPA user.
• The OVSF code and channel element management is differently dealt with in case of
“softer” handover. Atoll allocates OVSF codes for each cell-mobile link while it
globally assigns channel elements to a site.
In the HSDPA part, HSDPA and HSPA users are assigned an HSDPA bearer (Fast link adaptation).
Therefore, Atoll allocates to the cell:
• 16-bit length OVSF codes per cell-receiver, for HS-PDSCH. This figure depends on the HSDPA bearer assigned to the
user and on the type of service.
HS – PDSCH
For HSDPA users, HSPA VBR and BE service users, Atoll needs 32  N Codes 512-bit-length codes for each user
HS – PDSCH
connected to the cell. N Codes is the number of HS-PDSCH channels required by the HSDPA bearer.
HS – PDSCH
For HSPA CBR service users, Atoll needs 32  N Codes  C HSDPABearer 512-bit-length codes for each user connected
HS – PDSCH
to the cell. N Codes is the number of HS-PDSCH channels required by the HSDPA bearer.
DC-HSDPA users have two HSDPA bearers, one for each serving cell. Therefore, one DC-HSDPA user consumes OVSF
codes in both cells.
277
When HSDPA bearer users (at least one) are connected to the cell, Atoll gives the cell
back the minimum number of OVSF codes reserved for HS-PDSCH ( N Codes ). On
the other hand, if no HSDPA bearer user is connected, Atoll still keeps these codes and
the codes for HS-SCCH too. This is the same with HSUPA bearer users. Even if no HSUPA
bearer user is connected to the cell, Atoll still keeps the codes for E-HICH, E-RGCH and E-
AGCH channels.
4.3.4.2.2 Channel Elements Management

Channel elements are controlled in the R99 and the HSUPA parts of the simulation. Atoll checks the availability of this resource
in the uplink and downlink.
In the R99 part, during the resource control, Atoll determines the number of channel elements required by each site for R99
bearers in the uplink and downlink. Then, in the HSUPA part, Atoll carries out another resource control after allocating HSUPA
bearers. It takes into account the channel elements consumed by HSUPA bearer users in the uplink and recalculates the
number of channel elements required by each site in the uplink.
In the uplink, Atoll consumes N CE – UL  j  channel elements for each cell j on a site NI. This figure includes:
• Channel elements for R99 bearers:

Overhead
• N CE – UL channel elements for control channels,
R99 – T CH
• N CE – UL per cell-receiver link, for R99 TCH (traffic channels).
• Channel elements for HSUPA bearers:
HSUPA
• N CE per cell-receiver link, for packet (HSPA - BE) and packet (HSPA - VBR) service users.
HSUPA
• N CE  C HSUPABearer per cell-receiver link, for CBR service users.
Therefore, the number of channel elements required in the uplink at the site level, N CE – UL  N I  , is:
N CE – UL  N I  =  NCE – UL  j 
j  NI
In the downlink, Atoll consumes N CE – DL  j  channel elements for each cell j on a site NI. This figure includes:
• Channel elements for R99 bearers

Overhead
• N CE – DL channel elements for control channels (Pilot channel, Synchronisation channel, common channels),
R99 – T CH
• N CE – DL per cell-receiver link, for R99 TCH (traffic channels).
Therefore, the number of channel elements required in the downlink at the site level, N CE – DL  N I  , is:
N CE – DL  N I  =  NCE – DL  j 
j  NI
• In DC-HSDPA, A-DPCH is only transmitted on the anchor carrier. Therefore, a DC-

same amount of R99 resources as a single-cell HSDPA user.
• In case of “softer” handover (the mobile has several links with co-site cells), Atoll
allocates channel elements for the best serving cell-mobile link only.
4.3.4.2.3 Iub Backhaul Throughput

The Iub backhaul throughput is controlled in the R99, the HSDPA and the HSUPA parts of the simulation. Atoll checks the
availability of this resource in the uplink and downlink.
In the R99 part, during the resource control, Atoll determines the Iub throughput required by each site for R99 bearers in the
uplink and downlink. Then, in the HSDPA part, Atoll performs a resource control in the downlink after allocating HSDPA
bearers. It takes into account the Iub backhaul throughput consumed by HSDPA bearer users in the downlink and recalculates
the Iub backhaul throughput required by each site in the downlink. Finally, in the HSUPA part, Atoll carries out a resource
control in the uplink after allocating HSUPA bearers. It takes into account the Iub backhaul throughput consumed by HSUPA
bearer users in the uplink and updates the Iub backhaul throughput required by each site in the uplink.
In the uplink, the Iub backhaul throughput consumed by each cell j on a site NI, TP Iub – UL  j  , includes:
278
• The Iub backhaul throughput required for R99 bearers:

R99 – T CH
• TP Iub – UL per cell-receiver link, for R99 TCH (traffic channels).
• The Iub backhaul throughput required for HSUPA bearers:
HSUPA
• TP Iub per cell-receiver link, for HSPA BE and VBR service users.
HSUPA
• TP Iub  C HSUPABearer per cell-receiver link, for HSPA CBR service users.
Therefore, the Iub backhaul throughput required on uplink at the site level, TP Iub – UL  N I  , is:
TP Iub – UL  N I  =  TPIub – UL  j 
j  NI
In the downlink, the Iub backhaul throughput consumed by each cell j on a site NI, TP Iub – DL  j  , includes:
• The Iub backhaul throughput required for R99 bearers:

Overhead
• TP Iub – DL for R99 control channels (Pilot channel, Synchronisation channel, common channels).
R99 – T CH
• TP Iub – DL per cell-receiver link, for R99 TCH (traffic channels).
• The Iub backhaul throughput required for HSDPA bearers:
HSDPA
• TP Iub per cell-receiver link, for HSDPA, HSPA BE and VBR service users.
HSDPA
• TP Iub  C HSDPABearer per cell-receiver link, for HSPA CBR service users.
HSDPA DL HSDPA DL
With TP Iub = TP P – RLC + Overhead Iub  TP P – RLC
Therefore, the Iub backhaul throughput required on downlink at the site level, TP Iub – DL  N I  , is:
TP Iub – DL  N I  =  TPIub – DL  j 
j  NI
• In DC-HSDPA, A-DPCH is only transmitted on the anchor carrier. Therefore, a DC-

same amount of R99 resources as a single-cell HSDPA user. On the other hand, the
DC-HSDPA user has two HSDPA bearers (one for each serving cell) and consumes
HSDPA resources in both cells.
• In case of “softer” handover (the mobile has several links with co-site cells), Iub
backhaul throughput is consumed by the best serving cell-mobile link only.
4.3.4.3 Downlink Load Factor Calculation

Atoll calculates a downlink load factor for each cell (available in the Cells tab of any simulation result) and each connected
mobile (available in the Mobiles tab of any given simulation result).
4.3.4.3.1 Downlink Load Factor per Cell

Approach for downlink load factor evaluation is highly inspired by the downlink load factor defined in the book “WCDMA for
UMTS by Harry Holma and Antti Toskala”.
DL
Q req
Let CI req = ---------
- be the required quality.
DL
Gp
DL DL
G p and Q req are the processing gain on downlink and the Eb/Nt target on downlink respectively.
In case of soft-handoff, required quality is limited to the effective contribution of the transmitter.
 Ptch  c 
DL
P tx  c  = P pilot  c  + P SCH  c  + P otherCCH  c  +
tch
 Ptch  c 
DL ortho nonOrtho
P tx  c  = P CCH  c  + P CCH c +
tch
where
279
ortho
P CCH  c  = P pilot  c  + P otherCCH  c 
nonOrtho
P CCH  c  = P SCH  c 
At mobile level, we have a required power, Ptch:

term
P tch  c  = CI req   I extra  c  + I inter – carrier  c  + I inter – techno log y  c  + I intra  c  + N 0   LT  r
DL
With r = 1 when the user is active on the downlink and r = r c when the user is inactive. In case of an HSDPA bearer user,
DL
r = f act – ADPCH .
 I extra  c  + I inter – carrier  c  + I inter – techno log y  c  

 
P tch  c  = CI req    P tx  c  – P CCH
DL nonOrtho
 c  – P tch  c   P CCH c
nonOrtho  L r
 T
 - + ------------------------------ + N term
+  1 – F ortho   BTS   ------------------------------------------------------------------------------- 0 
 LT LT 
DL
 I extra  c  + I inter – carrier  c  + I inter – techno log y  c    L T  r +  1 – F ortho   BTS   P tx  c   r
+
nonOrtho term
F ortho   BTS  P CCH  c   r + N0  LT  r
P tch  ic  = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1
-------------------- +  1 – F ortho   BTS 
CIreq  r
I intra  c  is the total power received at the receiver from the cell with which it is connected.
I extra  c  is the total power received at the receiver from other cells.
I inter – carrier  c  is the inter-carrier interference received at the terminal.
I inter – techno log y  c  is the inter-technology interference received at the terminal from an external transmitter.
We have:
ortho nonOrtho
P CCH  c  + P CCH c
  I extra  c  + I inter – carrier  c  + I inter – techno log y  c    L T  r 

 
DL
P tx  c  =   + 1 – F    
DL
   +   
nonOrtho
   +
term
 

 ortho BTS P tx c r F ortho BTS P CCH c r N 0 L T r 
+ 

-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1
-------------------
- +  1 – F   
-

tch  
CI req  r ortho BTS
 
 
 I extra  c  + I inter – carrier  c  + I inter – techno log y  c    L T  r

-------------------------------------------------------------------------------------------------------------------------------------------  P DLtx  c 
DL
DL P tx  c 
P tx  c  = P ortho  c  + P nonOrtho  c  + ----------------------------------------------------------------------------------------------------------------------------------------------------------------
CCH CCH  - +
1
tch -------------------- +  1 – F ortho   BTS 
CI req  r
DL nonOrtho term
 1 – F ortho   BTS   P tx  c   r F ortho   BTS  P CCH  c   r + N0  LT  r
 --------------------------------------------------------------------------- +
1  -----------------------------------------------------------------------------------------------------------------
1 - 1 F
tch -------------------- +  1 – F ortho   BTS  tch ------------------- +  – ortho   BTS 
CI req  r CI req  r

------------------------------------------------------------------------------------------------------------------------------------------
- + 1 – F ortho   BTS  r 
 P
DL
 c  
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------   P DL  ic 

DL tx
P tx  c  –
 1 - + 1 – F  tx
tch  ------------------- ortho   BTS  
 CI req  r 
nonOrtho term

ortho nonOrtho
= P CCH  c  + P CCH c + -----------------------------------------------------------------------------------------------------------------
1
tch -------------------- +  1 – F ortho   BTS 
CI req  r
280
nonOrtho term

ortho nonOrtho
P CCH  c  + P CCH c + -----------------------------------------------------------------------------------------------------------------
1 - 1 F
tch ------------------- +  – ortho   BTS 
DL CIreq  r
P tx  c  = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------------------------------------------------
I extra  c  + I inter – carrier  c  + I inter – techno log y  c    L T  r
- + 1 – F ortho   BTS  r 
 P
DL
 c  
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

tx
1–
 1 - + 1 – F 
tch  ------------------- ortho   BTS  
 CI req  r 
Therefore, the downlink load factor can be expressed as:
------------------------------------------------------------------------------------------------------------------------------------------- + 1 – F ortho   BTS  r
DL
P tx  c 

DL
X = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
1 - + 1 – F
tch -------------------   
CI req  r ortho BTS
The downlink load factor represents the signal degradation in relation to the reference interference (thermal noise plus
synchronisation channel power).
4.3.4.3.2 Downlink Load Factor per Mobile

Atoll evaluates the downlink load factor for any connected mobile as follows:
DL
DL I tot  c 
X = ----------------
-
DL
N tot  c 
4.3.4.4 Uplink Load Factor Due to One User

UL
This part details how Atoll calculates the contribution of one user to the UL load factor ( X k ).
UL
In this calculation, we assume that the cell UL reuse factor ( F  txi ic  ) is constant.
The result depends on the option used to calculate Nt (Without useful signal or Total noise that you may select in Global
parameters).
Without Useful Signal Option
UL
UL  P b  k   req
W - ----------------------------------------------------------------------------------------------------------------
Q req  k  = -------------------------
R99
 UL
-
tx
TP P – UL  k  I intra –  P b  k   req + I extra + I inter – carrier + N 0
UL
UL W  P b  k   req
Q req  k  = -------------------------
R99
-  ------------------------------------------------------------------------
UL UL
-
tx
TP P – UL  k  I intra  F –  P b  k   req + N 0
R99 R99
UL  UL TP P – UL  k  TP P – UL  k 
- = Q UL
 P b  k   req   1 + Qreq  k   ------------------------- req  k   --------------------------   I intra  F
UL tx
+ N0 
 W  W
R99 R99
UL TP P – UL  k  TP P – UL  k 
Q req  k   ------------------------- -  I intra  F UL Q UL req  k   --------------------------  N 0
tx
UL W W
 P b  k   req = ----------------------------------------------------------------------------------- + -----------------------------------------------------------------
R99 R99
UL TP P – UL  k  UL TP P – UL  k 
1 + Q req  k   -------------------------- 1 + Q req  k   --------------------------
W W
R99
req TP P – UL  k 
We note  ------  k 
Ec UL
= Q req  k   -------------------------
-
 Nt  E – DPDCH W
UL tx
UL I intra  F N0
 P b  k   req = ------------------------------------------------------
- + ------------------------------------------------------
-
   
 1   1 
 ---------------------------------------- + 1  --------------------------------------- - + 1
  -----
Ec  req
   -----
Ec  req

  Nt-  k     Nt-  k  
E – DPDCH E – DPDCH
  Pb
UL
As I intra =  k   req , we have:
K
281
1 1
 ------------------------------------------------------
- + N 0   -------------------------------------------------------
UL tx
I intra = I intra  F 
   
K  1  K  1 
 ---------------------------------------
- + 1  ---------------------------------------
- + 1
  -----
Ec-  req    -----
Ec-  req 
  Nt  k     Nt  k  
E – DPDCH E – DPDCH
1
 ------------------------------------------------------
tx
N0  -
 
K  1 
 ---------------------------------------- + 1
  -----
Ec  req

  Nt-  k  
E – DPDCH
I intra = --------------------------------------------------------------------------------------
1

UL
1–F  -------------------------------------------------------
 
K  1 
 ---------------------------------------- + 1
  -----
Ec  req

  Nt-  k  
E – DPDCH
tx UL
N0  F
I intra = --------------------------------------------------------------------------------------
-
1
----------------------------------------------------------------------------- –1
1

UL
F  -------------------------------------------------------
 
K  1 
 ---------------------------------------- + 1
  -----
Ec  req

  Nt-  k  
E – DPDCH
UL
UL I intra + I extra + I inter – carrier I intra  F 1
X = ------------------------------------------------------------------------------- = --------------------------------------- = ----------------------------------
-
tx UL tx tx
I intra + I extra + I inter – carrier + N 0 I intra  F + N 0 N0
1 + -------------------------
UL
I intra  F
Therefore, we have:
1
 ------------------------------------------------------
UL UL
X = F  -
 
K  1 
 ---------------------------------------
- + 1
  -----
Ec  req

  Nt-  k  
E – DPDCH
So, we can conclude that the contribution of one user to the UL load is defined as:
UL UL 1
X k = F  -------------------------------------------------------
 
 1 
 ---------------------------------------- + 1
  -----
Ec  req

  Nt-  k  
E – DPDCH
Total Noise Option
UL
UL W  P b  k   req
Q req  k  = -------------------------
R99
-  -------------------------------------------------------------------------------
tx
TP P – UL  k  I intra + I extra + I inter – carrier + N 0
UL
UL  P b  k   req
W - ---------------------------------------
Q req  k  = -------------------------
R99
 UL tx
TP P – UL  k  I intra  F + N 0
R99
UL UL TP P – UL  k 
-   I intra  F UL + N tx
 P b  k   req = Q req  k   ------------------------- 0 
W
R99
req TP P – UL  k 
We note  ------  k 
Ec UL
= Q req  k   -------------------------
-
 Nt  E – DPDCH W
req
 P b  k   req =  ------  k 
UL Ec UL tx
  I intra  F + N 0 
 Nt  E – DPDCH
282
  Pb
UL
As I intra =  k   req , we have:
K
req
  -----
-  k 
UL tx Ec
I intra =  I intra  F + N0  
Nt  E – DPDCH
K
req
  -----
-  k 
tx Ec
N0 
Nt  E – DPDCH
K
I intra = -------------------------------------------------------------
-
UL
1–F
UL
UL I intra + I extra + I inter – carrier I intra  F 1
X = ------------------------------------------------------------------------------- = --------------------------------------- = ----------------------------------
-
tx UL tx tx
I intra + I extra + I inter – carrier + N 0 I intra  F + N 0 N0
1 + -------------------------
UL
I intra  F
Therefore, we have:
req
 -----
Ec- 

UL UL
X = F  k
 Nt  E – DPDCH
K
So, we can conclude that the contribution of one user to the UL load is defined as:
req
  ------  k 
UL UL Ec
X k = F
Nt E – DPDCH
4.3.4.5 Inter-carrier Power Sharing Modelling

Inter-carrier power sharing enables the network to dynamically allocate available power from R99-only and HSDPA carriers
among HSDPA carriers.
In this part, we will consider the most common scenario, a network consisting of an R99-only carrier (c1) and an HSDPA carrier
with dynamic power allocation (c2) (c2 does not support HSUPA).
As explained in The User Manual, the maximum power of the HSDPA cell must be set to the same value as the maximum
shared power in order to use power sharing efficiently. In this case, the HSDPA cell can use 100% of the available power, i.e,
all of the R99-only cell’s unused power can be allocated to the HSDPA cell.
Let’s take the following example to measure the impact of the inter-carrier power sharing.
• 1st case: Inter-carrier power sharing is not activated

On c1, we have: P max  Tx c 1  = 43dBm and P tx – R99  Tx c 1  = 39.1dBm .
On c2, we have: P max  Tx c 2  = 43dBm , P tx – R99  Tx c 2  = 36.1dBm and P Headroom  Tx c 2  = 0dB .
Therefore, P HSDPA  Tx c 2  = P max  Tx c 2  – P tx – R99  Tx c 2  – P Headroom  Tx c 2  = 42dBm
• 2nd case: Inter-carrier power sharing is activated and P max  Tx  = 46dBm
On c1, we have: P max  Tx c 1  = 43dBm and P tx – R99  Tx c 1  = 39.1dBm .
On c2, we have: P max  Tx c 2  = 46dBm , P tx – R99  Tx c 2  = 36.1dBm and P Headroom  Tx c 2  = 0dB .
Therefore, P HSDPA  Tx c 2  = P max  Tx  – P tx – R99  Tx c 1  – P tx – R99  Tx c 2  – P Headroom  Tx c 2  = 44.4dBm
4.3.4.6 Best Serving Cell Determination in Monte Carlo Simulations - Old Method
Before Atoll 2.8.0, best serving cell determination used to be performed by selecting the best carrier within transmitters
according to the selected method (site equipment) and then the best transmitter using the best carrier. To switch back to this
method, add the following lines in the Atoll.ini file:
[CDMA]
MultiBandSimu = 0
The method is described below:
283
For each station txi containing Mb in its calculation area and using a frequency band supported by the Mb’s terminal.
Determination of BestCarrier k  txi M b  .
If a given carrier is specified for the service requested by Mb and if it is used by txi
BestCarrier k  txi M b  is the carrier specified for the service.
Else the carrier selection mode defined for txi is considered.

For each carrier ic used by txi, we calculate current loading factor:
UL
UL I tot  txi ic  UL
- + X
X k  txi ic  = ----------------------------
UL
EndFor
UL
BestCarrier k  txi M b  is the carrier with the lowest X k  txi ic 

BestCarrier k  txi M b  is the carrier with the lowest P tx  txi ic  k

BestCarrier k  txi M b  is randomly selected

UL UL
BestCarrier k  txi M b  is the first carrier so that X k  txi ic   X max
Calculation of
   BTS  P c  txi M b BestCarrier 
Q pilot  txi BestCarrier  = --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
k
 DL DL 
 P tot  txi BestCarrier k  txi M b   + I extra  BestCarrier k  txi M b   + 
 DL DL Term 
 inter – carrier
I  BestCarrier k  txi  M b   + I inter – techno log y  BestCarrier k  txi  M b   + N 0 

   BTS  P c  txi M b BestCarrier 
Q pilot  txi BestCarrier  = ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
k
 DL DL 
 P tot  txi BestCarrier k  txi M b   + I extra  BestCarrier k  txi M b   
 
 + I DL  BestCarrier k  txi M b   + I inter – techno log y  BestCarrier k  txi M b   
DL
 inter – carrier 
 + N
Term
–  1 –      P  txi  M  BestCarrier  
 0 BTS c b 
Rejection of station txi if the pilot is not received

pilot
If Q pilot  txi M b BestCarrier   Q req  Mobility  M b   then txi is rejected by Mb
k
max
If Q pilot  txi M b BestCarrier   Q pilot  M b 
k k
Admission control (If simulation respects a loading factor constraint and Mb was not connected in previous iteration).
UL UL
If X k  txi BestCarrier  txi M b    X max , then txi is rejected by Mb
Else
max
Q pilot  M b  = Q pilot  txi M b BestCarrier 
k k
Tx BS  M b  = txi
Endif
EndFor
If no TxBS has been selected, Mb has failed to be connected to the network and is rejected.
284
4.4 UMTS HSPA Prediction Studies

4.4.1 Best Serving Cell and Active Set Determination
The mobile active set is the list of the cells to which the mobile is connected. The active set may consist of one or more cells
depending on whether the service supports soft handover and on the terminal active set size. The best serving cell and other
cells of the active set are selected among a list of potential serving cells which fulfil a set of conditions. Potential serving cells
must use a frequency band with which the terminal is compatible. They must also belong to layers supported by the service
and the terminal, and these layers must support a speed higher than the user mobility. In addition, the pilot signal level
received from these cells must exceed the defined minimum RSCP threshold.
The layer priority, the quality of the pilot ( Q pilot ), the handover margin ( M HO ) and the cell individual offset ( CIO ) are consid-
ered to rank all potential serving cells and determine the best serving cell.
Among all potential serving cells, Atoll first selects the cells which belong to the highest priority layer and then, the one with
the highest RSCP. This cell is referred to as the best serving cell candidate ( c BC ).
Then, Atollcalculates the best server indicator ( I BS ) for the best serving cell candidate and the other potential serving cells
( c OC ):
I BS  c BC  = Q pilot  c BC  + M HO  c BC  + C IO  c BC  for the best serving cell candidate,
I BS  c OC  = Q pilot  c OC  + C IO  c OC  for the other potential serving cells.
Atoll ranks the potential serving cells according to the best server indicator ( I BS ). The cell with the highest I BS is selected as
the best serving cell if its best server indicator ( I BS ) exceeds the Ec/I0 threshold defined in the properties of the mobility type.
Each other cell of the active set is selected among the other potential serving cells as follows:
• It must use the same carrier as the best serving cell.
• The pilot quality difference between the cell and the best serving cell must not exceed the AS-threshold set per cell.
• It must belong to the neighbour list of the best serving cell if it is located on a site where the equipment imposes this
restriction (the “restricted to neighbours” option selected in the equipment properties).
You can return to the old best serving cell selection mechanism as in Atoll 3.2.1, by setting
an option in the Atoll.ini file. For more information about setting options in the Atoll.ini file,
see the Administrator Manual.
4.4.2 Point Analysis - AS Analysis Tab

Let us suppose a receiver with a terminal, a service and a mobility type. This receiver does not create any interference. You
can make the prediction for a specific carrier, for all carriers, for a specific layer, or for all layers. For DC-HSDPA, MC-HSDPA
and DB-MC-HSDPA users, selecting one specific carrier or one layer associated with one unique carrier is not suitable. If you
have selected a DC-HSDPA user or a MC-HSDPA user, select "Best (All/Specific band)" as the carrier or layers associated with
several carriers. For a DB-MC-HSDPA user, select "Best (All bands)" as the carrier or layers associated with several carriers on
different frequency bands.
The analysis is based on the following parameters:
• The uplink load factor and the downlink total power of cells,
• The available HSDPA power of the cell in case of an HSDPA bearer user,
• The cell UL reuse factor, the cell UL load factor due to HSUPA and the maximum cell UL load factor for HSUPA bearer
users.
These parameters can be results of a given simulation, average values calculated from a group of simulations, or user-defined
cell inputs. In the last case, when no value is defined in the Cells table, Atoll uses the following default values:
• Total transmitted power = 50% of the maximum power (i.e, 40 dBm if the maximum power is set to 43 dBm)
• Uplink load factor = 50%.
• Uplink reuse factor = 1
• Uplink load factor due to HSUPA = 0%
• Maximum uplink load factor = 75%
On the other hand, no default value is used for the HSDPA power; this parameter must be defined by the user.
Results are displayed for any point of the map where the pilot signal level exceeds the defined minimum RSCP.
285
4.4.2.1 Bar Graph and Pilot Sub-Menu

Atoll performs a first selection of potential serving cells depending on if you have chosen "Carrier" or "Layer".
We can consider the following cases:
1st case: Analysis based on a specific carrier

The carrier that can be used is fixed. The potential serving cells are selected among all transmitters i that contain the receiver
in their calculation areas and that use the selected carrier ic.
2nd case: Analysis based on all carriers of all frequency bands/a specific frequency band
If you have selected "Best (All bands)", the potential serving cells are selected among all transmitters i that contain the
receiver in their calculation areas.
If the frequency band is fixed ("Best (Specific band)"), the potential serving cells are selected among all transmitters i that
contain the receiver in their calculation areas and that use a carrier of the selected frequency band.
3rd case: Analysis based on the best layer

The layer that can be used is fixed. The potential serving cells are selected among all transmitters i that contain the receiver
in their calculation areas and that have cells using the selected layer.
4th case: Analysis based on all layers

The potential serving cells are selected among all transmitters i that contain the receiver in their calculation areas.
In addition, potential serving cells must satisfy the following conditions:
• They must use a frequency band with which the terminal is compatible.
• They must also belong to layers supported by the service and the terminal, and these layers must support a speed
higher than the user mobility.
• The pilot signal level received from these cells must exceed the defined minimum RSCP threshold.
Ec/I0 (or Q pilot  ic  ) Evaluation
Atoll calculates the pilot quality for all potential serving cells (i, ic).
ic is the studied carrier and icadj is another carrier adjacent to ic. The interference reduction factor, RF  ic ic adj  , is defined
between ic and icadj and set to a value different from 0.
Two ways may be used to calculate I0.
Option Total noise: Atoll considers the noise generated by all the transmitters and the thermal noise.
Option Without pilot: Atoll considers the total noise deducting the pilot signal.
Calculation option may be selected in Global parameters.
Therefore, we have:
 BTS    P c  i ic 
Q pilot  i ic  = ---------------------------------------------
DL
I 0  ic 
With,
DL DL DL DL DL term
I 0  ic  = P tot  i ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0 for the total noise option,
And
DL DL DL DL DL term
I 0  ic  = I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0 –  1 –     BTS  P c  i ic  for the without
pilot option.
1st step: P c  i ic  calculation for each potential serving cell (i, ic)
P c  i ic  is the pilot power of a transmitter i on carrier ic at the receiver.
P pilot  i ic 
P c  i ic  = -------------------------
-
LT
I
L T is the total loss between transmitter i and receiver.

I
286
L Tx  L path  L term  L body  L Indoor  M Shadowing – Ec  Io

L T = ------------------------------------------------------------------------------------------------------------------------------------
-
I G Tx  G term
DL DL DL
2nd step: P tot  j ic  , P tot  i ic  and P tot  j ic adj  calculations
We have:

DL DL
I extra  ic  = P tot  j ic 
txj j  i
P SCH  ic 
I intra  ic  = P tot  i ic  –  BTS     P tot  i ic  – -------------------
DL DL DL
-
 LT 
 Ptot  j icadj 
DL
DL txj j
I inter – carrier  ic  = ----------------------------------------
-
and
Tx
 -------------------------------------
DL
I inter – techno log y  ic  = -
Tx Tx m
ni
L total  ICP ic i ic
DL
For each transmitter of the network, P tot  ic  is the total power received at the receiver from the transmitter on the carrier ic.
DL P Tx  ic 
P tot  ic  = ----------------
-
LT
P Tx  ic  is the total power transmitted by the transmitter on the carrier ic.Total power transmitted by each cell is either a
simulation result (provided in Simulation properties (Cells tab)) or a value user-defined in Cell properties.
DL
For each transmitter of the network, P tot  ic adj  is the total power received at the receiver from the transmitter on the carrier
icadj.
DL P Tx  ic adj 
P tot  ic adj  = ----------------------
-
LT
P Tx  ic adj  is the total power transmitted by the transmitter on the carrier icadj. Total power transmitted by each cell is either
a simulation result (provided in Simulation properties (Cells tab)) or a value user-defined in Cell properties.
term
3rd step: N 0 calculation
term Tx DL
N0 = NF Term  K  T  W  NR inter – techno log y
DL
4th step: I 0  ic  and Q pilot  i ic  evaluation using formulas described above
DL
5th step: G macro – diversity calculation
DL
The macro-diversity gain, G macro – diversity , models the decrease in shadowing margin due to the fact there are several
available pilot signals at the mobile.
DL npaths
G macro – diversity = M Shadowing – Ec  Io – M Shadowing – Ec  Io
npaths
M Shadowing – Ec  Io is the shadowing margin when the mobile receives n pilot signals (not necessarily from transmitters
belonging to the mobile active set).
This parameter is determined from cell edge coverage probability and Ec/I0 standard
deviation. When the Ec/I0 standard deviation is set to 0, the macro-diversity gain equals
0.
6th step: Determination of the best serving cell
287
Among all potential serving cells, Atoll first selects the cells which belong to the highest priority layer and then, the one with
the highest RSCP. This cell is referred to as the best serving cell candidate ( c BC ).
Then, Atoll calculates the best server indicator ( I BS ) for the best serving cell candidate ( c BC ) and the other potential serving
cells ( c OC ):
I BS  c BS  = Q pilot  c BS  + M HO  c BS  + C IO  c BS  for the best serving cell candidate,
I BS  c OC  = Q pilot  c OC  + C IO  c OC  for the other potential serving cells.
Atoll takes the cell with the highest best server indicator ( c max  I ) and calculates the best pilot quality received with a fixed
BS 
Resulting Resulting DL
cell edge coverage probability, Q pilot . Q pilot = G macro – diversity  Q pilot  c max  I 
BS 
Resulting req Resulting

If Q pilot  Q pilot , it means pilot quality at the receiver exceeds Q pilot x% of time (x is the fixed cell edge coverage
probability). The cell enters the active set as best serving cell, BS. Its carrier (icBS) will be used by other transmitters of the
active set (when active set size is greater than 1). Pilot is available.
Resulting req
If Q pilot  Q pilot , no cell among the potential serving cells can enter the active set. Pilot is unavailable.
7th step: Determination of active-set

Then, pilot qualities received from all potential serving cells other than BS ( Q pilot  i ic BS  ) are recalculated to determine the
entire receiver active set (when active set size is greater than 1). Same formulas and calculation method are used to update
DL
I 0  ic BS  value and determine Q pilot  i ic BS  .
We have:
 BTS    P c  i ic 
Q pilot  i ic  = ---------------------------------------------
DL
I 0  ic 
With,
DL DL DL DL DL term
I 0  ic  = P tot  i ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0 for the total noise option,
And
DL DL DL DL DL term
I 0  ic  = I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0 –  1 –     BTS  P c  i ic  for the without
pilot option.
Other cells (i,icBS) in the active set must satisfy the following criteria:
Q pilot  i ic BS  – Q pilot  BS   AS_threshold  BS 
 i ic BS   neighbour list  BS  (optionally)
Number of Cells in Active Set
This is a user-specified input in the Terminal properties. It corresponds to the active set size.
Thermal Noise
This parameter is calculated as described above (3rd step).
I0 (Best Server)
I0 (Best server) is the total noise received at the receiver on icBS. The notation “Best server” refers to the best serving cell of
active set. This is relevant when using the calculation option “Without pilot”. In this case, it informs that the pilot signal of the
best serving cell (BS) is deducted from the total noise.
Downlink Macro-Diversity Gain
This parameter is calculated as described above (5th step).
288
4.4.2.2 Downlink R99 Sub-Menu

The Downlink R99 sub-menu contains R99-related results.
Atoll calculates the traffic channel quality from each cell (k,icBS) of the receiver’s active set at the receiver. No power control
is performed as in simulations. Here, Atoll determines the downlink traffic channel quality at the receiver for the maximum
allowed traffic channel power per transmitter. Then, after combination, the total downlink traffic channel quality is evaluated
and compared with the specified target quality.
Eb/Nt Target
DL
Eb/Nt target ( Q req ) is defined for a given R99 bearer, a mobility type and a reception equipment. This parameter is available
in the R99 Bearer Selection table.
Compressed mode is operated when amobile supporting compressed mode is connected

to a cell located on a site with a compressed-mode-capable equipment, and
• Either the received Ec/I0 is lower than the Ec/I0 activation threshold (Global
parameters): Q pilot  Q pilot ,
• Or the pilot RSCP is lower than the pilot RSCP activation threshold (Global
CM – activation
parameters): P c  RSCP pilot .
When compressed mode is activated, the downlink Eb/Nt target is increased by the value
DL
user-defined for the DL Eb/Nt target increase field (Global parameters), Q req .
Required transmitter power on traffic channels
req
The calculation of the required transmitter power on traffic channels ( P tch ) may be divided into three steps.
DL
1st step: Q max  k ic BS  evaluation for each cell
DL
Let us assume the following notation: Eb/Nt max corresponds to Q max
Therefore, for each cell (k,icBS), we have:

DL
DL  BTS  P b – max  k ic BS 
-  G DL
Q max  k ic BS  = ------------------------------------------------------
DL
p  G Div
DL
N tot  ic BS 
max
DL P tch
With P b – max  k ic BS  = ----------
-
LT
k
DL DL DL DL DL term
and N tot  ic BS  = I intra  ic BS  + I extra  ic BS  + I inter – carrier  ic BS  + I inter – techno log y  ic BS  + N 0
Where
max
P tch is the maximum power allowed on traffic channels. This parameter is user-defined in the R99 Radio Bearers table.
DL
N tot  ic BS  is the total noise at the receiver on the carrier of the best serving cell.
DL
I intra  ic BS  is the intra-cell interference at the receiver on the carrier of the best serving cell.
P SCH  k ic BS 
I intra  ic BS  = P DL  k ic  –  BTS  F ortho   P DL  k ic  – ------------------------------
DL
tot BS tot BS L  T
DL
I extra  ic BS  is the extra-cell interference at the receiver on the carrier of the best serving cell.
 Ptot  j icBS 
DL DL
I extra  ic BS  =
j j  k
DL
I inter – carrier  ic BS  is the inter-carrier interference at the receiver on the carrier of the best serving cell.
289
DL
DL txj j
I inter – carrier  ic BS  = ----------------------------------------
-
RF  ic BS ic adj 
icadj is a carrier adjacent to icBS.
RF  ic BS ic adj  is the interference reduction factor, defined between ic and icadj and set to a value different from 0.
DL
I inter – techno log y  ic BS  is the inter-technology interference at the receiver on the carrier of the best serving cell.
Tx

DL
I inter – techno log y  ic BS  = -----------------------------------------
Tx Tx m
ni
i BS
th
Tx m
i BS
frequency gap between ic i (external network) and ic BS .
2nd step: Calculation of the total traffic channel quality

DL
Q MAX is the traffic channel quality at the receiver on icBS after signal combination of all the transmitters k of the active set.
On downlink, if there is no handoff, we have:

DL DL
Q MAX  ic BS  = Q max  k ic BS 
For any other handoff status, we have:
 Qmax  k icBS 
DL DL DL
Q MAX  ic BS  = f rake efficiency 
k
Where
DL
f rake efficiency is the downlink rake efficiency factor defined in Terminal properties.
req
3rd step: P tch calculation
DL
req Q req
-  P max
P tch = -------------------------- tch
DL
Q MAX  ic BS 
Compressed mode is operated when a mobile supporting compressed mode is

connected to a cell located on a site with a compressed-mode-capable equipment, and
• Either the received Ec/I0 is lower than the Ec/I0 activation threshold (Global
parameters): Q pilot  Q pilot .
• Or the pilot RSCP is lower than the pilot RSCP activation threshold (Global
CM – activation
parameters): P c  RSCP pilot
When compressed mode is activated, the downlink Eb/Nt target is increased by the value
DL
user-defined for the DL Eb/Nt target increase field (Global parameters), Q req . In this
DL DL
req Q req  Q req
-  P max
case, we have: P tch = ------------------------------ tch
DL
Q MAX  ic BS 
Max Eb/Nt for Each Cell of Active Set
For each cell (k,icBS), we have:

DL
DL  BTS  P b – max  k ic BS 
-  G DL
Q max  k ic BS  = ------------------------------------------------------
DL
p  G Div
DL
N tot  ic BS 
290
max
DL P tch
With P b – max  k ic BS  = ----------
-
LT
k
DL DL DL DL DL term
N tot  ic BS  = I intra  ic BS  + I extra  ic BS  + I inter – carrier  ic BS  + I inter – techno log y  ic BS  + N 0
max req
P SCH  k ic BS  P tch – P tch
I intra  ic BS  = P DL  k ic  –  BTS  F ortho   P DL  k ic  – ------------------------------
DL
–  1 –  BTS   max (--------------------------
-,0)
tot BS  tot BS L  T L Tk
DL DL
j j  k
DL
DL  j
I inter – carrier  ic BS  = txj
----------------------------------------
-
Tx

DL
Tx Tx m
ni
i BS
Where
req
P tch is the required transmitter power on traffic channels.
Max Eb/Nt
DL
Q MAX is the traffic channel quality at the receiver on icBS after signal combination of all the transmitters k of the active set.

DL DL
Q MAX  ic BS  = Q max  k ic BS 
 Qmax  k icBS 
DL DL DL
k
Where
DL
DL DL DL DL DL
Therefore, the service on the downlink traffic channel is available if Q MAX  ic BS   Q req (or Q MAX  ic BS   Q req  Q req when
compressed mode is activated).
Effective Eb/Nt
DL
Q eff is the effective traffic channel quality at the receiver on icBS.
DL DL DL DL DL DL DL
Q eff = min  Q MAX Q req  (or Q eff = min  Q MAX Q req  Q req  when compressed mode is activated).
Downlink Soft Handover Gain
DL
G SHO corresponds to the DL soft handover gain.
DL
DL Q MAX  ic BS 
G SHO = ------------------------------------------------
DL
max  Qmax  k ic BS  
DL DL
max  Qmax  k ic BS   corresponds to the highest Q max  k ic BS  value.
4.4.2.3 Uplink R99 Sub-Menu

The Uplink R99 sub-menu contains R99-related results.
291
For each cell (k,icBS) in the receiver’s active set, Atoll calculates uplink traffic channel quality from receiver. No power control
is performed as in simulations. Here, Atoll determines the uplink traffic channel quality at the cell for the maximum terminal
power allowed. Then, the total uplink traffic channel quality is evaluated with respect to the receiver handover status. From
this value, Atoll calculates the terminal power required to obtain the R99 bearer and compares it to the maximum terminal
power allowed.
Max Terminal Power
max
Max terminal power ( P term ) is an input user-defined for each terminal. It corresponds to the terminal’s maximum power.
Required Terminal Power
req
The calculation of the terminal power required to obtain an R99 bearer ( P term – R99 ) may be divided into three steps.
UL
1st step: Q max  k ic BS  evaluation for each cell
For each cell (k,icBS) in the receiver’s active set, we have:

UL
UL  term  P b – max  k ic BS 
-  G UL
Q max  k ic BS  = --------------------------------------------------------
UL
p  G Div
UL
N tot  k ic BS 
max UL
UL P term   1 – r c 
With P b – max  k ic BS  = ---------------------------------------
-
LT
k
UL
N tot  k ic BS  is the total noise at the transmitter on the carrier of the best serving cell. This value is calculated from the cell
UL
uplink load factor X  k ic BS  .
tx
UL N0
N tot  k ic BS  = ------------------------------------
-
UL
1 – X  k ic BS 
tx
N 0 is the transmitter thermal noise.
2nd step: Calculation of the total traffic channel quality

UL
Q MAX  ic BS  is the traffic channel quality at the transmitter on icBS after signal combination of all the transmitters k of the
active set.
UL UL
If there is no handoff (1/1): Q MAX  ic BS  = Q max  k ic BS 
For soft handoff (2/2):

UL UL UL
Q MAX  ic BS  =  G macro – diversity  2 links  max  Q max  k ic BS  
UL
 G macro – diversity  2 links is the uplink macro-diversity gain. This parameter is determined from the fixed cell edge coverage
probability and the uplink Eb/Nt standard deviation. When the option “Shadowing taken into account” is not selected
(Prediction properties), Atoll considers the uplink macro-diversity gain defined by the user in Global parameters.
UL UL
max  Q max  k ic BS   corresponds to the highest Q max  k ic BS  value.
For soft-soft handoffs (3/3):

UL UL UL
UL
For softer and softer-softer handoffs (1/2 and 1/3):
  Qmax  k icBS  
UL UL UL
k
For softer-soft handoffs (2/3), there are two possibilities. If the MRC option is selected (option available in Global parameters),
we have:
292
 UL 

UL UL UL UL
Q MAX  ic BS  =  G macro – diversity  2 links  max  f rake efficiency   Q max  k ic BS   Q max  k ic BS 
 k on the same site 
k on the same site
Else,
UL UL UL
req
3rd step: P term – R99 calculation
req
P term – R99 is the required terminal power.
UL
req Q req
-  P max
P term – R99 = -------------------------- term
UL
Q MAX  ic BS 
UL
Q req is the uplink traffic quality target defined by the user for a given reception equipment, a given R99 bearer and a given
mobility type. This parameter is available in the R99 Bearer Selection table.
Compressed mode is operated when a mobile supporting compressed mode is

connected to a cell located on a site with a compressed-mode-capable equipment, and
• The received Ec/I0 is lower than the Ec/I0 activation threshold (Global parameters):
Q pilot  Q pilot .
• The pilot RSCP is lower than the pilot RSCP activation threshold (Global parameters):
CM – activation
P c  RSCP pilot
When compressed mode is activated, the uplink Eb/Nt target is increased by the value
UL
user-defined for the UL Eb/Nt target increase field (Global parameters), Q req . In this
UL UL
-  P max
case, we have: P term – R99 = ------------------------------ term
UL
Q MAX  ic BS 
req max
Therefore, the service on the uplink traffic channel is available if P term – R99  P term .
Eb/Nt Max
For each cell (k,icBS) in the receiver’s active set, we have:

UL
UL  term  P b – max  k ic BS 
-  G UL
Q max  k ic BS  = --------------------------------------------------------
UL
p  G Div
UL
N tot  k ic BS 
max UL
UL P term   1 – r c 
With P b – max  k ic BS  = ---------------------------------------
-
LT
k
UL
N tot  k ic BS  is the total noise at the transmitter on the carrier of the best serving cell. This value is calculated from the cell
UL
uplink load factor X  k ic BS  .
tx max req
UL N0 P term – P term – R99
- +  1 –  term   max (------------------------------------------
N tot  k ic BS  = ------------------------------------ ,0)
UL LT
1 – X  k ic BS  k
tx
UL
Q MAX  ic BS  is the traffic channel quality at the transmitter on icBS after signal combination of all the transmitters k of the
active set.
UL UL
If there is no handoff (1/1): Q MAX  ic BS  = Q max  k ic BS 
For soft handoff (2/2):

UL UL UL
293
UL
UL UL
For soft-soft handoffs (3/3):

UL UL UL
UL
For softer and softer-softer handoffs (1/2 and 1/3):
  Qmax  k icBS  
UL UL UL
k
For softer-soft handoffs (2/3), there are two possibilities. If the MRC option is selected (option available in Global parameters),
we have:
 UL 

UL UL UL UL
Q MAX  ic BS  =  G macro – diversity  2 links  max  f rake efficiency   Q max  k ic BS   Q max  k ic BS 
 k on the same site 
k on the same site
Else,
UL UL UL
Effective Eb/Nt
UL
Q eff is the effective traffic channel quality at the transmitter on icBS.
UL UL UL UL UL UL UL
Q eff = min  Q MAX Q req  (or Q eff = min  Q MAX Q req  Q req  when compressed mode is activated).
Uplink Soft Handover Gain
UL
G SHO corresponds to the uplink soft handover gain.
UL
UL Q MAX  ic BS 
G SHO = ------------------------------------------------
UL
max  Q max  k ic BS  
UL UL
4.4.2.4 HSDPA Sub-Menu

The HSDPA sub-menu contains HSDPA-related results for HSDPA and HSPA users when the HS-SCCH quality is sufficient and
if the user can obtain an HSDPA bearer.
Atoll determines the best HSDPA bearer that the user can obtain. The HSDPA bearer user is processed as if he is the only user
in the cell, i.e. he uses the entire HSDPA power available in the cell. For further information on the fast link adaptation
modelling, see "Fast Link Adaptation Modelling" on page 242.
When modelling MC-HSDPA users (including DC-HSDPA users) and DB-MC-HSDPA users, Atoll determines the serving cells and
the best HSDPA bearer obtained in each serving cell. In each cell, the user is processed as if he is the only user in the cell. Atoll
details the results for each cell to which the user is connected. For further information on MC-HSDPA user modelling, see "MC-
HSDPA Users" on page 296. For further information on DB-MC-HSDPA user modelling, see "DB-MC-HSDPA Users" on
page 296.
General Results
Atoll displays the name of the cell to which the user is connected, the frequency band used by the transmitter, the selected
294
carrier, and the maximum available HSDPA power of the cell.
HS-PDSCH Ec/Nt
Atoll calculates the best HS-PDSCH quality (HS-PDSCH Ec/Nt). The way of calculating it depends on the selected option in the
transmitters global parameters (HSDPA part): CQI based on CPICH quality or CQI based on HS-PDSCH quality.
For further details on the HS-PDSCH quality calculation, see either "HS-PDSCH Quality Calculation" on page 244 if the selected
option is "CQI based on CPICH quality" or "HS-PDSCH Quality Calculation" on page 249 if the selected option is "CQI based on
HS-PDSCH quality".
HS-PDSCH Power
Atoll calculates the available HS-PDSCH power. For further details on the HS-PDSCH power calculation, see either "HS-PDSCH
Quality Calculation" on page 244 if the selected option is "CQI based on CPICH quality" or "HS-PDSCH Quality Calculation" on
page 249 if the selected option is "CQI based on HS-PDSCH quality".
HS-SCCH Ec/Nt Threshold
Atoll displays the HS-SCCH Ec/Nt threshold set for the selected mobility type.
HS-SCCH Ec/Nt
Atoll displays the obtained HS-SCCH quality.

When the HS-SCCH power allocation strategy is dynamic, this parameter corresponds to the HS-SCCH Ec/Nt threshold defined
for the selected mobility type.
When the HS-SCCH power allocation strategy is static, the HS-SCCH Ec/Nt is calculated from the fixed HS-SCCH power.
We have:
 BTS  P c  ic 
 -----
Ec-
 ic  i
= -------------------------------
N tot  ic 
And
 BTS  P c  ic 
 -----
Ec- 
 Nt  ic  HS – SCCH
i
= ----------------------------------------------------------------------------------------------------------------------------
DL term
i
With
DL DL DL DL DL term
DL  DL P SCH  ic   DL P SCH  ic 

- –  BTS   P tot  ic  – -------------------
DL term
-
 LT   LT 

DL DL
txj j  i
DL
DL  j
------------------------------------
DL
Tx

DL
Tx Tx m
n
i i
th
295
Tx m
i
P c  ic  = -------------------------------
-
i LT
i
And
L T = ------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
term term
CQI
It corresponds to the HS-PDSCH CQI.

The way of calculating it depends on the selected option in the transmitters global parameters (HSDPA part): CQI based on
CPICH quality or CQI based on HS-PDSCH quality.
For further details on the HS-PDSCH quality calculation, see either "HS-PDSCH CQI Determination" on page 246 if the selected
option is "CQI based on CPICH quality" or "HS-PDSCH CQI Determination" on page 252 if the selected option is "CQI based on
HS-PDSCH quality".
HSDPA Bearer Parameters
Knowing the HS-PDSCH CQI, Atoll calculates the best HSDPA bearer that can be used and selects a bearer compatible with cell
and terminal user equipment HSDPA capabilities. For further details on the HSDPA bearer selection, see "HSDPA Bearer
Selection" on page 246.
Atoll displays the parameters of the selected HSDPA bearer:
• The transport block size,
• The modulation scheme used,
• The number of HS-PDSCH channels used.
Peak RLC Throughput
DL
Once the bearer selected, Atoll determines the peak RLC throughput that can be provided to the user TP P –R LC .
Effective RLC Throughput
DL
Atoll displays the Effective RLC throughput ( TP E – RLC ) provided to the user. The Effective RLC throughput is calculated as
follows:
DL
DL TP P –RLC
TP E – RLC = -----------------
-
TTI
Where TTI is the minimum number of TTI (Transmission Time Interval) between two TTI used; it is defined in the terminal
user equipment category properties.
BLER
Atoll reads the BLER in the quality graph BLER = f(HS-PDSCH Ec/Nt) that is defined for the selected bearer and mobility type.
Knowing the HS-PDSCH Ec/Nt, it finds the corresponding BLER.
Bearer Consumption
Atoll provides this result for HSPA CBR service users only. The minimum throughput demand required by the service is
allocated to these users. Therefore, they partly consume the HSDPA bearer. The bearer consumption expressed in %,
C HSDPABearer , is calculated as follows:
TPD Min – DL
C HSDPABearer = ---------------------------------------------------
-
DL
TP P –R LC  I HSDPABearer 
296
MC-HSDPA Users
When multi-cell HSDPA is active, MC-HSDPA users can simultaneously connect to several HSDPA cells of the transmitter for
data transfer. The maximum number of cells to which the user can simultaneously connect depends on the DL multi-cell mode
set for the HSDPA UE category of the terminal.
Atoll determines the best serving cell using the best serving cell selection algorithm. For information on how the best serving
cell is selected, see "Best Serving Cell and Active Set Determination" on page 284. If the best carrier belongs to a transmitter
that supports the multi-cell HSDPA mode and if the transmitter has several HSDPA carriers, Atoll selects the other serving
cells, i.e., the secondary cells. The secondary cells belong to the same transmitter and are chosen among the adjacent carriers
according to the CQI. When two adjacent carriers are available, Atoll takes the one with the highest CQI value. Atoll selects
secondary cells as long as HSDPA carriers are available in the transmitter and the maximum number of cells to which the user
can simultaneously connect is not exceeded. In each serving cell (i.e., the best cell and the secondary cells), Atoll determines
the best HSDPA bearer obtained. In each cell, the user is processed as if he is the only user in the cell. The user is connected
to a cell if he obtains an HSDPA bearer.
DB-MC-HSDPA Users
When multi-cell HSDPA and dual-band HSDPA modes are active, DB-MC-HSDPA users can simultaneously connect to HSDPA
cells of two co-site transmitters using different frequency bands. If the two co-site transmitters work on the same frequency
band, then the users can only connect to the HSDPA cells of one transmitter. The maximum number of cells to which the user
can simultaneously connect depends on the DL multi-cell mode set for the HSDPA UE category of the terminal.
Let’s consider the following configuration:
• A site with transmitters working on two different frequency bands,
• The site equipment supports the dual-band HSDPA mode,
• Each transmitter has several HSDPA carriers,
• The multi-cell HSDPA mode is active for each transmitter.
Atoll determines the best serving cell using the best serving cell selection algorithm. For information on how the best serving
cell is selected, see "Best Serving Cell and Active Set Determination" on page 284. The secondary cells are taken in the same
band as the best carrier (i.e., they belong to the same transmitter), as long as carriers are available. Then, if additional carriers
are required and if there are no more carriers available in this transmitter, Atoll selects the carriers in a transmitter using the
second frequency band. Within one frequency band, the secondary cells are first selected according to an adjacency criterion
and then, according to the CQI value. When two adjacent carriers are available, Atoll takes the one with the highest CQI value.
In each serving cell (i.e., the best cell and the secondary cells), Atoll determines the best HSDPA bearer obtained. In each cell,
the user is processed as if he is the only user in the cell. The user is connected to a cell if he obtains an HSDPA bearer.
Results for MC-HSDPA and DB-MC-HSDPA Users
When the user is simultaneously connected to several HSDPA cells, Atoll details the results for each cell. In addition, it displays
the following results under Total:
• The Peak RLC Throughput

DL DL
TP P – RLC = TP P –RLC  c 
c  Serving cell
• The Effective RLC Throughput

DL
DL TP P –RLC
TP E – RLC = -----------------
-
TTI
• The Application Throughput

DL
 TP P –RLC  c    1 – BLER HSDPA  
DL  Serving cells -  f TP – Scaling – TP Offset
TP A = c----------------------------------------------------------------------------------------------------------
TTI
Where:
quality (HS-PDSCH Ec/Nt). Knowing the HS-PDSCH Ec/Nt, Atoll finds the corresponding BLER.
297
4.4.2.5 HSUPA Sub-Menu

The HSUPA sub-menu contains HSUPA-related results for HSPA users if the user can obtain an HSUPA bearer.
Atoll determines the best HSUPA bearer that the user can obtain. The HSUPA bearer user is processed as if he is the only user
in the cell, i.e. he uses the entire remaining load of the cell.
For further information on the HSUPA bearer selection, see "HSUPA Bearer Allocation Process" on page 261.
Required E-DPDCH Ec/Nt
req
It corresponds to the E-DPDCH Ec/Nt required to obtain the HSUPA bearer (  ------
Ec
). This value is defined for an HSUPA
Nt E – DPDCH
bearer ( Index HSUPABearer ) and a number of retransmissions ( N Rtx ) in the HSUPA Bearer Selection table.
Required Terminal Power
Ec req
From  ------
req
, Atoll calculates the terminal power required to obtain the HSUPA bearer, P term – HSUPA .
req
P term – HSUPA =  ------
req Ec UL
 L T  N tot
Nt E – DPDCH
With
UL intra UL tx
UL tx extra UL

UL
L T = ----------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
UL UL
tx intra extra UL tx
Max E-DPDCH Ec/Nt
Ec max
Atoll calculates the maximum E-DPDCH Ec⁄Nt allowed (  ------ ). For further details on the calculation, see "Max E-
DPDCH Ec/Nt" on page 298.
HSUPA Bearer Parameters
Atoll selects the best HSUPA bearer from the HSUPA compatible bearers. This is the HSUPA bearer with the highest potential
UL
TP P – RLC  Index HSUPABearer 
throughput ( ------------------------------------------------------------------
- ) where:
Ec  req
 ----- Ec max
• -   ------
 Nt E – DPDCH  Nt E – DPDCH
req max
• And P term – HSUPA  P term
With
max
P term : the maximum terminal power allowed.
Atoll displays the parameters of the selected HSUPA bearer:

• The radio bearer index,
• The TTI duration,
• The modulation scheme used,
• The number of E-DPDCH codes used.
298
Peak RLC Throughput
UL
After selecting the HSUPA bearer, Atoll determines the corresponding RLC peak throughput, TP P – RLC .
Peak RLC Throughput/No. of RTX
UL
TPP – RLC  Index HSUPABearer 
Atoll displays the peak RLC throughput to number of retransmissions ratio ( ------------------------------------------------------------------- ). Atoll considers the
ratio to select the HSUPA bearer when several HSUPA bearers meet the selection criteria.
Min Effective RLC Throughput
UL
From the RLC peak throughput, Atoll calculates the minimum effective RLC throughput, TP Min – E – RLC .
UL
UL TP P – RLC   1 – BLER HSUPA 
TP Min – E – RLC  M b  = ------------------------------------------------------------------
-
N Rtx
Where:
BLER HSUPA is the residual BLER after N Rtx retransmissions.
Application Throughput
UL
Atoll displays the provided application throughput ( TP A ). The application throughput represents the net throughput after
deduction of coding (redundancy, overhead, addressing, etc.). This one is calculated as follows:
UL
UL TP P – RLC   1 – BLER HSUPA   f TP – Scaling – TP Offset
TP A  M b  = ---------------------------------------------------------------------------------------------------------------------------
-
N Rtx
BLER
Knowing the E-DPDCH Ec/Nt, Atoll finds the corresponding BLER. It is read in the quality graph defined for the quartet
“reception equipment-selected bearer-number of retransmissions-mobility” (HSUPA Quality Graphs tab in the Reception
equipment properties). This graph describes the variation of BLER as a function of the measured quality (E-DPDCH Ec/Nt).
Bearer Consumption
Atoll provides this result for CBR service users only. The minimum bit rate required by the service is allocated to these users.
Therefore, they parly consume the HSUPA bearer. The bearer consumption expressed in %, C HSUPABearer , is calculated as
follows:
TPD min – UL
C HSUPABearer = -----------------------------------------------------
-
UL
TP P – RLC  I HSUPABearer 

Atoll calculates UMTS-specific coverage studies on each pixel where the pilot signal level exceeds the minimum RSCP
threshold. Let us assume each pixel on the map corresponds to a probe receiver with a terminal, a mobility type and a service.
This receiver does not create any interference. You can make the coverage prediction for a specific carrier, for all carriers, for
a specific layer, or for all layers. Coverage predictions are based on parameters that can be either simulation results, or user-
defined cell inputs.
4.4.3.1 Pilot Quality Analysis

Atoll determines the best serving cell (BS) for each pixel and calculates the pilot quality received with a fixed cell edge
Resulting
coverage probability, Q pilot  BS  .
Potential serving cells are filtered depending on the prediction definition (selected layers or carriers, layers supported by the
service and the terminal, mobility type) and the pilot signal level which must exceed the defined minimum RSCP threshold.
For further information on formulas, see "Definitions" on page 214. For information on the best serving cell selection and
pilot quality calculation, see "Bar Graph and Pilot Sub-Menu" on page 285.
299
4.4.3.1.1 Prediction Study Inputs

The Pilot Quality Analysis depends on the downlink total transmitted power of cells. This parameter can be either a simulation
output, or a user-defined cell input. In the last case, when no value is defined in the Cells table for the total transmitted power,
Atoll considers 50% of the maximum power as default value (i.e. 40 dBm if the maximum power is set to 43 dBm).
4.4.3.1.2 Study Display Options

Atoll displays the best pilot quality received with a fixed cell edge coverage probability.
Single colour
Resulting req req

Atoll displays a coverage if Q pilot  BS   Q pilot . Coverage consists of a single layer with a unique colour. Q pilot is a target
value defined in the Mobility table by the user.
Colour per transmitter
Resulting req
Atoll displays a coverage if Q pilot  BS   Q pilot . Coverage consists of several independent layers that can be displayed and
hidden on the map. There is a layer per transmitter. Layer colour is the colour assigned to the transmitter of the best serving
cell (BS).
Colour per mobility
In this case, receiver is not completely defined and no mobility is assigned.

Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per user-
Resulting req
defined mobility defined in the Mobility Types sub-folder. For each layer, area is covered if Q pilot  BS   Q pilot .
Colour per probability
This display option is available only if analysis is based on all simulations in a group (i.e. if you select a group of simulations
and the “All” option in the Condition tab of prediction properties). Coverage consists of several independent layers that can
be displayed and hidden on the map. There is a layer per user-defined probability level defined in the Display tab (Prediction
Resulting req
properties). For each layer, area is covered if Q pilot  BS   Q pilot in the required number of simulations.
Colour per cell edge coverage probability
defined cell edge coverage probability, p, defined in the Display tab (Prediction properties). For each layer, area is covered if
Resulting req
Q pilot  BS p   Q pilot .
Colour per quality level (Ec/I0)
defined quality threshold defined in the Display tab (Prediction properties). For each layer, area is covered if
Resulting
Q pilot  BS    Q pilot  threshold .
Colour per quality margin (Ec/I0 margin)
defined quality margin defined in the Display tab (Prediction properties). For each layer, area is covered if
Resulting req
Q pilot  BS  – Q pilot   Q pilot  m arg in .
Colour per pilot signal level (Ec)
defined pilot signal level defined in the Display tab (Prediction properties). For each layer, area is covered if
Resulting
Q pilot  BS    Q pilot  threshold .
4.4.3.2 Downlink Service Area Analysis

As in point predictions, Atoll calculates traffic channel quality at the receiver for each cell (k,icBS) in the receiver’s active set.
No power control is performed as in simulations. Here, Atoll determines downlink traffic channel quality at the receiver for a
300
DL
maximum allowed traffic channel power for transmitters. Then, the total downlink traffic channel quality ( Q MAX  ic BS  ) is
evaluated after recombination.
Atoll displays the total traffic channel quality in the downlink.
For information on best serving cell selection and active set determination, see "Best Serving Cell and Active Set
Determination" on page 284. For further details on calculations, see "Downlink R99 Sub-Menu" on page 288.

The Downlink Service Area Analysis depends on the downlink total transmitted power of cells. This parameter can be either
a simulation output, or a user-defined cell input. In the last case, when no value is defined in the Cells table for the total
transmitted power, Atoll considers 50% of the maximum power as default value (i.e. 40 dBm if the maximum power is set to
43 dBm).

Single colour
DL DL DL DL DL
Atoll displays a coverage with a unique colour if Q MAX  ic BS   Q req (or Q MAX  ic BS   Q req  Q req if compressed mode is
activated).
DL
Q req is the downlink traffic quality target defined by the user for a given reception equipment, a given R99 bearer and a given
mobility type. This parameter is available in the R99 Bearer Selection table.
DL
Q req is the DL Eb/Nt target increase; this parameter is user-defined in the Global parameters.
DL DL DL DL DL
Atoll displays a coverage if Q MAX  ic BS   Q req (or Q MAX  ic BS   Q req  Q req if compressed mode is activated). Coverage
consists of several independent layers that can be displayed and hidden on the map. There is a layer per transmitter. Layer
colour is the colour assigned to best serving transmitter.
Colour per mobility
In this case, receiver is not completely defined and no mobility is assigned. Coverage consists of several independent layers
that can be displayed and hidden on the map. There is a layer per user-defined mobility defined in Mobility sub-folder. For
DL DL DL DL DL
each layer, area is covered if Q MAX  ic BS   Q req (or Q MAX  ic BS   Q req  Q req if compressed mode is activated).
Colour per service
In this case, receiver is not completely defined and no service is assigned. Coverage consists of several independent layers that
can be displayed and hidden on the map. There is a layer per user-defined service defined in Services sub-folder. For each
DL DL DL DL DL
layer, area is covered if Q MAX  ic BS   Q req (or Q MAX  ic BS   Q req  Q req if compressed mode is activated).
DL DL
properties). For each layer, area is covered if Q MAX  ic BS   Q req in the required number of simulations.
Colour per cell edge coverage probability
defined cell edge coverage probability, p, defined in the Display tab (Prediction properties). For each layer, area is covered if
DL DL DL DL DL
Q MAX  ic BS p   Q req (or Q MAX  ic BS   Q req  Q req if compressed mode is activated).
Colour per maximum quality level (max Eb/Nt)
DL
Q MAX  ic BS   Threshold .
301
Colour per effective quality level (Effective Eb/Nt)
DL DL DL DL DL DL DL DL
Q eff  ic BS   Threshold . Q eff  ic BS  = min  Q MAX  ic BS  Q req  (or Q eff  ic BS  = min  Q MAX  ic BS  Q req  Q req 
when compressed mode is activated).
Colour per quality margin (Eb/Nt margin)
DL DL DL DL DL
Q MAX  ic BS  – Q req  M arg in (or Q MAX  ic BS  – Q req  Q req  M arg in when compressed mode is activated).
Colour per required power
req
Atoll calculates the downlink required power, P tch  ic BS  , as follows:
DL
req Q req
-  P max
P tch  ic BS  = -------------------------- tch
DL
Q MAX  ic BS 
Where
DL
Q req is the Eb/Nt target on downlink. This parameter, available in the R99 Bearer Selection table, is user-defined for a given
R99 bearer, a given reception equipment and a mobility type.
max
P tch is a user-defined input for each bearer related to a service. It corresponds to the maximum allowable traffic channel
power for a transmitter.
DL DL
-  P max
When compressed mode is activated, we have: P tch  ic BS  = ------------------------------ tch .
DL
Q MAX  ic BS 
defined required power threshold defined in the Display tab (Prediction properties). For each layer, area is covered if
req
P tch  ic BS   Threshold .
Colour per required power margin
defined power margin defined in the Display tab (Prediction properties). For each layer, area is covered if
req max
P tch  ic BS  – P tch  M arg in .
4.4.3.3 Uplink Service Area Analysis

As in point prediction, Atoll calculates uplink traffic channel quality from the receiver for each cell (k,icBS) in receiver active
set. No power control simulation is performed. Atoll determines uplink traffic channel quality at the transmitter for the
UL
maximum terminal power allowed. Then, the total uplink traffic channel quality ( Q MAX  ic BS  ) is evaluated with respect to
receiver handover status.
Atoll displays the total traffic channel quality in the uplink.
For information on best serving cell selection and active set determination, see "Best Serving Cell and Active Set
Determination" on page 284. For further details on calculations, see "Uplink R99 Sub-Menu" on page 291.

The Uplink Service Area Analysis depends on the UL load factor of cells. This parameter can be either a simulation output, or
a user-defined cell input. In the last case, when no value is defined in the Cells table for the uplink load factor, Atoll uses 50%
as default value.
302

Single colour
UL UL UL UL UL
colour is unique.
UL
Q req is defined for a reception equipment, a R99 bearer and a mobility type. This parameter is available in the R99 Bearer
Selection table.
UL
Q req is the UL Eb/Nt target increase; this parameter is user-defined in the Global parameters.
UL UL UL UL UL
consists of several independent layers that can be displayed and hidden on the map. There is a layer per transmitter. Layer
colour is the colour assigned to best server transmitter.
Colour per mobility
In this case, receiver is not completely defined and no mobility is assigned. Coverage consists of several independent layers
that can be displayed and hidden on the map. There is a layer per user-defined mobility defined in Mobility sub-folder. For
UL UL UL UL UL
each layer, area is covered if Q MAX  ic BS   Q req (or Q MAX  ic BS   Q req  Q req if compressed mode is activated).
Colour per service
In this case, receiver is not completely defined and no service is assigned. Coverage consists of several independent layers that
can be displayed and hidden on the map. There is a layer per user-defined service defined in Services sub-folder. For each
UL UL UL UL UL
layer, area is covered if Q MAX  ic BS   Q req (or Q MAX  ic BS   Q req  Q req if compressed mode is activated).
UL UL UL UL UL
properties). For each layer, area is covered if Q MAX  ic BS   Q req (or Q MAX  ic BS   Q req  Q req if compressed mode is
activated) in the required number of simulations.
Colour per maximum quality level (Max Eb/Nt)
UL
Q MAX  ic BS   Threshold .
Colour per effective quality level (Effective Eb/Nt)
UL
Q effective  ic BS   Threshold .
UL UL UL UL UL UL UL
Q eff  ic BS  = min  Q MAX  ic BS  Q req  (or Q eff  ic BS  = min  Q MAX  ic BS  Q req  Q req  when compressed mode is
activated).
Colour per quality margin (Eb/Nt margin)
UL UL UL UL UL
Q MAX  ic BS  – Q req  M arg in (or Q MAX  ic BS  – Q req  Q req  M arg in if compressed mode is activated).
303
Colour per required power
defined power threshold defined in the Display tab (Prediction properties). For each layer, area is covered if
req
P term – R99  ic BS   Threshold .
Colour per required power margin
defined power margin defined in the Display tab (Prediction properties). For each layer, area is covered if
req max
P term – R99  ic BS  – P term  M arg in .
Colour per soft handover gain
Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per soft
UL
handover gain value defined in the Display tab (Prediction properties). For each layer, area is covered if G SHO  Threshold .
4.4.3.4 Downlink Total Noise Analysis

Atoll determines the downlink total noise level generated by cells. You can make the coverage prediction for a specific carrier,
for all carriers, for a specific layer, or for all layers. We assume that each pixel on the map corresponds to a probe receiver
with a terminal, a mobility type and a service.
When you select "Best (All/Specific band)" as the carrier or layers associated with several carriers, Atoll determines the DL
total noise level on each carrier supported by the user service. When only one carrier is analysed, Atoll determines DL total
noise or DL noise rise on this carrier.
DL
Tx
 
DL DL txj j term
N tot  ic  = P tot  ic  + ------------------------------------ + -------------------------------------- + N 0
RF  ic ic adj  Tx
Tx m
txj j ni i
DL
Downlink noise rise, NR DL  ic  , is calculated from the downlink total noise, N tot , as follows:
term
 N0 
NR DL  ic  = – 10 log  ------------
-
 N DL tot 
4.4.3.4.1 Study Inputs

The Downlink Total Noise Analysis depends on the downlink total transmitted power of cells. This parameter can be either a
simulation output, or a user-defined cell input. In the last case, when no value is defined in the Cells table for the total
transmitted power, Atoll considers 50% of the maximum power as default value (i.e. 40 dBm if the maximum power is set to
43 dBm).
4.4.3.4.2 Display Options

The following display options are available for the prediction:
Colour per minimum noise level
defined noise level defined in the Display tab (Prediction properties). For each layer, area is covered if
DL
minN tot  ic   Threshold .
ic
Colour per maximum noise level
DL
maxN tot  ic   Threshold .
ic
304
Colour per average noise level
DL
averageNtot  ic   Threshold .
ic
Colour per minimum noise rise
Atoll displays bins where minNR DL  ic   Threshold . Coverage consists of several independent layers that can be displayed
ic
and hidden on the map. There is a layer per user-defined noise rise threshold defined in the Display tab.
Colour per maximum noise rise
Atoll displays bins where maxNR DL  ic   Threshold . Coverage consists of several independent layers that can be displayed
ic
and hidden on the map. There is a layer per user-defined noise rise threshold defined in the Display tab.
Colour per average noise rise
Atoll displays bins where averageNR DL  ic   Threshold . Coverage consists of several independent layers that can be
ic
displayed and hidden on the map. There is a layer per user-defined noise rise threshold defined in the Display tab.
When only one carrier is analysed, Atoll determines DL total noise or DL noise rise on this
carrier. In this case, the displayed coverage is the same for any selected display option
(average, minimum, maximum).
4.4.3.5 HSDPA Prediction Study

When calculating the HSDPA coverage prediction, either you can take all the possible HSDPA radio bearers into consideration,
or you can study a certain HSDPA radio bearer. Then, available display options depend on what you have selected.
When considering all the HSDPA radio bearers, you can set display parameters:
• To analyse the uplink and downlink A-DPCH qualities on the map,
• To analyse the HS-SCCH quality/power,
• To model fast link adaptation for a single HSDPA bearer user or for a defined number of HSDPA bearer users.
When studying a certain HSDPA radio bearer, you can display areas where a certain RLC peak throughput is available with
different cell edge coverage probabilities (i.e. the probability of having a certain RLC peak throughput). This type of analysis is
not relevant when modelling MC-HSDPA and DB-MC-HSDPA users.
Here we assume that each pixel on the map corresponds to one or several users with HSDPA capable terminal, mobility and
HSDPA service.The user does not create any interference. You can make the coverage prediction for a specific carrier, for all
carriers, for a specific layer, or for all layers. For DC-HSDPA, MC-HSDPA and DB-MC-HSDPA users, selecting one specific carrier
or one layer associated with one unique carrier is not suitable. To display the global throughput, you have to select several
carriers ("Best HSPA (All/Specific band)" as the carrier) or layers associated with several carriers.
For information on the best serving cell and secondary cells selection, see "MC-HSDPA Users" on page 296 and "DB-MC-
HSDPA Users" on page 296.
Note that the HSDPA service area is limited by the pilot quality, the A-DPCH quality and the HS-SCCH quality.

Parameters used as input for the HSDPA prediction study are:
• The available HSDPA power of the cell,
• The downlink total transmitted power of the cell,
• The number of HSDPA bearer users within the cell if the study is calculated for several users.
These parameters can be either simulation outputs, or user-defined cell inputs. In the last case, when no value is defined in
the Cells table for the total transmitted power and the number of HSDPA bearer users, Atoll uses the following default values:
• Total transmitted power = 50% of the maximum power (i.e, 40 dBm if the maximum power is set to 43 dBm)
• Number of HSDPA bearer users = 1
On the other hand, no default value is used for the available HSDPA power; this parameter must be defined by the user.
305

When considering all the HSDPA radio bearers, several display options are available in the study properties dialogue. They can
be regrouped in four categories according to the objective of the study:
• To analyse the uplink and downlink A-DPCH qualities on the map,
• To analyse the HS-SCCH quality/power,
• To model fast link adaptation for a single HSDPA bearer user,
• To model fast link adaptation for a defined number of HSDPA bearer users.
When studying a certain HSDPA radio bearer, only one display option is available. It allows you to display where a certain RLC
peak throughput is available with different cell edge coverage probabilities.
Analysis of UL And DL A-DPCH Qualities
• Colour per Max A-DPCH Eb/Nt DL

DL
Atoll displays the A-DPCH quality at the receiver ( Q MAX  BS  ) for the best serving cell (BS). No power control is performed as
in simulations. Here, Atoll determines downlink traffic channel quality at the receiver for a maximum traffic channel power
allowed for the best serving cell.
For information on calculation methods, see "Downlink R99 Sub-Menu" on page 288.
DL
Q MAX  BS   Threshold .
• Colour per Max A-DPCH Eb/Nt UL

UL
Atoll displays the A-DPCH quality at the best serving cell ( Q MAX  BS  ). No power control is performed as in simulations. Here,
Atoll determines uplink traffic channel quality at the receiver for a maximum terminal power allowed.
For information on calculation methods, see "Uplink R99 Sub-Menu" on page 291.
UL
Q MAX  BS   Threshold .
Analysis of The HS-SCCH Quality/Power
• Colour per HS-SCCH Power

This display option is relevant in case of dynamic HS-SCCH power allocation only. In this case, Atoll displays on each pixel the
HS-SCCH power per HS-SCCH channel. Coverage consists of several independent layers that can be displayed and hidden on
the map. There is a layer per threshold. For each layer, area is covered if P HS – SCCH  BS   Threshold .
• Colour per HS-SCCH Ec/Nt

This display option is relevant in case of static HS-SCCH power allocation only. In this case, Atoll displays on each pixel the HS-
SCCH quality per HS-SCCH channel. Coverage consists of several independent layers that can be displayed and hidden on the
map. There is a layer per threshold. For each layer, area is covered if  ------  BS 
Ec
 HS – SCCH  Threshold
.
 Nt
Fast Link Adaptation Modelling For A Single User
When you calculate the study with the following display options, Atoll considers one user on each pixel and determines the
best HSDPA bearer that the user can obtain. For MC-HSDPA and DB-MC-HSDPA users, Atoll determines the best HSDPA
bearers that the user can obtain in each serving cell. On each pixel, the user is processed as if he is the only user in the cell i.e.
he uses the entire HSDPA power available in the cell.
For further information on the fast link adaptation modelling, see "Fast Link Adaptation Modelling" on page 242.
• Colour per HS-PDSCH Ec/Nt
Atoll displays on each pixel the HS-PDSCH quality. For MC-HSDPA and DB-MC-HSDPA users, it corresponds to the HS-PDSCH
Ec/Nt of the best serving cell. Coverage consists of several independent layers that can be displayed and hidden on the map.
There is a layer per threshold. For each layer, area is covered if  ------  BS 
Ec
 HS – PDSCH  Threshold
.
 Nt
• Colour per CQI
306
Atoll displays either the CPICH CQI (see the calculation detail in "CPICH CQI Determination" on page 244) when the selected
option in Global parameters (HSDPA part) is CQI based on CPICH quality, or the HS-PDSCH CQI (see the calculation detail in
the section 10.7.1.2.2) when considering the CQI based on HS-PDSCH quality option.
For MC-HSDPA and DB-MC-HSDPA users, it corresponds to the CQI of the best serving cell.
Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per CQI
threshold (  CQI  threshold ). For each layer, area is covered if CQI   CQI  threshold .
• Colour per Peak MAC Throughput

DL
Atoll displays the Peak MAC throughput ( TPP –M AC ) provided on each pixel. The Peak MAC throughput is calculated as follows:
S block  c 

DL
TP P – M AC = ---------------------
T TTI
c  Serving cells
Where,
S block  c  is the transport block size (in kbits) of the HSDPA bearer selected in the cell, c, for the user; it is defined for each
HSDPA bearer in the HSDPA Radio Bearers table.
–3
Coverage consists of several independent layers that can be displayed and hidden on the map. There is a layer per possible
DL
Peak MAC throughput ( TPP –M AC ). For each layer, area is covered if the Peak MAC throughput exceeds the user-defined
thresholds.
• Colour per Effective MAC Throughput
DL
Atoll displays the Effective MAC throughput ( TP E – M AC ) provided on each pixel. The Effective MAC throughput is calculated as
follows:
S block  c 

DL
TP E – M AC = --------------------------
-
T TTI   TTI
c  Serving cells
Where,
S block  c  is the transport block size (in kbits) of the selected HSDPA bearer in the cell, c; it is defined for each HSDPA bearer
in the HSDPA Radio Bearers table.
–3
DL
Effective MAC throughput ( TP E – M AC ). For each layer, area is covered if the Effective MAC throughput exceeds the user-
defined thresholds.
• Colour per Peak RLC Throughput
After selecting the bearer, Atoll reads the corresponding RLC peak throughput ( TP DL  I ). This is the highest
P – RLC HSDPABearer 
throughput that the bearer can provide on each pixel. Then, it determines the peak RLC throughput provided by the serving
DL
cell, c, in the downlink, TP P – RLC  c  .
DL DL
For an HSDPA user, we have: TP P – RLC = TP P –RLC  c 
For MC-HSDPA and DB-MC-HSDPA users, the peak RLC throughput provided to the user is calculated as follows:

DL DL
TP P – RLC = TP P –RLC  c 
c  Serving cell
DL
RLC peak throughput ( TP P – RLC ). For each layer, area is covered if the peak RLC throughput can be provided.
• Colour per Effective RLC Throughput
307
DL
Atoll displays the Effective RLC throughput ( TP E – RLC ) provided on each pixel. The Effective RLC throughput is calculated as
follows:
DL
DL TP P –RLC
TP E – RLC = -----------------
-
TTI
DL
Effective RLC throughput ( TP E – RLC ). For each layer, area is covered if the Effective RLC throughput exceeds the user-defined
thresholds.
• Colour per Average Effective RLC Throughput
DL
Atoll displays the average effective RLC throughput ( TP Av – E – RLC ) provided on each pixel.
For an HSDPA user, we have:

DL
DL TP P –RLC  c    1 – BLER HSDPA 
TP Av –E –RLC = ------------------------------------------------------------------------
-
TTI
For MC-HSDPA and DB-MC-HSDPA users, we have:

DL
 TP P –RLC  c    1 – BLER HSDPA  
DL  Serving cells
TP Av –E –RLC = c----------------------------------------------------------------------------------------------------------
-
TTI
Where,
DL
average effective RLC throughput ( TP Av –E –RLC ). For each layer, area is covered if the average effective RLC throughput
exceeds the user-defined thresholds.
• Colour per Application Throughput
DL
Atoll displays the application throughput ( TP A ) provided on each pixel. The application throughput represents the net
throughput after deduction of coding (redundancy, overhead, addressing, etc.).
It is calculated as follows:
DL DL
TP A = TP Av –E – RLC  f TP – Scaling – TP Offset
Where:
DL
TP Av –E –RLC is the average effective RLC throughput.
DL
application throughput ( TPA ). For each layer, area is covered if the application throughput exceeds the user-defined
thresholds.
308
Fast Link Adaptation Modelling For Several Users
When you calculate the study with the following display options, Atoll considers several users per pixel and determines the
best HSDPA bearer that each user can obtain. In this case, the cell available HSDPA power is shared between HSDPA bearer
users. When the coverage prediction is not based on a simulation, the number of HSDPA bearer users is taken from the cell
properties. The displayed results of the coverage prediction will be an average result for one user.
For further information on the HSDPA bearer allocation process when there are several users, see "HSDPA Bearer Allocation
Process" on page 239 For further information on the fast link adaptation modelling, see "Fast Link Adaptation Modelling" on
page 242.
• Colour per Effective MAC Throughput per User
DL
Atoll displays the average Effective MAC throughput per user (  TP E –M AC  Av ) provided on each pixel. The average Effective
MAC throughput per user is calculated as follows:
n HSDPA

DL
TP E –M AC  x 
DL
 TP E –M AC  Av = ----------------------------------------------
x=1 -
Max  n HSDPA  c  
c  Serving cells  x 
Where,
n HSDPA  c  is the number of HSDPA bearer users within the cell, c.
DL
TP E – M AC  x  is the Effective MAC throughput of each HSDPA bearer user. For further information on the calculation of the
Effective MAC throughput, see "Colour per Effective MAC Throughput" on page 306.
DL
average Effective MAC throughput per user (  TP E –M AC  Av ). For each layer, area is covered if the average Effective MAC
throughput per user exceeds the user-defined thresholds.
• Colour per RLC Throughput per User
DL
Atoll displays the average effective RLC throughput per user (  TP E –R LC  Av ) provided on each pixel. The average effective RLC
throughput per user is calculated as follows:
n
HSDPA

DL
TP E –R LC  x 
DL
 TP E –R LC  Av = -----------------------------------------
x=1 -
n HSDPA
Where,
n HSDPA is the number of HSDPA bearer users within the cell.
DL
TP E – R LC  x  is the Effective RLC throughput of each HSDPA bearer user. For further information on the calculation of the
Effective RLC throughput, see "Colour per Effective RLC Throughput" on page 307.
DL
average effective RLC throughput per user (  TP E –R LC  Av ). For each layer, area is covered if the average effective RLC
throughput per user exceeds the user-defined thresholds.
• Colour per ApplicationThroughput per User
DL
Atoll displays the average application throughput per user (  TPA  Av ) provided on each pixel. The average application
throughput per user is calculated as follows:
n HSDPA

DL
TP A  x 
DL
 TP A  Av = ----------------------------------
x=1 -
n HSDPA
Where,
n HSDPA is the number of HSDPA bearer users within the cell.
309
DL
TP A  x  is the application throughput of each HSDPA bearer user. For further information on the calculation of the
application throughput, see "Colour per Application Throughput" on page 308.
DL
average application throughput per user (  TP A  Av ). For each layer, area is covered if the average application throughput per
user exceeds the user-defined thresholds.
Probability of Having a Certain Peak RLC Throughput
This result can be obtained only if you have selected an HSDPA radio bearer in the Condition tab.
• Colour per Cell Edge Coverage Probability
Atoll shows areas where the selected HSDPA radio bearer is available with different cell edge coverage probabilities. Coverage
consists of several independent layers that can be displayed and hidden on the map. There is a layer per cell edge coverage
probability defined in the Display tab. For each layer, area is covered if the selected HSDPA radio bearer is available.
4.4.3.6 HSUPA Prediction Study

A dedicated HSUPA study is available with different calculation and display options. Atoll determines on each pixel the best
HSUPA bearer that can be obtained; it can consider either a single HSPA user or several ones on each pixel. For further
information on the HSUPA bearer selection, see "HSUPA Bearer Allocation Process" on page 261. By calculating this study with
suitable display options, it is possible:
• To analyse the power required by the selected terminal,
• To analyse the required E-DPDCH quality,
• To analyse peak and effective throughputs.
We assume that each pixel on the map corresponds to one or several users with HSUPA capable terminal, mobility and HSUPA
service. You can make the coverage prediction for a specific carrier, for all carriers, for a specific layer, or for all layers. The
user does not create any interference.
Note that the HSUPA service area is limited by the pilot quality and the A-DPCH-EDPCCH quality.

Parameters used as input for the HSUPA prediction study are:
• The cell UL load factor,
• The cell UL reuse factor,
• The cell UL load factor due to HSUPA,
• The maximum cell UL load factor,
• The number of HSUPA bearer users within the cell if the study is calculated for several users.
These parameters can be either simulation outputs, or user-defined cell inputs. In the last case, When no value is defined in
the Cells table, Atoll uses the following default values:
• Uplink load factor = 50%
• Uplink reuse factor = 1
• Uplink load factor due to HSUPA = 0%
• Maximum uplink load factor = 75%
• Number of HSUPA beare users = 1
4.4.3.6.2 Calculation Options

Atoll can calculate the HSUPA coverage prediction in one of two ways:
• HSUPA resources can be dedictated to a single user: On each pixel, the user is processed as if he is the only user in
the cell i.e he will use the entire remaining load after allocating capacity to all R99 users.
• HSUPA resources can be shared by HSUPA users defined or calculated per cell: Atoll considers several HSUPA bearer
users per pixel. After allocating capacity to all R99 users, the remaining load of the cell will be shared equally between
all the HSUPA bearer users. When the coverage prediction is not based on a simulation, the number of HSUPA bearer
users is taken from the cell properties. The displayed results of the coverage prediction will be an average result for
one user.
4.4.3.6.3 Display Options

The following display options are available in the prediction property dialogue.
310
Colour per Required E-DPDCH Ec/Nt
Atoll displays on each pixel the E-DPDCH Ec/Nt required to obtain the selected HSUPA bearer. Coverage consists of several
independent layers that can be displayed and hidden on the map. There is a layer per threshold. For each layer, area is covered
Ec req
if  ------  Threshold .
Colour per Required Terminal Power
Atoll displays on each pixel the terminal power required to obtain the selected HSUPA bearer. The required terminal power
is calculated from the required E-DPDCH Ec/Nt. Coverage consists of several independent layers that can be displayed and
req
hidden on the map. There is a layer per threshold. For each layer, area is covered if P term  Threshold .
Colour per Peak MAC Throughput
UL
Atoll displays the Peak MAC throughput ( TPP –M AC ) provided on each pixel. The Peak MAC throughput is calculated as follows:
UL
UL S block
TP P – M AC = ------------
-
T TTI
Where,
UL
S block is the transport block size (in kbits) for the selected HSUPA bearer; it is defined for each HSUPA bearer in the HSUPA
Radio Bearers table.
T TTI is the duration of one TTI for the selected HSUPA bearer; it is defined for each HSUPA bearer in the HSUPA Radio Bearers
table.
UL
Peak MAC throughput ( TPP –M AC ). For each layer, area is covered if the Peak MAC throughput exceeds the user-defined
thresholds.
Colour per Peak RLC Throughput
After selecting the HSUPA bearer, Atoll reads the corresponding RLC peak throughput. This is the highest throughput that the
selected HSUPA bearer can provide on each pixel.
UL
RLC peak throughput ( TP P – RLC ). For each layer, area is covered if the peak RLC throughput can be provided.
Colour per Minimum Effective RLC Throughput
UL
Atoll displays the minimum effective RLC throughput ( TP Min – E – RLC ) provided on each pixel. The minimum effective RLC
throughput corresponds to the RLC throughput obtained for a given BLER and the maximum number of retransmissions. It is
UL
UL TP P – RLC   1 – BLER HSUPA 
TP Min – E – RLC = ----------------------------------------------------------------
-
N Rtx
Where,
BLER HSUPA is the residual BLER for the selected uplink transmission format (HSUPA bearer with N Rtx retransmissions). It is
read in the quality graph defined for the quartet “reception equipment-selected bearer-number of retransmissions-mobility”
(HSUPA Quality Graphs tab in the Reception equipment properties). This graph describes the variation of BLER as a function
of the measured quality (E-DPDCH Ec/Nt). Knowing the E-DPDCH Ec/Nt, Atoll finds the corresponding BLER.
N Rtx is the maximum number of retransmissions for the selected HSUPA bearer. This figure is read in the HSUPA Bearer
Selection table.
UL
minimum effective RLC throughput ( TP Min – E –RLC ). For each layer, area is covered if the minimum effective RLC throughput
311
Colour per Average Effective RLC Throughput
When HARQ (Hybrid Automatic Repeat Request) is used, the required average number of retransmissions is smaller and the
UL
Effective RLC throughput is an average effective RLC throughput ( TP Av –E –RL C ). This is the RLC throughput obtained for a given
BLER and the average number of retransmissions. It is calculated as follows:
UL
UL TP P –RLC   1 – BLER HSUPA 
TP Av –E –RL C = ----------------------------------------------------------------
-
 N Rtx  av
BLER HSUPA is the residual BLER for the selected uplink transmission format (HSUPA bearer with N Rtx retransmissions). It is
read in the quality graph defined for the quartet “reception equipment-selected bearer-number of retransmissions-mobility”
(HSUPA Quality Graphs tab in the Reception equipment properties). This graph describes the variation of BLER as a function
of the measured quality (E-DPDCH Ec/Nt). Knowing the E-DPDCH Ec/Nt, Atoll finds the corresponding BLER.
The average number of retransmissions (  N Rtx  av ) is determined from early termination probabilities defined for the selected
HSUPA bearer (in the HSUPA Bearer Selection table). The Early Termination Probability graph shows the probability of early
termination ( p ) as a function of the number of retransmissions ( N Rtx ). Atoll calculates the average number of
retransmissions (  N Rtx  av ) as follows:
N 
Rtx max
  p  N Rtx  – p  N Rtx – 1    N Rtx

N =1
 N Rtx  av = -----------------------------------------------------------------------------------------------
Rtx
p   N Rtx  max 
UL
average effective RLC throughput ( TP Av – E – RL C ). For each layer, area is covered if the minimum effective RLC throughput
Colour per Application Throughput
UL
Atoll displays the application throughput ( TP A ) provided on each pixel. The application throughput represents the net
throughput after deduction of coding (redundancy, overhead, addressing, etc.). This one is calculated as follows:
UL UL
TP A  M b  = TP Min – E –RLC  f TP – Scaling – TP Offset
Where:
f TP – Scaling and TP Offset respectively represent the scaling factor between the application throughput and the minimum RLC
(Radio Link Control) throughput and the throughput offset. These two parameters model the header information and other
UL
application throughput ( TP A ). For each layer, area is covered if the application throughput exceeds the user-defined
thresholds.
Colour per Average Application Throughput
UL
Atoll displays the average application throughput ( TP Av – A ) provided on each pixel. It is calculated as follows:
UL UL
TP Av – A  M b  = TP Av –E –RL C  f TP – Scaling – TP Offset
Where:
f TP – Scaling and TP Offset respectively represent the scaling factor between the average application throughput and the
average RLC (Radio Link Control) throughput and the throughput offset. These two parameters model the header information
and other supplementary data that does not appear at the application level. They are defined in the service properties.
UL
average application throughput ( TP Av – A ). For each layer, area is covered if the average application throughput exceeds the
user-defined thresholds.
312

Atoll permits the automatic allocation of intra-technology neighbours in the current network. Two allocation algorithms are
available, one dedicated to intra-carrier neighbours and the other for inter-carrier neighbours.
The intra-technology neighbour allocation algorithms take into account all the cells of TBC transmitters. It means that all the
cells of TBC transmitters of your .atl document are potential neighbours.
The cells to be allocated will be called TBA cells. They must fulfil following conditions:
Only TBA cells may be assigned neighbours.
zone.

• "Neighbour Allocation for All Transmitters" on page 312.
• "Neighbour Allocation for a Group of Transmitters or One Transmitter" on page 316.
• "Importance Calculation" on page 316.

We assume that we have a reference, cell A, and a candidate neighbour, cell B. When the automatic neighbour allocation
starts, Atoll checks the following conditions:
• The distance between both cells must be less than the user-definable maximum inter-site distance. If the distance
between the reference cell and the candidate neighbour is greater than this value, then the candidate neighbour is
discarded.
azimuths of antennas. For information on the effective distance calculation, see "Calculation of the Inter-Transmitter
• Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. You may choose
one or more carriers. Atoll will allocate neighbours only to the cells using the selected carriers.
For inter-carrier neighbours, in addition to the carrier(s) on which you want to run the allocation, you can select
the carrier(s) of potential neighbours.
• Force co-site cells as neighbours: This option enables you to force cells located on the reference cell site in the
candidate neighbour list. This constraint can be weighted among the others and ranks the neighbours through the
importance field (see after).
• Force adjacent cells as neighbours (only for intra-carrier neighbours): This option enables you to force cells
geographically adjacent to the reference cell in the candidate neighbour list. This constraint can be weighted
among the others and ranks the neighbours through the importance field (see below).
313
Let CellA be a candidate neighbour cell of CellB. CellA is considered adjacent to CellB if
there exists at least one pixel in the CellB Best Server coverage area where CellA is Best
Server (if several cells have the same best server value) or CellA is the second best server
that enters the Active Set (respecting the HO margin of the allocation).
When Force adjacent cells as neighbours is selected, adjacent cells are sorted and listed
from the most adjacent to the least adjacent, depending on the above criterion.
Adjacency is relative to the number of pixels satisfying the criterion.
• Force adjacent layers as neighbours: If selected, Atoll adds all the cells adjacent across network layers to the
reference cell to the candidate neighbour list. The weight of this constraint is always the average of the Min and
Max values defined for the adjacency factor. This weight is used to calculate the rank of each neighbour and its
importance. Cells are considered adjacent across layers if they belong to different layers and have a coverage
overlap of at least one pixel.
reference cell is a candidate neighbour of another cell, this one will be considered as candidate neighbour of the
reference cell.
If the neighbours list of a cell is full, the reference cell will not be added as a neighbour
of that cell and that cell will be removed from the reference cell’s neighbours list. You can
force Atoll to keep that cell in the reference cell’s neighbours list by adding the following
option in the Atoll.ini file:
[Neighbours]
you may force/forbid a cell to be candidate neighbour of the reference cell.
• Intra-carrier neighbours: intra-carrier handover is a soft handover.

The reference cell A and the candidate cell B are using the carrier c1 (c1 is the selected carrier on which you run the allocation).
SA is the area where the cell A is the best serving cell. It means that the cell A is the first one in the active set.
• The pilot signal level received from A is greater than the minimum pilot signal level (min RSCP).
• The best server indicator of A ( I BS  A  ) exceeds the minimum pilot quality (min Ec/I0).
• I BS  A  is the highest one.
For information on the best server indicator calculation, see "Best Serving Cell Determination in Monte Carlo
Simulations - Old Method" on page 283.
314
SB is the area where the cell B can enter the active set.
• The pilot signal received from the cell B is greater than the minimum pilot signal level (min RSCP).
• The pilot quality from B is greater than the pilot quality from A minus the Ec/I0 margin (AS Threshold).
Figure 4.16: Overlapping Zone for Intra-carrier Neighbours
• Inter-carrier neighbours: inter-frequency handover is a hard handover, triggered in multi-carrier W-CDMA networks
for coverage reasons (1st case) and to balance the load between carriers (2nd case).
The reference cell A is using the carrier c1 (c1 is the carrier selected in Source) and the candidate cell B is using the carrier c2
(c2 is the carrier selected in Destination).
SA is the area where the reference cell A is either the best serving cell among all cells using c1 (1st case) or a cell that can enter
the active set of a user connected to c1 (2nd case).
• 1st case: The cell A is the best serving cell among all cells using c1 but its pilot quality starts significantly decreasing.
• The best server indicator of A ( I BS  A  ) is the highest one.
• I BS  A  is lower than the minimum pilot quality (min Ec/I0) plus the handover margin.
• 2nd case: The cell A is not the best serving cell among all cells using c1 but it can enter the active set of a user
connected to c1.
• The best server indicator of A ( I BS  A  ) exceeds the minimum pilot quality (min Ec/I0).
• I BS  A  is not the highest one. It is strictly lower than the best server indicator of the best serving cell and
greater than the best server indicator of the best serving cell minus the handover margin.
SB is the area where the cell B is the best serving cell among all cells using c2.
• The pilot signal level received from B is greater than the minimum pilot signal level (min RSCP).
• The best server indicator of B ( I BS  B  ) exceeds the minimum pilot quality (min Ec/I0).
• I BS  B  is the highest one.
315
Figure 4.17: Overlapping Zone for Inter-carrier Neighbours - 1st case
Figure 4.18: Overlapping Zone for Inter-carrier Neighbours - 1st Case
For information on the best server indicator calculation, see "Best Serving Cell Determination in Monte Carlo Simulations - Old
Method" on page 283.
• Two ways enable you to determine the I0 value:

1. Global Value: A percentage of the cell maximum power is considered. If the % of
maximum power is too low, i.e. if %  Pmax  P pilot , Atoll takes into account the
pilot power of the cell. Then, I0 represents the sum of values calculated for each
cell.
2. Defined per Cell: Atoll takes into account the total downlink power defined per
cell. I0 represents the sum of total transmitted powers.
• For calculating the overlapping coverage areas, Atoll uses the service with the lowest
body loss, the terminal that has the highest difference between gain and losses or the
lowest noise figure when all terminals have the same (gain-losses) value, and the
shadowing margin calculated using the defined cell edge coverage probability, if the
option is selected. The service and terminal are selected such that the selection gives
the largest possible coverage areas for the cells.
SA  SB
Atoll calculates the percentage of covered area ( ------------------  100 ) and compares this value to the % minimum covered area. If
SA
this percentage is not exceeded, the candidate neighbour B is discarded.
• The importance of neighbours.
316
For information on the importance calculation, see "Importance Calculation" on page 316.
Importance values are used by the allocation algorithm to rank the neighbours according to the allocation reason. Atoll lists
all neighbours and sorts them by importance value so as to eliminate some of them from the neighbour list if the maximum
number of neighbours to be allocated to each transmitter is exceeded. If we consider the case for which there are 15
candidate neighbours and the maximum number of neighbours to be allocated to the reference cell is 8. Among these 15
candidate neighbours, only 8 (having the highest importance values) will be allocated to the reference cell. Note that specific
maximum numbers of neighbours (maximum number of intra-carrier neighbours, maximum number of inter-carrier
neighbours) can be defined at the cell level (property dialogue or cell table). If defined there, this value is taken into account
instead of the default one available in the Neighbour Allocation dialogue.
In the Results part, Atoll only displays the cells for which it finds new neighbours. For these cells, it provides the list of
neighbours, the number of neighbours and the maximum number of neighbours allowed for each cell. In addition, it indicates
the importance (in %) of each neighbour and the allocation reason. Therefore, a neighbour may be marked as exceptional pair,
co-site, adjacent, coverage or symmetric. For neighbours accepted for co-site, adjacency and coverage reasons, Atoll displays
the percentage of area meeting the coverage conditions and the corresponding surface area (km2), the percentage of area
meeting the adjacency conditions and the corresponding surface area (km2). Finally, if cells have previous allocations in the
list, neighbours are marked as existing.
site distance with the effective inter-cell distance. As a consequence, there can
be cases where the real distance between assigned neighbours is higher than the
Max inter-site distance, because the effective distance is smaller. You can force
Atoll to compare the Max inter-site distance with the real inter-site distance by
adding the following lines in Atoll.ini:
[Neighbours]
Atoll to prioritise the individual distances between reference cells and their
respective neighbour candidates by adding the following lines in Atoll.ini:
[Neighbours]

Transmitter
Atoll allocates neighbours to:
• TBA cells,
• Neighbours of TBA cells marked as exceptional pair, adjacent and symmetric,
• Neighbours of TBA cells that satisfy coverage conditions.
4.5.3 Importance Calculation

Importance values are used by the allocation algorithm to rank the neighbours according to the allocation reason and the
distance, and to quantify the neighbour importance.
4.5.3.1 Importance of Intra-carrier Neighbours


Co-site cell Importance Function (IF)
selected
317

Only if the Force adjacent layers as neighbours option is
selected
Only if the Force adjacent cells as neighbours option is
Adjacent cell Importance Function (IF)
selected
relationship
using the Importance Function (IF). The IF considers fourfactors for calculating the importance:
transmitter.
d-
 Di  = 1 – ----------
d max
antennas. For information on the effective distance calculation, see "Calculation of the Inter-Transmitter Distance" on
page 319.


Coverage 10%+20%{10%(Di)+90%(O)}+1%+9%(Di)
Adjacent cells
Co-site cells
Where:
318
overlapping.
causes.
4.5.3.2 Importance of Inter-carrier Neighbours

As indicated in the table below, the neighbour importance depends on the distance and on the neighbourhood cause; this
value varies between 0 to 100%.

Existing neighbour If the Delete existing neighbours option is not selected Existing importance
Exceptional pair If the Force exceptional pairs option is selected 100 %
Co-site cell If the Force co-site cells as neighbours option is selected Importance Function (IF)
If the % minimum covered area is exceeded Importance Function (IF)
If the Force neighbour symmetry option is selected Importance Function (IF)
relationship
using the Importance Function (IF). The IF considers threefactors for calculating the importance:
transmitter.
d-
 Di  = 1 – ----------
d max
page 319.

The IF is user-definable using the Min importance and Max importance fields.
The IF evaluates importance as follows:

Co-site Neighbourhood Resulting IF using the default values
IF
cause from the table above
No Min(O)+Delta(O){Max(Di)(Di)+(100%-Max(Di))(O)} 10%+50%{10%(Di)+90%(O)}
Min(C)+Delta(C){Max(Di)(Di)/(Max(Di)+Max(O))+
Yes 60%+40%{1/7%(Di)+6/7%(O)}
Max(O)(O)/(Max(Di)+Max(O))}
Where
319
co-site neighbours and neighbours allocated based on coverage overlapping.
causes.
4.5.4 Appendices
4.5.4.1 Calculation of the Inter-Transmitter Distance
d = D   1 + x  cos  – x  cos  
where x = 0.3% so that the maximum D variation does not exceed 1%.
4.6 Primary Scrambling Code Allocation

Downlink primary scrambling codes enable you to distinguish cells from one another (cell identification).
By default, there are 512 primary scrambling codes numbered (0...511).
The cells to which Atoll allocates scrambling codes are referred to as the TBA cells (cells to be allocated). TBA cells fulfil
following conditions:
• They belong to the folder on which allocation has been executed. This folder can be either the Transmitters folder
or a group of transmitters or a single transmitter.
zone.
320
4.6.1 Automatic Allocation Description

4.6.1.1 Options and Constraints
The scrambling code allocation algorithm can take into account following constraints and options:
• Neighbourhood between cells,
You may consider:
• First order neighbours: The neighbours of TBA cells listed in the Intra-technology neighbours table,
• Second order neighbours: The neighbours of neighbours,
• Third order neighbours: The neighbour’s neighbour’s neighbours.
• In the context of the primary scrambling code allocation, the term "neighbours" refers
to intra-carrier neighbours.
• Atoll can take into account inter-technology neighbour relations as constraints to
allocate different scrambling codes to the UMTS neighbours of a GSM transmitter. In
order to consider inter-technology neighbour relations in the scrambling code
allocation, you must make the Transmitters folder of the GSM .atl document
accessible in the UMTS .atl document. For information on making links between GSM
and UMTS .atl documents, see the User Manual.
• Atoll considers symmetry relationship between a cell, its first order neighbours, its
second order neighbours and its third order neighbours.
• Cells fulfilling a criterion on Ec/I0 (option “Additional Overlapping Conditions”),

Atoll reuses the intra-carrier neighbour allocation algorithm to determine the list of cells which cannot be allocated the same
scrambling code, and to calculate their importance.
For a reference cell “A”, Atoll considers all the cells “B” that can enter the active set on the area where the reference cell is
the best server (area where (Ec/I0)A exceeds the minimum Ec/I0 and is the highest one and (Ec/I0)B is within a Ec/I0 margin
of (Ec/I0)A).
• Atoll considers either a percentage of the cell maximum powers or the total downlink
power used by the cells in order to evaluate I0. In this case, I0 equals the sum of total
transmitted powers. When this parameter is not specified in the cell properties, Atoll
uses 50% of the maximum power.
• Reuse distance,
• Reuse distance is a constraint on the allocation of scrambling codes. A code cannot be

reused at a cell that is not at least as far away as the reuse distance from the cell
allocated with the particular code.
• Scrambling code reuse distance can be defined at cell level. If this value is not defined,
then Atoll will use the default reuse distance defined in the Scrambling Code
Automatic Allocation dialogue.
• Exceptional pairs,
• Domains of scrambling codes,
When no domain is assigned to cells, Atoll considers the 512 primary scrambling codes
available.
• The number of primary scrambling codes per cluster. In Atoll, we call "cluster", a group of scrambling codes as defined
in 3GPP specifications. 3GPP specifications define 64 clusters consisting of 8 scrambling codes (in this case, clusters
are numbererd from 0 to 63). However, you can define another value (e.g. if you set the number of codes per cluster
to 4, scrambling codes will be distributed in 128 clusters).
When the allocation is based on a Distributed strategy (Distributed per Cell or Distributed per Site), this parameter can
also be used to define the interval between the primary scrambling codes assigned to cells on a same site. The defined
interval is applied by adding the following lines in the Atoll.ini file:
[PSC]
321
ConstantStep = 1
For more information about setting options in the atoll.ini file, see the Administrator Manual.
• The carrier on which the allocation is run: It can be a given carrier or all of them. In this case, either Atoll independently
plans scrambling codes for the different carriers, or it allocates the same primary scrambling code to each carrier of a
transmitter if the option "Allocate carriers identically" is selected.
• The possibility to use a maximum of codes from the defined domains (option "Use a Maximum of Codes"): Atoll will
try to spread the scrambling code spectrum the most.
• The "Delete All Codes" option: When selecting this option, Atoll deletes all the current scrambling codes and carries
out a new scrambling code allocation. If not selected, the existing scrambling codes are kept.
In addition, it depends on the selected allocation strategy. Allocation strategies can be:
• Clustered allocation: The purpose of this strategy is to choose for a group of mutually constrained cells, scrambling
codes among a minimum number of clusters. In this case, Atoll will preferentially allocate all the codes within the
same cluster.
• Distributed per cell allocation: This strategy consists in using as many clusters as possible. Atoll will preferentially
allocate codes from different clusters.
• One cluster per site allocation: This strategy allocates one cluster to each site, then, one code from the cluster to each
cell of each site. When all the clusters have been allocated and there are still sites remaining to be allocated, Atoll
reuses the clusters as far as possible at another site.
• Distributed per site allocation: This strategy allocates a group of adjacent clusters to each site, then, one cluster to
each transmitter on the site according to its azimuth and finally, one code from the cluster to each cell of each
transmitter. The number of adjacent clusters per group depends on the number of transmitters per site you have in
your network; this information is required to start allocation based on this strategy. When all the groups of adjacent
clusters have been allocated and there are still sites remaining to be allocated, Atoll reuses the groups of adjacent
clusters as far as possible at another site.
In the Results table, Atoll only displays scrambling codes allocated to TBA cells.
4.6.1.2 Allocation Process

For each TBA cell, Atoll lists all cells which have constraints with the cell. They are referred to as near cells. The near cells of a
TBA cell may be:
• Its neighbour cells: the neighbours listed in the Intra-technology neighbours table (options “Existing neighbours” and
"First Order"),
• The neighbours of its neighbours (options “Existing neighbours” and “Second Order”),
• The third order neighbours (options “Existing neighbours” and “Third Order”),
• The cells that fulfil Ec/I0 condition (option “Additional Overlapping Conditions”),
• The cells with distance from the TBA cell less than the reuse distance,
• The cells that make exceptional pairs with the TBA cell.
Additional constraints are considered when:
• The cell and its near cells are neighbours of a same GSM transmitter (only if the Transmitters folder of the GSM .atl
document is accessible in the UMTS .atl document),
• The neighbour cells cannot share the same cluster (for the "Distributed per site" allocation strategy only).
These constraints have a certain weight taken into account to determine the TBA cell priority during the allocation process
and the cost of the scrambling code plan. During the allocation, Atoll tries to assign different scrambling codes to the TBA cell
and its near cells. If it respects all the constraints, the cost of the scrambling code plan is 0. When a cell has too many
constraints and there are not anymore scrambling codes available, Atoll breaks the constraint with the lowest cost so as to
generate the scrambling code plan with the lowest cost. For information on the cost generated by each constraint, see "Cell
Priority" on page 323.
4.6.1.2.1 Single Carrier Network

The allocation process depends on the selected strategy. Algorithm works as follows:
Strategies: Clustered and Distributed per Cell
Atoll processes TBA cells according to their priority. It allocates scrambling codes starting with the highest priority cell and its
near cells, and continuing with the lowest priority cells not allocated yet and their near cells. For information on calculating
cell priority, see "Cell Priority" on page 323.
322
Strategy: One Cluster per Site
All sites which have constraints with the studied site are referred to as near sites.
Atoll assigns a cluster to each site, starting with the highest priority site and its near sites, and continuing with the lowest
priority sites not allocated yet and their near sites. When all the clusters have been allocated and there are still sites remaining
to be allocated, Atoll reuses the clusters at another site. When the Reuse Distance option is selected, the algorithm reuses
the clusters as soon as the reuse distance is exceeded. Otherwise, when the option is not selected, the algorithm tries to assign
reused clusters as spaced out as possible.
Then, Atoll allocates a primary scrambling code from the cluster to each cell located on the sites (codes belong to the assigned
clusters). It starts with the highest priority cell and its near cells and goes on with the lowest priority cells not allocated yet
and their near cells.
For information on calculating site priority, see "Site Priority" on page 325. For information on calculating cell priority, see
"Cell Priority" on page 323.
Strategy: Distributed per Site
Atoll assigns a group of adjacent clusters to each site, starting with the highest priority site and its near sites, and continuing
with the lowest priority sites not allocated yet and their near sites. When all the groups of adjacent clusters have been
allocated and there are still sites remaining to be allocated, Atoll reuses the groups of adjacent clusters at another site. When
the Reuse Distance option is selected, the algorithm reuses the groups of adjacent clusters as soon as the reuse distance is
exceeded. Otherwise, when the option is not selected, the algorithm tries to assign reused groups of adjacent clusters as
spaced out as possible. Then, Atoll assigns each cluster of the group to each transmitter of the site according to the
transmitter azimuth and selected neighbourhood constraints (options "Neighbours in Other Clusters" and "Secondary
Neighbours in Other Clusters"). Then, Atoll allocates a primary scrambling code to each cell located on the transmitters (codes
belong to the assigned clusters). It starts with the highest priority cell and its near cells and goes on with the lowest priority
cells not allocated yet and their near cells.
Determination of Groups of Adjacent Clusters
In order to determine the groups of adjacent clusters to be used, Atoll proceeds as follows: It defines theoretical groups of
adjacent clusters, independently of the defined domain, considering the 512 primary scrambling codes available and the
specified number of codes per cluster (if this one is set to 8, 64 clusters are supposed to be available). It starts the division in
group from the cluster 0 (hard coded) and takes into account the maximum number of transmitters per site user-specified in
order to determine the number of clusters in each group and then, the number of possible groups.
Let us assume that the number of codes per cluster is set to 8 and the maximum number of transmitters per site in the
network is 3. In this case, we have the following theoretical groups:
Group 1 Group 2 Group 3 Group 4 ... Group 21
Cluster 0 Cluster 3 Cluster 6 Cluster 9 Cluster 61
Cluster 1 Cluster 4 Cluster 7 Cluster 10 ... Cluster 62
If no domain is assigned to cells, Atoll can use all these groups for the allocation. On the other hand, if a domain is used, the
tool compares adjacent clusters really available in the assigned domain to the theoretical groups and only keeps adjacent
clusters mapping the theoretical groups.
Let us assume that we have a domain consisted of 12 clusters: clusters 1 to 8 and clusters 12 to 15.
Therefore, Atoll will be able to use the following groups of adjacent clusters:
• Group 2 with cluster 3, 4 and 5,
• Group 3 with cluster 6, 7 and 8,
• Group 6 with cluster 12, 13 and 14.
• The clusters 1, 2 and 15 will not be used.
If a domain does not contain any adjacent clusters, the user is warned through the 'Event Viewer'.
4.6.1.2.2 Multi-Carrier Network

In case you have a multi-carrier network and you run the scrambling code allocation on all the carriers, the allocation process
depends on the allocation strategy as detailed above and in addition, wether the option "Allocate Carriers Identically" is
selected or not.
323
When the option is not selected, algorithm works for each strategy, as explained above. On the other hand, when the option
is selected, allocation order changes. It is no longer based on the cell priority but depends on the transmitter priority. All
transmitters which have constraints with the studied transmitter will be referred to as near transmitters.
In case of a "Per cell" strategy (Clustered and Distributed per cell), Atoll starts scrambling code allocation with the highest
priority transmitter and its near transmitters and continues with the lowest priority transmitters not allocated yet and their
near transmitters. The same scrambling code is assigned to each cell of the transmitter.
In case of the "One cluster per site" strategy, Atoll assigns a cluster to each site and then, allocates a scrambling code to each
transmitter. It starts with the highest priority transmitter and its near transmitters and continues with the lowest priority
transmitters not allocated yet and their near transmitters. The same scrambling code is assigned to each cell of the
transmitter.
In case of the "Distributed per site" strategy, Atoll assigns a group of adjacent clusters to each site, then a cluster to each
transmitter and finally, allocates a scrambling code to each transmitter. It starts with the highest priority transmitter and its
near transmitters and continues with the lowest priority transmitters not allocated yet and their near transmitters. The same
scrambling code is assigned to each cell of the transmitter.
For information on calculating transmitter priority, see "Transmitter Priority" on page 325.
When cells, transmitters or sites have the same priority, processing is based on an
alphanumeric order.
4.6.1.3 Priority Determination

4.6.1.3.1 Cell Priority
Scrambling code allocation algorithm in Atoll allots priorities to cells before performing the actual allocation. Priorities
assigned to cells depend upon how much constrained each cell is and the cost defined for each constraint. A cell without any
constraint has a default cost, C , equal to 0. The higher the cost on a cell, the higher the priority it has for the scrambling code
allocation process.
There are six criteria employed to determine the cell priority:
• Scrambling Code Domain Criterion
The cost due to the domain constraint, C i  Dom  , depends on the number of scrambling codes available for the allocation.
The domain constraint is mandatory and cannot be broken.
When no domain is assigned to cells, 512 scrambling codes are available and we have:
C i  Dom  = 0
When domains of scrambling codes are assigned to cells, each unavailable scrambling code generates a cost. The higher the
number of codes available in the domain, the less will be the cost due to this criterion. The cost is given as:
C i  Dom  = 512 – Number of scrambling codes in the domain
• Distance Criterion
The constraint level of any cell i depends on the number of cells (j) present within a radius of "reuse distance" from its centre.
The total cost due to the distance constraint is given as:
C i  Dist  =  Cj  Dist  i  
j
Each cell j within the reuse distance generates a cost given as:
C j  Dist  i   = w  d ij   c dis tan ce
Where
w  d ij  is a weight depending on the distance between i and j. This weight is inversely proportional to the inter-cell distance.
For a reuse distance of 2000m, the weight for an inter-cell distance of 1500m is 0.25, the weight for co-site cells is 1 and the
weight for two cells spaced out 2100m apart is 0.
c dis tan ce is the cost of the distance constraint. This value can be defined in the Constraint Cost dialogue.
• Exceptional Pair Criterion

The constraint level of any cell i depends on the number of exceptional pairs (j) for that cell. The total cost due to exceptional
pair constraint is given as:
324
C i  EP  =  cEP  i – j 
j
Where
c EP is the cost of the exceptional pair constraint. This value can be defined in the Constraint Cost dialogue.
• Neighbourhood Criterion
The constraint level of any cell i depends on the number of its neighbour cells j, the number of second order neighbours k and
the number of third order neighbours l.
Let’s consider the following neighbour schema:
Figure 4.20: Neighbourhood Constraints
The total cost due to the neighbour constraint is given as:

     
Ci  N  = 
  Cj  N1  i   +  Cj – j  N1  i   +   Ck  N2  i   +  Ck – k  N2  i   +   Cl  N3  i   +  Cl – l  N3  i  
j j k k l l
Each first order neighbour cell j generates a cost given as:

C j  N1  i   = I j  c N1
Where
I j is the importance of the neighbour cell j.
c N1 is the cost of the first order neighbour constraint. This value can be defined in the Constraint Cost dialogue.
Because two first order neighbours must not have the same scrambling code, Atoll considers the cost created by two first
order neighbours to be each other.
C j  N1  i   + C j  N1  i  
C j – j  N1  i   = -----------------------------------------------------
-
2
Each second order neighbour cell k generates a cost given as:
C k  N2  i   = Max ( C j  N1  i    C k  N1  j   , C j  N1  i    C k  N1  j   )  c N2
Where
c N2 is the cost of the second order neighbour constraint. This value can be defined in the Constraint Cost dialogue.
Because two second order neighbours must not have the same scrambling code, Atoll considers the cost created by two
second order neighbours to be each other.
C k  N2  i   + C k  N2  i  
C k – k  N2  i   = -------------------------------------------------------
-
2
Each third order neighbour cell l generates a cost given as:
 C  N1  i    C k  N1  j    C l  N1  k   C j  N1  i    C k  N1  j    C l N1  k  
C l  N3  i   = Max  j   c N3
  C j  N1  i    C k  N1  j     C l N1  k  C j  N1  i    C k  N1  j    C l N1  k  
Where
c N3 is the cost of the third order neighbour constraint. This value can be defined in the Constraint Cost dialogue.
Because two third order neighbours must not have the same scrambling code, Atoll considers the cost created by two third
C l  N3  i   + C l  N3  i  
C l – l  N3  i   = -----------------------------------------------------
-
2
325
Atoll considers the highest cost of both links when a neighbour relation is symmetric and
the importance value is different.
In this case, we have:

C j  N1  i   = Max  I i – j I j – i   c N1
And
C k  N2  i   = Max (C j  N1  i    C k  N1  j  ,C j  N1  k    C i  N1  j  )  c N2
• GSM Neighbour Criterion

This criterion is considered when the co-planning mode is activated (i.e. the Transmitters folder of the GSM .atl document is
made accessible in the UMTS .atl document) and inter-technology neighbours have been allocated. If the cell i is neighbour of
a GSM transmitter, the cell constraint level depends on how many cells j are neighbours of the same GSM transmitter. The
total cost due to GSM neighbour constraint is given as:
C i  N 2G  =  cN2G  j – Tx2G 
j
Where
cN is the cost of the GSM neighbour constraint. This value can be defined in the Constraint Cost dialogue.
2G
• Cluster Criterion
When the "Distributed per Site" allocation strategy is used, you can consider additional constraints on allocated clusters (one
cell, its first order neighbours and its second order neighbours must be assigned scrambling codes from different clusters). In
this case, the constraint level of any cell i depends on the number of first and second order neighbours, j and k. The total cost
due to the cluster constraint is given as:
C i  Cluster  =  Cj  N1  i    cCluster +  Ck  N2  i    cCluster

j k
Where
c Cluster is the cost of the cluster constraint. This value can be defined in the Constraint Cost dialogue.
Therefore, the total cost due to constraints on any cell i is defined as:
C i = C i  Dom  + C i  U 
With
C i  U  = C i  Dist  + C i  EP  + C i  N  + C i  N 2G  + C i  Cluster 
4.6.1.3.2 Transmitter Priority

In case you have a multi-carrier network and you run scrambling code allocation on "all" the carriers with the option "allocate
carriers identically", algorithm in Atoll allots priorities to transmitters. Priorities assigned to transmitters depend on how
much constrained each transmitter is and the cost defined for each constraint. The higher the cost on a transmitter, the higher
the priority it has for the scrambling code allocation process.
Let us consider a transmitter Tx with two cells using carriers 0 and 1. The cost due to constraints on the transmitter is given as:
C Tx = C Tx  Dom  + C Tx  U 
With C Tx  U  = Max  C  U   and C  Dom  = 512 – Number of scrambling codes in the domain
i  Tx i Tx
Here, the domain available for the transmitter is the intersection of domains assigned to cells of the transmitter. The domain
constraint is mandatory and cannot be broken.
4.6.1.3.3 Site Priority

In case of "Per Site" allocation strategies (One cluster per site and Distributed per site), algorithm in Atoll allots priorities to
sites. Priorities assigned to sites depend on how much constrained each site is and the cost defined for each constraint. The
higher the cost on a site, the higher the priority it has for the scrambling code allocation process.
326
Let us consider a site S with three transmitters; each of them has two cells using carriers 0 and 1. The cost due to constraints
on the site is given as:
C S = C S  U  + C S  Dom 
With C S  U  = Max  C  U   and C  Dom  = 512 – Number of scrambling codes in the domain
Tx  S Tx S
Here, the domain considered for the site is the intersection of domains available for transmitters of the site. The domain
4.6.2 Allocation Examples

4.6.2.1 Allocation Strategies and Use a Maximum of Codes
In order to understand the differences between the different allocation strategies and the behaviour of algorithm when using
a maximum of codes or not, let us consider the following sample scenario:
Figure 4.21: Primary Scrambling Codes Allocation
Let Site0, Site1, Site2 and Site3 be four sites with 3 cells using carrier 0 whom scrambling codes have to be allocated out of
three clusters consisted of 8 primary scrambling codes. This implies that the domain of scrambling codes for the four sites is
from 0 to 23 (cluster 0 to cluster 2). The reuse distance is supposed to be less than the inter-site distance. Only co-site
neighbours exist.
The following section lists the results of each combination of options with explanation where necessary.
4.6.2.1.1 Strategy: Clustered

Since the restrictions of neighbourhood only apply to co-sites with the same importance and sites distances are greater than
reuse distances, every cell has the same priority. Then, scrambling code allocation to cells is performed in an alphanumeric
order.
327
Without ‘Use a Maximum of Codes’ With ‘Use a Maximum of Codes’
Atoll starts allocating the codes from the start of cluster 0 at As it is possible to use a maximum of codes, Atoll starts
each site. allocation at the start of a different cluster at each site. When
a cluster is reused, and there are non allocated codes left in
the cluster, Atoll first allocates those codes before reusing
the already used ones.
4.6.2.1.2 Strategy: Distributed

reuse distances, every cell has the same priority. Then, scrambling code allocation to cells is performed in an alphanumeric
order.
Atoll allocates codes from different clusters to each cell of Atoll allocates codes from different clusters to each site’s
the same site. Under given constraints of neighbourhood and cells. As it is possible to use a maximum of codes, Atoll
reuse distance, same codes can be allocated to each site’s allocates the codes so that there is least repetition of codes.
cells.
328
4.6.2.1.3 Strategy: ‘One Cluster per Site

reuse distances, every site has the same priority. Then, cluster allocation to sites is performed in an alphanumeric order.
In this strategy, a cluster of codes is limited to be used at just When it is possible to use a maximum of codes, Atoll can
one site at a time unless all codes and clusters have been allocate different codes from a reused cluster at another site.
allocated and there are still sites remaining to be allocated. In
this case Atoll reuses the clusters as far as possible at another
site.
4.6.2.1.4 Strategy: ‘Distributed per Site

reuse distances, every site has the same priority. Then, the group of adjacent clusters allocation to sites is performed in an
alphanumeric order.
In this strategy, a group of adjacent clusters is limited to be When it is possible to use a maximum of codes, Atoll can
used at just one site at a time unless all codes and groups of allocate different codes from a reused group of adjacent
adjacent clusters have been allocated and there are still sites cluster at another site.
remaining to be allocated. In this case (here only one group
of adjacent clusters (clusters 0, 1 and 2) is available), Atoll
reuses the group at another site.
329
4.6.2.2 Allocate Carriers Identically

In order to understand the behaviour of algorithm when using the option "Allocate Carriers Identically" or not, let us consider
the following sample scenario:
Let Site0, Site1, Site2 and Site3 be four sites with 3 cells using carrier 0 and 3 cells using carrier 1. Scrambling codes have to
be allocated out of 3 clusters consisted of 8 primary scrambling codes. This implies that the domain of scrambling codes for
the five sites is from 0 to 23 (cluster 0 to cluster 2). The reuse distance is supposed to be less than the inter-site distance. Only
co-site neighbours exist. Allocation algorithm will be based on the "One Cluster per Site" strategy and the option "Use a
Maximum of Codes" is selected.
Without ‘Allocate Carriers Identically’ With ‘Allocate Carriers Identically’
Atoll allocates one cluster at each site as detailed in the In this case, Atoll allocates one cluster at each site and then,
previous section. Then, it allocates a code from the cluster to one code to each transmitter so as to use a maximum of
each cell of the site so as to use a maximum of codes. codes. Then, the same code is given to each cell of the
transmitter.
In both cases (with and without ’Allocate Carriers Identically’), every site has the same priority. Then, cluster allocation to sites
is performed in an alphanumeric order.
4.7 Automatic GSM-UMTS Neighbour Allocation

4.7.1 Overview
You can automatically calculate and allocate neighbours between GSM and UMTS networks. In Atoll, it is called inter-
technology neighbour allocation.
Inter-technology handover is used in two cases:
• When the UMTS coverage is not continuous. In this case, the UMTS coverage is extended by UMTS-GSM handover into
the GSM network,
• And in order to balance traffic and service distribution between both networks.
Note that the automatic inter-technology neighbour allocation algorithm takes into account both cases.
In order to be able to use the inter-technology neighbour allocation algorithm, you must have:
• An .atl document containing the GSM network, GSM.atl, and another one describing the UMTS network, UMTS.atl,
• An existing link on the Transmitters folder of GSM.atl into UMTS.atl.
The external neighbour allocation algorithm takes into account all the GSM TBC transmitters. It means that all the TBC
transmitters of GSM.atl are potential neighbours. The cells to be allocated will be called TBA cells which, being cells of
UMTS.atl, satisfy following conditions:
• They satisfy the filter criteria applied to Transmitters folder,
• They belong to the folder for which allocation has been executed. This folder can be either the Transmitters folder or
a group of transmitters subfolder.
Only UMTS TBA cells may be assigned neighbours.
330

The allocation algorithm takes into account criteria listed below:
• The inter-transmitter distance,
• The maximum number of neighbours fixed,
• Allocation options,
• The selected allocation strategy,
Two allocation strategies are available: the first one is based on distance and the second one on coverage overlapping.
We assume we have a UMTS reference cell, A, and a GSM candidate neighbour, transmitter B.
4.7.2.1 Algorithm Based on Distance

When the automatic allocation starts, Atoll checks the following conditions:
• The distance between the UMTS reference cell and the GSM neighbour must be less than the user-definable maximum
inter-site distance. If the distance between the UMTS reference cell and the GSM neighbour is greater than this value,
then the candidate neighbour is discarded.
• The calculation options,
Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. You may choose one or more
carriers. Atoll will allocate neighbours to cells using the selected carriers.
Force co-site cells as neighbours: It enables you to automatically include GSM transmitters located on the same site as the
reference UMTS cell in the candidate neighbour list. This option is automatically selected.
Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. Therefore, you may
force/forbid a GSM transmitter to be candidate neighbour of the reference UMTS cell.
Delete existing neighbours: When selecting the Delete existing neighbours option, Atoll deletes all the current neighbours and
carries out a new neighbour allocation. If not selected, existing neighbours are kept.
Importance values are used by the allocation algorithm to rank the neighbours. Atoll lists all neighbours and sorts them by
importance value so as to eliminate some of them from the neighbour list if the maximum number of neighbours to be
allocated to each cell is exceeded. If we consider the case for which there are 15 candidate neighbours and the maximum
number of neighbours to be allocated to the reference cell is 8. Among these 15 candidate neighbours, only 8 (having the
highest importance values) will be allocated to the reference cell. Note that the maximum number of inter-technology
neighbours can be defined at the cell level (property dialogue or cell table). If defined there, this value is taken into account
Co-site transmitter If the Force co-site cells as neighbours option is selected 100 %
Neighbourhood relationship that d-
If the maximum distance is not exceeded 1 – ----------
fulfils distance conditions d max
Where d is the effective distance between the UMTS reference cell and the GSM neighbour and d max is the maximum inter-
site distance.
In the Results part, Atoll provides the list of neighbours, the number of neighbours and the maximum number of neighbours
allowed for each cell. In addition, it indicates the importance (in %) of each neighbour and the allocation reason. Therefore, a
neighbour may be marked as exceptional pair, co-site, or distance. For neighbours accepted for distance reasons, Atoll
displays the distance from the reference cell (m). Finally, if cells have previous allocations in the list, neighbours are marked
as existing.
331
4.7.2.2 Algorithm Based on Coverage Overlapping

• The distance between the UMTS reference cell and the GSM neighbour must be less than the user-definable maximum
inter-site distance. If the distance between the UMTS reference cell and the GSM neighbour is greater than this value,
then the candidate neighbour is discarded.
Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. You may choose one or more
carriers. Atoll will allocate neighbours to cells using the selected carriers.
reference UMTS cell in the candidate neighbour list. This option is automatically selected.
force/forbid a GSM transmitter to be candidate neighbour of the reference UMTS cell.
• There must be an overlapping zone ( S A  S B ) with a given cell edge coverage probability.
Four different cases may be considered for SA:
• 1st case: SA is the area where the cell A is the best serving cell of the UMTS network.
• The pilot signal received from A is greater than the minimum pilot signal level,
• The pilot quality from A exceeds a user-definable minimum value (minimum Ec/I0) and is the highest one.
In this case, the Ec/I0 margin must be equal to 0dB and the max Ec/I0 option disabled.
• 2nd case: SA represents the area where the pilot quality from the cell A strats decreasing but the cell A is still the
best serving cell of the UMTS network.
The Ec/I0 margin must be equal to 0dB, the max Ec/I0 option selected and a maximum Ec/I0 user-defined.
• The pilot quality from A exceeds the minimum Ec/I0 but is lower than the maximum Ec/I0.
• The pilot quality from A is the highest one.
• 3rd case: SA represents the area where the cell A is not the best serving cell but can enter the active set.
Here, the Ec/I0 margin has to be different from 0dB and the max Ec/I0 option disabled.
• The pilot quality from A is within a margin from the best Ec/I0, where the best Ec/I0 exceeds the minimum Ec/
I0.
• 4th case: SA represents the area where:
• The pilot quality from A is within a margin from the best Ec/I0 (where the best Ec/I0 exceeds the minimum Ec/
I0) and lower than the maximum Ec/I0.
In this case, the margin must be different from 0dB, the max Ec/I0 option selected and a maximum Ec/I0 user-
defined.
Two different cases may be considered for SB:
• 1st case: SB is the area where the cell B is the best serving cell of the GSM network.
In this case, the margin must be set to 0dB.
• The signal level received from B on the BCCH TRX type exceeds the user-defined minimum threshold and is the
highest one.
• 2nd case: The margin is different from 0dB and SB is the area where:
• The signal level received from B on the BCCH TRX type exceeds the user-defined minimum threshold and is
within a margin from the best BCCH signal level.
SA  SB
-  100 ) and compares this value to the % minimum covered area. If
Atoll calculates the percentage of covered area ( -----------------
SA
332
Candidate neighbours fulfilling coverage conditions are sorted in descending order with respect to % of covered area.
When the automatic allocation is based on coverage overlapping, we recommend you to

perform two successive automatic allocations:
• A first allocation in order to find handovers due to non-continuous UMTS coverage.
In this case, you have to select the max Ec/I0 option and define a high enough value.
• A second allocation in order to complete the previous list with handovers motivated
for reasons of traffic and service distribution. Here, the max Ec/I0 option must be
disabled.

Importance values are used by the allocation algorithm to rank the neighbours according to the distance and the allocation
reason. Atoll lists all neighbours and sorts them by importance value so as to eliminate some of them from the neighbour list
if the maximum number of neighbours to be allocated to each cell is exceeded. If we consider the case for which there are 15
candidate neighbours, only 8 (having the highest importance values) will be allocated to the reference cell. Note that the
maximum number of inter-technology neighbours can be defined at the cell level (property dialogue or cell table). If defined
there, this value is taken into account instead of the default one available in the Neighbour Allocation dialogue.
Neighbourhood reason When Importance value
Co-site transmitter If the Force co-site cells as neighbours option is selected Importance Function (IF)
using the Importance Function (IF). The IF considers threefactors for calculating the importance:
transmitter.
d
 Di  = 1 – -----------
d max
page 319.


IF
Yes 60%+40%{1/7%(Di)+6/7%(O)}
Where
333
causes.
neighbour may be marked as exceptional pair, co-site or coverage. For neighbours accepted for co-site and coverage reasons,
Atoll displays the percentage of area meeting the coverage conditions and the corresponding surface area (km2). Finally, if
cells have previous allocations in the list, neighbours are marked as existing.
• No prediction study is needed to perform an automatic neighbour allocation. When

starting an automatic neighbour allocation, Atoll automatically calculates the path
loss matrices if not found.
• A forbidden neighbour must not be listed as neighbour except if the neighbourhood
relationship already exists and the Delete existing neighbours option is unchecked
when you start the new allocation. In this case, Atoll displays a warning in the Event
viewer indicating that the constraint on the forbidden neighbour will be ignored by
algorithm because the neighbour already exists.
• In the Results, Atoll displays only the cells for which it finds new neighbours.
Therefore, if a TBA cell has already reached its maximum number of neighbours
before starting the new allocation, it will not appear in the Results table.
4.7.2.3 Appendices
4.7.2.3.1 Delete Existing Neighbours Option
As explained above, Atoll keeps the existing inter-technology neighbours when the Delete existing neighbours option is not
checked. We assume that we have an existing allocation of inter-technology neighbours.
A new TBA cell i is created in UMTS.atl. Therefore, if you start a new allocation without selecting the Delete existing
neighbours option, Atoll determines the neighbour list of the cell i.
If you change some allocation criteria (e.g. increase the maximum number of neighbours or create a new GSM TBC
transmitter) and start a new allocation without selecting the Delete existing neighbours option, it examines the neighbour list
of TBA cells and checks allocation criteria if there is space in their neighbour lists. A new GSM TBC transmitter can enter the
TBA cell neighbour list if allocation criteria are satisfied. It will be the first one in the neighbour list.
334
Chapter 5
CDMA2000 Networks

• "General Prediction Studies" on page 337
• "Definitions and Formulas" on page 340
• "Active Set Management" on page 358
• "CDMA2000 Prediction Studies" on page 392
• "PN Offset Allocation" on page 430
• "Automatic GSM-CDMA Neighbour Allocation" on
page 438
© Forsk 2015
336
AT330_TRR_E1 Chapter 5: CDMA2000 Networks
5 CDMA2000 Networks
This chapter describes all the calculations performed in Atoll CDMA2000 documents.
All the calculations are performed on TBC (to be calculated) transmitters. For the definition of TBC transmitters please refer
to "Path Loss Matrices" on page 103.
5.1 General Prediction Studies

5.1.1 Calculation Criteria
Three criteria can be studied in point analysis (Profile tab) and in common coverage studies. Study criteria are detailed in the
table below:
Signal level ( P rec ) in dBm
P rec  ic  = EIRP  ic  – L path – M Shadowing – model – L Indoor + G term – L term
Path loss ( L path ) in dBm L path = L model + L ant

Tx
Total losses ( L total ) in dBm L total =  L path + L Tx + L term + L indoor + M Shadowing – model  –  G Tx + G term 
where,
EIRP is the effective isotropic radiated power of the transmitter,
ic is a carrier rank,
L model is the loss on the transmitter-receiver path (path loss) calculated by the propagation model,
L ant is the transmitter antenna attenuation (from antenna patterns),

Tx
M Shadowing – model is the shadowing margin. This parameter is taken into account when the option “Shadowing taken into
account” is selected,
L Indoor are the indoor losses, taken into account when the option “Indoor coverage” is selected,
L term are the receiver losses,
G term is the receiver antenna gain,
G Tx is the transmitter antenna gain,
L Tx is the transmitter loss ( L Tx = L total – DL ). For information on calculating transmitter loss, "UMTS, CDMA2000, TD-SCDMA,
WiMAX, and LTE Documents" on page 30.
• For CDMA2000 1xRTT systems, EIRP  ic  = P pilot  ic  + G Tx – L Tx (where, P pilot  ic 

is the cell pilot power).
• For CDMA2000 1xEV-DO systems, EIRP  ic  = P max  ic  + G Tx – L Tx (where P max  ic 
is the maximum cell power).
• When you make the prediction, you can consider the best carrier of all bands or the
best carrier of a particular frequency band (Best (All Bands/Specific Band) option). In
this case, Atoll displays the best signal level received from a transmitter. Therefore, if
the network consists of 1xRTT and 1xEV-DO carriers, Atoll takes the highest power of
both cells for each transmitter (i.e. the highest value between the pilot power of the
1xRTT cell and the maximum power of the 1xEV-DO cell) to calculate the received
signal level.
• Atoll considers that G term and L term equal zero.
337
Chapter 5: CDMA2000 Networks ©Forsk 2015

5.1.2.1 Profile Tab
Atoll displays either the signal level received from the selected transmitter on a carrier ( P rec  ic  ), or the highest signal level
received from the selected transmitter on the best carrier.

Analysis provided in the Reception tab is based on path loss matrices. So, you can study reception from TBC transmitters for
which path loss matrices have been computed on their calculation areas.
For each transmitter, Atoll displays either the signal level received on a carrier, ( P rec  ic  ), or the highest signal level received
on the best carrier.
Reception bars are displayed in a decreasing signal level order. The maximum number of reception bars depends on the signal
level received from the best server. Only reception bars of transmitters whose signal level is within a 30 dB margin from the
best server can be displayed.
• You can use a value other than 30 dB for the margin from the best server signal level,
for example a smaller value for improving the calculation speed. For more
information on defining a different value for this margin, see the Administrator
Manual.

For each TBC transmitter, Txi, Atoll determines the selected criterion on each bin inside the Txi calculation area. In fact, each
bin within the Txi calculation area is considered as a potential (fixed or mobile) receiver.
• The study conditions in order to determine the service area of each TBC transmitter,
• The display settings to select how to colour service areas.

Atoll uses parameters entered in the Condition tab of the coverage study property dialogue to predetermine areas where it
will display coverage.
We can distinguish three cases:

Txi Txi Txi
Minimum threshold  P rec  ic   or L total or L path   Maximum threshold

Txi Txi Txi
And
Txi Txj
P rec  ic   Best  P rec  ic   – M
ji

Best function: considers the highest value.
338
Txi is the highest.
from Txi is either the highest or 2dB lower than the highest.
from Txi is 2dB higher than the signal levels from transmitters, which are 2nd best
servers.

Txi Txi Txi
And
Txi nd Txj
P rec  ic   2 Best  P rec  ic   – M
ji
2nd Best function: considers the second highest value.
Txi is the second highest.
from Txi is either the second highest or 2dB lower than the second highest.
from Txi is 2dB higher than the signal levels from transmitters, which are 3rd best
servers.

5.1.3.2.1 Plot Resolution
Prediction plot resolution is independent of the matrix resolutions and can be defined on a per study basis. Prediction plots
are generated from multi-resolution path loss matrices using bilinear interpolation method (similar to the one used to
evaluate site altitude).

It is possible to display the transmitter service area with colours depending on any transmitter attribute or other criteria such
as:
Atoll calculates signal level received from the transmitter on each bin of each transmitter service area. A bin of a service area
is coloured if the signal level is greater than or equal to the defined minimum thresholds (bin colour depends on signal level).
as transmitter service areas. Each layer shows the different signal levels available in the transmitter service area.
Atoll calculates signal levels received from transmitters on each bin of each transmitter service area. Where other service
areas overlap the studied one, Atoll chooses the highest value. A bin of a service area is coloured if the signal level is greater
than or equal to the defined thresholds (the bin colour depends on the signal level). Coverage consists of several independent
layers whose visibility in the workspace can be managed. There are as many layers as defined thresholds. Each layer
corresponds to an area where the signal level from the best server exceeds a defined minimum threshold.
Path Loss (dB)
Atoll calculates path loss from the transmitter on each bin of each transmitter service area. A bin of a service area is coloured
if path loss is greater than or equal to the defined minimum thresholds (bin colour depends on path loss). Coverage consists
of several independent layers whose visibility in the workspace can be managed. There are as many layers as service areas.
Each layer shows the different path loss levels in the transmitter service area.
339
Total Losses (dB)
Atoll calculates total losses from the transmitter on each bin of each transmitter service area. A bin of a service area is
coloured if total losses is greater than or equal to the defined minimum thresholds (bin colour depends on total losses).
Atoll calculates signal levels received from transmitters on each bin of each transmitter service area. Where other service
areas overlap the studied one, Atoll determines the best transmitter and evaluates path loss from the best transmitter. A bin
of a service area is coloured if the path loss is greater than or equal to the defined thresholds (bin colour depends on path
layers as defined thresholds. Each layer corresponds to an area where the path loss from the best server exceeds a defined
minimum threshold.
Atoll calculates signal levels received from transmitters on each bin of each transmitter service area. Where service areas
overlap the studied one, Atoll determines the best transmitter and evaluates total losses from the best transmitter. A bin of
a service area is coloured if the total losses is greater than or equal to the defined thresholds (bin colour depends on total
layers as defined thresholds. Each layer corresponds to an area where the total losses from the best server exceed a defined
minimum threshold.
Number of Servers
Atoll evaluates how many service areas cover a bin in order to determine the number of servers. The bin colour depends on
the number of servers. Coverage consists of several independent layers whose visibility in the workspace can be managed.
There are as many layers as defined thresholds. Each layer corresponds to an area where the number of servers is greater than
or equal to a defined minimum threshold.
On each bin of each transmitter service area, the coverage corresponds to the pixels where the signal level from this
transmitter fulfils signal conditions defined in Conditions tab with different Cell edge coverage probabilities. There is one
coverage area per transmitter in the explorer.
On each bin of each transmitter service area, the coverage corresponds to the pixels where the best signal level received fulfils
signal conditions defined in Conditions tab. There is one coverage area per cell edge coverage probability in the explorer.
5.2 Definitions and Formulas

5.2.1 Parameters Used for CDMA2000 1xRTT Modelling
5.2.1.1 Inputs
This table lists simulation and prediction inputs (calculation options, quality targets, active set management conditions, etc.)

Tx
cn last Frequency band parameter None Last carrier number
340

cn – cn first
ic =  -------------------------
- – cn lower
 cn 
Active set upper threshold
req req req
Q pilot Q pilot  txi ic  + Q pilot None (used to determine the best server in
the active set)
Active set lower threshold
min min min
Q pilot Q pilot  txi ic  + Q pilot None (used to determine other members
of the active set)
Minimum Ec/I0 required from the
req
Q pilot  txi ic  Min. Ec/I0 - Cell parameter None cell to be the best server in the active
set
min
Q pilot  txi ic  T_Drop - Cell parameter None cell not to be rejected from the
active set
Variation of the minimum Ec/I0
req
Q pilot Delta Min. Ec/I0 - Mobility parameter None required from the cell to be the best
server in the active set
min
Q pilot Delta T_Drop - Mobility parameter None required from the cell not to be
rejected from the active set
E bFCH – DL
DL
 Q req  FCH  ----
- (Service, Terminal, Mobility) parameter None
Eb/Nt target for FCH channel on
 N t req downlink
E bSCH – DL
 ----
- Eb/Nt target for SCH channel on
DL
 Q req  SCH  N t req None
downlink
(Service, Terminal, Mobility, SCH throughput multiple) parameter
E bFCH – UL
UL
 Q req  FCH  ----
- (Service, Terminal, Mobility) parameter None
Eb/Nt target for FCH channel on
 N t req uplink
E bSCH – UL
 ----
- Eb/Nt target for SCH channel on
UL
 Q req  SCH  N t req None
uplink
(Service, Terminal, Mobility, SCH throughput multiple) parameter
N CE –U L  N I  Site parameter None
available for a site on uplink
N CE –D L  N I  Site parameter None
available for a site on downlink
N CE –U L  N I  Simulation result None
site consumed by users on uplink
N CE –D L  N I  Simulation result None
site consumed by users on downlink
Overhead
N CE –U L Site equipment parameter None the cell for common channels on
uplink
Overhead
N CE –D L Site equipment parameter None the cell for common channels on
downlink
341

FCH Number of channel elements used
N CE –U L (Terminal, site equipment) parameter None
for FCH on uplink
FCH Number of channel elements used
N CE –D L (Terminal, site equipment) parameter None
for FCH on downlink
Max Maximum number of Walsh codes
N Codes  txi ic  Simulation constraint None
available per cell (128)
Number of Walsh codes used by the
N Codes  txi ic  Simulation result None
cell

Tx DL
NR inter – techno log y Cell parameter None Inter-technology downlink noise rise
Tx UL
NR inter – techno log y Cell parameter None Inter-technology uplink noise rise


Tx m and received by m assuming the
ICP ic  ic If not defined, it is assumed that there is no inter-technology None
i
network) and ic
UL
X max Simulation constraint (global parameter or cell parameter) % Maximum uplink load factor
DL
%Powermax Simulation constraint (global parameter or cell parameter) % Maximum percentage of used power
Tx Tx UL
Term Tx DL
Rc W bps Chip rate

UL
f rake efficiency Equipment parameter None Uplink rake receiver efficiency factor
DL Downlink rake receiver efficiency

f rake efficiency Terminal parameter None
factor
SCH throughput factor (drawn
SCH
TPF DL Simulation result None following the SCH probabilities of the
service)
FCH
TPP – DL Terminal parameter bps Downlink FCH peak throughput
SCH FCH SCH

TPP – DL TP P – DL  TPF DL bps Downlink SCH bit rate
SCH throughput factor (drawn

SCH
TPF UL Simulation result None following the SCH probabilities of the
service)
FCH
TP P – UL Terminal parameter bps Uplink FCH peak throughput
SCH FCH SCH

TP P – UL TPP – UL  TPF UL bps Uplink SCH bit rate
342

W - Downlink service processing gain on
Gp
FCH – DL ----------------- None
FCH
TP P – DL FCH
W - Downlink service processing gain on

Gp
SCH – DL ----------------- None
SCH
TP P – DL SCH
W - Uplink service processing gain on

Gp
FCH – UL ----------------- None
FCH
TP P – UL FCH
W - Uplink service processing gain on

Gp
SCH – UL ----------------- None
SCH
TP P – UL SCH
DL
AF FCH Service parameter None Downlink activity factor on FCH
UL
AF FCH Service parameter None Uplink activity factor on FCH
P Sync  txi ic  Cell parameter W Cell synchronisation channel power

Cell other common channels (except
P paging  txi ic  Cell parameter W
CPICH and SCH) power
P pilot  txi ic  Cell parameter W Cell pilot power
P max  txi ic  Cell parameter W Maximum cell power

Maximum amount of power
M pooling  txi ic  Cell parameter dB
reserved for pooling
min
P FCH Service parameter W Minimum power allowed for FCH
max
P FCH Service parameter W Maximum power allowed for FCH
min
P SCH Service parameter W Minimum power allowed for SCH
max
P SCH Service parameter W Maximum power allowed for SCH
Cell FCH power for a traffic channel

P FCH  txi ic tch  Simulation result including the term AF FCH  Serv 
DL
W
on carrier ic
P FCH  txi ic   P FCH  txi ic tch  W Total FCH power on carrier ic
tch  FCH  ic  
Transmitter SCH power for a traffic

P SCH  txi ic tch  Simulation result W
channel on carrier ic
P SCH  txi ic   P SCH  ic tch  W Total SCH power on carrier ic

tch  SCH  ic  
P pilot  txi ic  + P Sync  txi ic  + P paging  txi ic  + P SCH  txi ic  Transmitter total transmitted power
P tx  txi ic  W
+ P FCH  txi ic  on carrier ic
min
max
FCH UL Terminal FCH power transmitted in

P term  ic  Simulation result including the term AF FCH  Serv  W
carrier ic
SCH Terminal SCH power transmitted on
P term  ic  Simulation result W
carrier ic
transmitted
343


L Tx
L indoor Clutter (and, optionally, frequency band) parameter Indoor loss
f Terminal parameter None Number of fingers
p Terminal parameter % Pilot power percentage

DL G macro – diversity = M Shadowing – Ec  Io – M Shadowing –Ec  Io
M Shadowing –  Eb  Nt  Result calculated from cell edge coverage probability and DL Eb/Nt DL Eb/Nt Shadowing margin
DL
None
UL
None
UL npaths
G macro – diversity = M Shadowing –  Eb  Nt  – M Shadowing –  Eb  Nt 
For Ec/I0 calculation

-------------------------------------------------------------------------------------------------------------------------------------
G Tx  G term
For DL Eb/Nt calculation
DL
LT ----------------------------------------------------------------------------------------------------------------------------------------------
- None Transmitter-terminal total loss
G Tx  G term
UL
----------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
In simulations
--------------------------------------------------------------------------------------------------------------------
G Tx  G term
P pilot  txi ic 
P c  txi ic  ------------------------------- W Chip power received at terminal
LT
FCH – DL P FCH  txi ic tch  Bit received power at terminal for
Pb  txi ic tch  ----------------------------------------- W
LT FCH on carrier ic
344
SCH – DL P SCH  txi ic tch  Bit received power at terminal for
Pb  txi ic tch  ----------------------------------------- W
LT SCH on carrier ic
DL FCH – DL SCH – DL Bit received power at terminal for

P b  txi ic tch  Pb  txi ic tch  + P b  txi ic tch  W
FCH+SCH on carrier ic
DL P tx  txi ic  Total received power at terminal

P tot  txi ic  -------------------------- W
P FCH  txi ic  + P SCH  txi ic  Total power received at terminal

DL
P traf  txi ic   ------------------------------------------------------------------
LT W from traffic channels of a transmitter
tch  ic  on carrier ic
FCH
FCH – UL P term Bit received power at transmitter for
Pb  ic  ------------ W
LT FCH on carrier ic
SCH
SCH – UL P term Bit received power at transmitter for
Pb  ic  ------------ W
LT SCH on carrier ic
UL FCH – UL SCH – UL Bit received power at transmitter for

P b  ic  Pb  ic  + P b  ic  W
SCH+FCH on carrier ic
UL
UL P b  ic  Total power transmitted by the
P tot  ic  UL UL
P b  ic  + P c  ic  = ----------------
- W
1 – p terminal on carrier ic
UL UL
P c  ic  p  Ptot  ic  W Chip received power at transmitter
a. L Tx = L total – UL on uplink and L Tx = L total – DL on downlink. For information on calculating transmitter losses on uplink and downlink,
see "UMTS, CDMA2000, TD-SCDMA, WiMAX, and LTE Documents" on page 30.
npaths
npaths
UL
UL
prediction studies, carrier power level and intra-cell interference are downgraded by the shadowing model ( M Shadowing –  Eb  Nt  or
DL
M Shadowing – Ec  Io ) while extra-cell interference level is not. Therefore, M Shadowing –  Eb  Nt  or M Shadowing – Ec  Io is set to 1 in downlink
DL
extra-cell interference calculation.
5.2.1.2 Ec/I0 Calculation

This table details the pilot quality ( Q pilot or Ec  Io ) calculations.

I intra  txi ic  DL
P tot  txi ic  W

DL Downlink extra-cell interference at
DL
txj j  i

DL
W
---------------------------------------------
DL
Tx
ni i a

I 0  ic  I intra  txi ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0 W
carrier ic b
345
E  BTS    P c  txi ic 
Q pilot  txi ic    ----c --------------------------------------------------
- Quality level at terminal on pilot for
 I0  DL None
I 0  ic  carrier ic
Term
b. In an active set, N 0 is calculated for all its members with Inter-technology downlink noise rise of the best server.
5.2.1.3 DL Eb/Nt Calculation

This table details calculations of downlink traffic channel quality ( Q tch (tch could be FCH or SCH) or  ------ ).
DL Eb
 Nt DL
Downlink intra-cell interference at

DL
I intra  txi ic   1 –  BTS  F ortho   P DL  txi ic  W
tot terminal on carrier ic

DL
txj j  i

DL
I inter – carrier  ic   j
W
---------------------------------------------
txj - at terminal on carrier ic
DL
Tx
ni i a

N tot  ic  I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0 W
carrier ic
FCH – DL
 BTS  Pb  txi ic tch  – DL
-  G FCH
------------------------------------------------------------------------------------------------------ p Quality level at terminal on a traffic
E DL DL
N tot  ic  –  1 – F ortho    BTS  P b  txi ic 
DL
ic    ----b-
DL
Q FCH  txi None channel from one transmitter for a
 N t FCH
 BTS  P b
FCH – DL
 txi ic tch  FCH channel on carrier ic b
– DL
-  G FCH
Total noise: ------------------------------------------------------------------ p
DL
N tot  ic 
Quality level at terminal for FCH


DL DL
DL f rake efficiency  Q FCH  tx k ic  using carrier ic due to combination of
Q FCH  ic  None
tx k  ActiveSet  FCH 
all transmitters of the active set
SCH – DL
 BTS  Pb  txi ic tch  – DL
-  G SCH
------------------------------------------------------------------------------------------------------ p Quality level at terminal on a traffic
E DL DL
N tot  ic  –  1 – F ortho    BTS  P b  txi ic 
DL
Q SCH  txi ic    ----b-
DL
None channel from one transmitter for a
 N t SCH
 BTS  P b
SCH – DL
 txi ic tch  SCH channel on carrier icc
– DL
-  G SCH
Total noise: ------------------------------------------------------------------ p
DL
N tot  ic 
Quality level at terminal for SCH


DL DL
DL f rake efficiency  Q SCH  tx k ic  using carrier ic due to combination of
Q SCH  ic  None
tx k  ActiveSet  SCH 
all transmitters of the active set
DL
DL Q FCH  ic  Downlink soft handover gain for FCH
 G SHO  FCH ----------------------------------------------------
- None
DL
Q FCH  BestServer ic  channel on carrier ic
DL
DL Q SCH  ic  Downlink soft handover gain for SCH
 G SHO  SCH ----------------------------------------------------
- None
DL
Q SCH  BestServer ic  channel on carrier ic
346

DL
 Q req  FCH Required transmitter FCH traffic
req
P FCH  txi ic  -----------------------  P FCH  txi ic  W channel power to achieve Eb/Nt
DL
Q FCH  ic  target at terminal on carrier ic
DL
 Q req SCH Required transmitter SCH traffic
req
P SCH  txi ic  -----------------------  P SCH  txi ic  W channel power to achieve Eb/Nt
DL
Q SCH  ic  target at terminal on carrier ic
req req req Required transmitter traffic channel
P tch  txi ic  P FCH  txi ic  + P SCH  txi ic  W
power on carrier ic
c. Calculation option may be selected in the Global parameters tab. The chosen option will be taken into account only in simulations. In point
5.2.1.4 UL Eb/Nt Calculation

This table details calculations of uplink traffic channel quality ( Q tch (tch could be FCH or SCH) or  ------ ).
UL Eb
 Nt UL
  Pb
UL UL
 ic  + P c  ic   Total power received at transmitter
UL
intra W from intra-cell terminals using carrier
I tot  txi ic  term
txi
ic

UL
 P b  ic  + P c  ic  
UL Total power received at transmitter
UL
extra W from extra-cell terminals using
txj j  i
carrier ic
  Pb
UL UL
 ic adj  + P c  ic adj  
txj j
----------------------------------------------------------------------
-

Tx
 txi ic  +  1 – F MUD   term I tot
UL W
 txi ic  +I inter – carrier  txi ic transmitter on carrier ic
UL UL tx Total noise at transmitter on carrier

N tot  txi ic  I tot  txi ic  + N 0 W
ic (Uplink interference) a
FCH – UL
 term  P b  ic  – UL
-  G FCH
-------------------------------------------------------------------------------------------------------- p Quality level at transmitter on a
E UL Tx UL
Q FCH  txi ic    ----b- N tot  txi ic  –  1 – F MUD    term  P b  ic 
UL
None traffic channel for the FCH channel
 N t UL
 term  P b
FCH – UL
 ic  on carrier icb
FCH – UL
Total noise: ------------------------------------------------
UL
 Gp

SCH – UL
 term  P b  ic  – UL
-  G SCH
-------------------------------------------------------------------------------------------------------- p Quality level at transmitter on a
E UL Tx UL
Q SCH  txi ic    ----b- N tot  txi ic  –  1 – F MUD    term  P b  ic 
UL
None traffic channel for the SCH channel
N t UL
 term  P b
SCH – UL
 ic  on carrier icc
SCH – UL
Total noise: ------------------------------------------------
UL
 Gp
347

UL
No HO: Q tch  txi ic 

UL UL
tx k  ActiveSet
 samesite 
Soft, Softer/Soft HO (No MRC): Quality level at site using carrier ic

due to combination of all
Max  Q UL tch  tx k
UL
ic    G macro – diversity transmitters of the active set located
tx k  ActiveSet at the same site and taking into
account increase of the quality due
Softer/Soft HO (MRC):
UL
Q tch  ic  None to macro-diversity (macro-diversity
gain).
Max   tch could be FCH or SCH
 UL 

UL UL
tx k ,tx l  ActiveSet  f rake efficiency  Q tch  tx k ic  Q tch  tx l ic 
In simulations,
 
tx k  samesite  tx k  UL
G macro – diversity = 1.
tx  othersite
l
UL
UL
UL Q FCH  ic  Uplink soft handover gain for FCH
 G SHO  FCH ----------------------------------------------------
- None
UL
Q FCH  BestServer ic  channel on carrier ic
UL
UL Q SCH  ic  Uplink soft handover gain for SCH
 G SHO  SCH ----------------------------------------------------
- None
UL
Q SCH  BestServer ic  channel on carrier ic
UL
 Q req  FCH Required terminal power to achieve
FCH – req
P term  ic  -  P FCH
---------------------- term  ic  W Eb/Nt target at transmitter for FCH
UL
Q FCH  ic  on carrier ic
UL
 Q req  SCH Required terminal power to achieve
SCH – req
P term  ic  -----------------------  P SCH
term  ic  W Eb/Nt target at transmitter for SCH
UL
Q SCH  ic  on carrier ic
req FCH – req SCH – req Required terminal power on carrier
P term  ic  P term  ic  + P term  ic  W
ic
tx
a. In an active set, N 0 is calculated for all its members with Inter-technology uplink noise rise of the best server.
c. Calculation option may be selected in the Global parameters tab. The chosen option will be taken into account only in simulations. In point
5.2.1.5 Simulation Results

This table contains some simulation results provided in the Cells and Mobiles tabs of the simulation property dialogue.
P tot  txi ic  – F ortho   BTS  P tot  txi ic 

DL DL
I intra  txi ic  None
DL
–  1 – F ortho   BTS   P b  txi ic 

DL
txj j  i
348

DL
W
---------------------------------------------
DL
Tx
ni i a
Total effective interference at

DL DL DL DL DL
I tot  ic  I intra  ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  W terminal on carrier ic (after
unscrambling)
DL DL Term Total received noise at terminal on
N tot  ic  I tot  ic  + N 0 W
carrier ic
  Pb
UL UL
 ic  + P c  ic   Total power received at transmitter
UL intra
term
txi
ic

UL
 P b  ic  + P c  ic  
UL Total power received at transmitter
UL extra
term
txj j  i
carrier ic
  Pb
UL UL
 ic adj  + P c  ic adj  
txj j
----------------------------------------------------------------------
-
UL UL UL Total received interference at

extra Tx
intra UL W

UL UL tx
N tot  txi ic  I tot  txi ic  + N 0 W ic
UL
UL
X  txi ic  ----------------------------
UL
UL
UL
UL intra
Tx
UL
1
--------------------------
- Cell uplink reuse efficiency factor on
E  txi ic  UL None
P tx  txi ic  
 ----------------------------- Percentage of max transmitter
DL
%Power  txi ic  -  100 None
 P max  txi ic  power used.
Simulation result available per cell

DL DL
 I extra  ic  + I inter – carrier  ic    L T
---------------------------------------------------------------------------------- + 1 – F ortho   BTS
P tx  txi ic 
 ---------------------------------------------------------------------------------------------------------------------------------
1 - + 1 – F
tch ----------
DL ortho   BTS 
CI req
DL
X  txi ic  SCH – DL FCH – DL
None Downlink load factor on carrier ic
DL Q req Q req
with CI req = -------------------- + --------------------
SCH – DL FCH – DL
Gp Gp
DL
I tot  ic 
Simulation result available per mobile: ------------------
DL
-
N tot  ic 
DL
I tot  ic 
DL
F  txi ic  -----------------------------
- None Downlink reuse factor on a carrier ic
DL
I intra  txi ic 
349

DL DL
UL UL
5.2.2 Parameters Used for CDMA2000 1xEV-DO Modelling

5.2.2.1 Inputs
This table lists simulation and prediction inputs (calculation options, quality targets, active set management conditions, etc.)

Tx
cn last Frequency band parameter None Last carrier number

cn – cn first
ic =  -------------------------
- – cn lower
 cn 
Active set upper threshold
req req req
Q pilot Q pilot  txi ic  + Q pilot None (used to determine the best server in
the active set)
Active set lower threshold
min min min
Q pilot Q pilot  txi ic  + Q pilot None (used to determine other members
of the active set)
req
Q pilot  txi ic  Min. Ec/I0 - Cell parameter None cell to be the best server in the active
set
min
Q pilot  txi ic  T_Drop - Cell parameter None cell not to be rejected from the
active set
req
Q pilot Delta Min. Ec/I0 - Mobility parameter None required from the cell to be the best
server in the active set
min
Q pilot Delta T_Drop - Mobility parameter None required from the cell not to be
rejected from the active set
Ec UL Minimum pilot quality required in

 ----
- Mobility parameter for 1xEV-DO Rev. 0 users None
 N t min – Rev0 the uplink to operate EV-DO Rev. 0
Minimum pilot quality required in

Ec
 ----
UL
- Transmitter parameter None the uplink to operate multi-carrier
 N t min – RevB EV-DO
350
Ec UL Parameter read in the 1xEV-DO Radio Bearer Selection (Uplink) table Minimum pilot quality level required
 ----
- None
 N t min for 1xEV-DO Rev. A and Rev. B users to obtain a radio bearer in the uplink
Number of subframes associated

n SF 1xEV-DO Radio Bearer Selection (Uplink) table None with the 1xEV-DO radio bearer in the
uplink
Uplink RLC peak throughput
UL
TP P – R LC 1xEV-DO Radio Bearer Selection (Uplink) table None provided by the 1xEV-DO radio
bearer
Minimum pilot quality level required
Ec DL Parameter read in the 1xEV-DO Radio Bearer Selection (Downlink)
 ----
- None to obtain a radio bearer in the
 N t min table for 1xEV-DO Rev. A and Rev. B users
downlink
Number of timeslots associated with
n TS 1xEV-DO Radio Bearer Selection (Downlink) table None the 1xEV-DO radio bearer in the
downlink
Downlink RLC peak throughput
DL
TP P – R LC Downlink 1xEV-DO Radio Bearer Table None provided by the 1xEV-DO radio
bearer
Number of EVDO channel elements
Max
N EVDO – CE  N I  Site parameter None available for a site on uplink and
downlink
Total number of EVDO channel
N EVDO – CE  N I  Simulation result None elements of a site consumed by
users on uplink and downlink
TCH Number of channel elements used
N CE – UL (Terminal, site equipment) parameter None
for TCH on uplink
Maximum number of MAC indexes
Max
N MacIndexes  txi ic  Simulation constraint None available per cell (59 for Rev0 and
114 for RevA)
Number of MAC indexes used by the
N MacIndexes  txi ic  Simulation result None
cell
Max Maximum number of EVDO users
n EVDO  txi ic  Simulation constraint (cell parameter) None
that can be connected to the cell
Number of EVDO users connected to
n EVDO  txi ic  Simulation result None
the cell

Tx DL
NRinter – techno log y Cell parameter None Inter-technology downlink noise rise
Tx UL
NRinter – techno log y Cell parameter None Inter-technology uplink noise rise


Tx m
ICP ic  ic If not defined, it is assumed that there is no inter-technology None and received by m assuming the
i
network) and ic
UL
X max Simulation constraint (global parameter or cell parameter) % Maximum uplink load factor
351

Tx Tx UL
Term Tx DL
Rc W bps Chip rate

UL
f rake efficiency Equipment parameter None Uplink rake receiver efficiency factor
UL Simulation result bps Uplink throughput

TP
UL Uplink throughput due to TCP
TP TCP – ACK Simulation result bps
aknowledgements
Downlink throughput for Broadcast/
TPBCMCS Cell parameter bps
Multicast services
Downlink peak throughput supplied

TP max – DL Simulation result bps
to the terminal
DL
TP avg Simulation result bps Downlink average cell throughput

TPD min – UL Service parameter kbps service should have in order to be
available in the uplink
TPD min – DL Service parameter kbps service should have in order to be
available in the downlink
Downlink user application
TPA
DL
f TP – Scaling  TP max – DL – TP Offset bps
throughput
f TP – Scaling Service parameter % Scaling factor
TPOffset Service parameter kbps Offset
TPD min – DL Downlink radio bearer consumption

C DL – Bearer ------------------------------------------------------------
- % for a (1xEV-DO Rev. A - Guaranteed
DL
TP P – R LC  Index DL – Bearer  Bit Rate) service user
TPD min – UL Uplink radio bearer consumption for
C UL – Bearer ------------------------------------------------------------
- % a (1xEV-DO Rev. A - Guaranteed Bit
UL
TP P – R LC  Index UL – Bearer  Rate) service user
W- Uplink service processing gain on
Gp
UL ---------- None
UL FCH
TP
G idle – power Cell parameter None Idle power gain
G MU Cell parameter None Multi user gain
P max  txi ic  Cell parameter W Max cell power

Pilot burst transmitted by the
P tx  txi ic b pilot  P max  txi ic  W
transmitter on carrier ic.
P max  txi ic  if users to support Traffic burst transmitted by the
P tx  txi ic b traffic  W
P max  txi ic   G idle – power if no user to support transmitter on carrier ic.
ER DRC Cell parameter % Error rate on the DRC channel
Pourcentage of EVDO timeslots

TS BCMCS Cell parameter % dedicated to Broadcast/Multicast
services
Pourcentage of EVDO timeslots
TS EVDO – CCH Cell parameter %
dedicated to control channels
352

Terminal power transmitted on
P term  ic  Simulation result W
carrier ic
min
max

transmitted

L Tx
L indoor Clutter (and, optionally, frequency band) parameter Indoor loss
G ACK Terminal parameter None Acknowledgement Channel gain
G RRI Terminal parameter (for 1xEV-DO Rev A terminals only) None Reverse Rate Indicator Channel gain
G DRC Terminal parameter None Data Rate Control Channel gain
G Auxiliary – pilot Terminal parameter (for 1xEV-DO Rev A terminals only) None Auxiliary Pilot Channel gain
G TCH Terminal parameter None Traffic data Channel gain
carriers Maximum number of carriers in

n max Terminal parameter None
multi-carrier mode
DL G macro – diversity = M Shadowing – Ec  Io – M Shadowing –Ec  Io
UL
None
UL npaths
G macro – diversity = M Shadowing –  Eb  Nt  – M Shadowing – Eb  Nt 
353
For Ec/I0 and Ec/Nt calculations
-------------------------------------------------------------------------------------------------------------------------------------
G Tx  G term
LT L path  L Tx  L term  L body  L indoor  M Shadowing –  Eb  Nt  None Transmitter-terminal total loss
UL
----------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
In simulations
--------------------------------------------------------------------------------------------------------------------
G Tx  G term
DL P tx  txi ic b pilot  Pilot burst received at terminal from

P tot  txi ic b pilot  ----------------------------------------- W
LT a transmitter on carrier ic
DL P tx  txi ic b traffic  Traffic burst received at terminal

P tot  txi ic b traffic  --------------------------------------------- W
UL P term Bit received power at transmitter on

P b  ic  ------------ W
LT carrier ic
UL
NRthreshold  txi ic  Cell parameter dB Cell uplink noise rise threshold
UL Cell uplink noise rise upgrading/

NR threshold  txi ic  Cell parameter dB
downgrading delta
a. L Tx = L total – UL on uplink and L Tx = L total – DL on downlink.

npaths
npaths
UL
UL
prediction studies, carrier power level and intra-cell interference are downgraded by the shadowing model ( M Shadowing – Ec  Io ) while
extra-cell interference level is not. Therefore, M Shadowing – Ec  Io is set to 1 in downlink extra-cell interference calculation.
5.2.2.2 Ec/I0 and Ec/Nt Calculations

E E E
This table details ----c  txi ic b pilot  , ----c-  txi ic b pilot  and ----c-  txi ic b traffic  calculations.
I0 Nt Nt

DL  txi ic 
I intra   0 W terminal on carrier ic (only one
 b pilot or b traffic  mobile is served at a time)
Downlink extra-cell interference

DL
DL
I extra  ic b pilot  P tot  txj ic b pilot  W based on pilot at terminal on carrier
txj j  i ic

DL
DL
I extra  ic b traffic  P tot  txj ic b traffic  W based on traffic at terminal on
txj j  i carrier ic
 Ptot  txj icadj bpilot 

DL based on pilot at terminal on carrier
I inter – carrier  ic b pilot  txj j
W
-------------------------------------------------------------
- ic
354
 Ptot  txj icadj btraffic 

DL based on traffic at terminal on
I inter – carrier  ic b traffic  txj j
W
----------------------------------------------------------------
- carrier ic
DL
Tx
ni i a
DL DL DL
DL P tot  txi ic b pilot  + I extra  ic b pilot  + I inter – carrier  ic b pilot  Total noise based on pilot received at
I 0  ic b pilot  W
DL term terminal on carrier ic
+ I inter – techno log y  ic  + N0
DL DL DL
DL P tot  txi ic b traffic  + I extra  ic b traffic  + I inter – carrier  ic b traffic  Total noise based on traffic received
I 0  ic b traffic  W
DL term at terminal on carrier ic
+ I inter – techno log y  ic  + N0
DL DL term Total noise based on pilot received at

N tot  ic b pilot  I extra  ic b pilot  + N 0 W
DL DL term Total noise based on traffic received
N tot  ic b traffic  I extra  ic b traffic  + N 0 W
at terminal on carrier ic
Q pilot  txi ic  DL
 BTS    P tot  txi ic b pilot  Pilot quality level at terminal on
Ec ---------------------------------------------------------------------- None
 ----  txi ic b pilot 
DL
I 0  ic b pilot  carrier ic
I0
DL
E  BTS    P tot  txi ic b pilot  Pilot quality level at terminal on
----c-  txi ic b pilot  ----------------------------------------------------------------------------------------------------------------
- None
Nt DL
N tot  ic b pilot  +  1 –  BTS   P tot  txi ic b pilot 
DL carrier ic
DL
E  BTS    P tot  txi ic b traffic  Traffic quality level at terminal on
----c-  txi ic b traffic  -----------------------------------------------------------------------------------------------------------------------
- None
Nt DL
N tot  ic b traffic  +  1 –  BTS   P tot  txi ic b traffic 
DL carrier ic
5.2.2.3 UL Eb/Nt Calculation

or  ------ ).
UL Eb
This table details calculations of uplink quality ( Q
 Nt UL
 Pb
UL
 ic  Total power received at transmitter
UL
txi
ic

UL
P b  ic  Total power received at transmitter
UL
extra W from extra-cell terminals using
txj j  i
carrier ic
 Pb
UL
 ic adj 
txj j
-----------------------------------
-

Tx
UL W

UL UL tx
N tot  txi ic  I tot  ic  + N 0 W ic
355

UL
 term  P b  ic 
-  G UL
-------------------------------------------------------------------------------------------------------- p
E UL Tx UL Quality level at transmitter on carrier
Q  txi ic    ----b- N tot  txi ic  –  1 – F MUD    term  P b  ic 
UL
 N t UL None
UL
ica
 term  P b  ic  UL
Total noise: ------------------------------------
UL
 Gp
UL
No HO: Q  txi ic 

UL UL
tx k  ActiveSet
 samesite 
Soft, Softer/Soft HO (No MRC):

Quality level at site using carrier ic
Max  Q UL UL due to combination of all
tch  tx k ic    G macro – diversity
tx k  ActiveSet transmitters of the active set located
at the same site and taking into
Softer/Soft HO (MRC): account increase of the quality due
UL
Q total  ic  None to macro-diversity (macro-diversity
gain).
Max  
 UL 

UL UL
tx ,tx  ActiveSet  f rake efficiency  Q tch  tx k ic  Q tch  tx l ic  In simulations,
k l   UL
tx k  samesite  tx k  G macro – diversity = 1.
tx l  othersite
UL
UL
UL Q total  ic  Uplink soft handover gain on carrier
G SHO -------------------------------------------------
- None
UL
Q  BestServer ic  ic
For 1xEV-DO Rev 0 terminal

E UL
 ----c- UL
 G p   1 + G ACK + G DRC + G TCH 
 N t min
For 1xEV-DO Rev A terminalb
UL
When the acknoledgement signal is considered
Q req UL None Eb/Nt target on uplink
Ec 
 ---- UL
-  G p   1 + G ACK + G RRI + G DRC + G TCH + G Auxiliary – Pilot 
 N t min
When the acknoledgement signal is not considered
E UL
 ----c- UL
 G p   1 + G RRI + G DRC + G TCH + G Auxiliary – Pilot 
 N t min
Q req
UL Required terminal power to achieve
req
P term  ic  -  P term
---------------------- W Eb/Nt target at transmitter on carrier
UL
Q total  ic  ic
a. Calculation option may be selected in the Global parameters tab. The chosen option will be taken into account only in simulations. In point
b. In simulations, the uplink Eb/Nt target is calculated without considering the acknowledgement signal.
5.2.2.4 Simulation Results

This table contains some simulation results provided in the Cells and Mobiles tabs of the simulation property dialogue.
356

DL DL
I intra  txi ic b traffic   1 – F ortho   BTS   P tot  txi ic b traffic  = 0 W terminal on carrier ic (only one
mobile is served at a time)

DL
DL
I extra  ic b traffic  P tot  txj ic b traffic  W based on traffic at terminal on
txj j  i carrier ic

DL based on traffic at terminal on
I inter – carrier  ic b traffic  txj j
W
----------------------------------------------------------------
- carrier ic
DL
I inter – techno log y  ic   -------------------------------------
L
Tx
 ICP
Tx m
- W interference at terminal on carrier ic
ni total n i ic a
DL DL DL Total effective interference based on

DL I intra  ic b traffic  + I extra  ic b traffic  + I inter – carrier  ic b traffic 
I tot  ic b traffic  W traffic at terminal on carrier ic (after
DL
+ I inter – techno log y  ic  unscrambling)
DL DL term Total noise based on traffic received

N tot  ic b traffic  I tot  ic b traffic  + N 0 W
at terminal on carrier ic
 Pb
UL
 ic  Total power received at transmitter
UL
txi
ic

UL
P b  ic  Total power received at transmitter
UL extra
term
txj j  i
carrier ic
 Pb
UL
 ic adj 
txj j
-----------------------------------
-

Tx
UL W

UL UL tx
N tot  txi ic  I tot  txi ic  + N 0 W ic
Number of mobiles connected to
N mobiles  txi ic  Simulation result None
transmitter txi on carrier ic
Number of (1xEV-DO Rev. A -
Guaranteed bit rate) service users
N GBR –m obiles  txi ic  Simulation result None
connected to transmitter txi on
carrier ic
Number of (1xEV-DO - Variable bit
N VBR –m obiles  txi ic  Simulation result None rate) service users connected to
transmitter txi on carrier ic
DL
DL I tot  ic b traffic 
X  txi ic  -------------------------------------
- None Cell downlink load factor on carrier ic
DL
N tot  ic b traffic 
UL
UL
X  txi ic  ----------------------------
UL
UL
UL
UL intra
Tx
357
1 Cell uplink reuse efficiency factor on

UL
E  txi ic  --------------------------- None
UL
DL DL
UL UL
5.3 Active Set Management

The mobile active set is the list of the transmitters to which the mobile is connected. The active set may consist of one or more
transmitters; depending on whether the service supports soft handoff and on the terminal active set size. Transmitters in the
mobile active set must use a frequency band with which the terminal is compatible and the pilot signal level received from
these transmitters must exceed the defined minimum RSCP threshold.
It is, however, the quality of the pilot (Ec⁄I0) that finally determines whether or not a transmitter can belong to the active set.
Cells entering the mobile’s active set must fulfill the following conditions:
• The best server (first cell entering active set)
In order for a given transmitter to enter the mobile active set as best server, the quality of this transmitter’s pilot must
be the highest one and it must exceed an upper threshold equal to the sum of the minimum Ec/I0 defined in the
properties of the best serving cell and the Delta minimum Ec/I0 defined in the properties of the mobility type. The
upper threshold is set for the carrier as defined in the cell properties and can also take into account the user mobility
type if the Delta minimum Ec/I0 defined in the mobility type is different from 0. The carrier used by the transmitters
in the active set corresponds to the best carrier of the best server. For information on the best carrier selection, see
the Technical Reference Guide.
• In order for a transmitter to enter the active set (other cells of active set):
• They must use the same carrier as the best server cell,
• The pilot quality from other candidate cells must exceed a lower threshold. The lower threshold depends both on
the type of carrier and the mobility type. It is equal to the sum of T_Drop defined in the properties of the best
server and the Delta T_Drop defined in the properties of the mobility type.
• If you have selected to restrict the active set to neighbours, the cell must be a neighbour of the best server (the
"restricted to neighbours” option is selected in the equipment properties).
For multi-carrier EVDO Rev.B users, the active set may consist of several sub-active sets, each one being associated with one
carrier. The number of sub-active sets depends on the maximum number of carriers supported by the terminal. As detailed
above, the quality of the pilot (Ec⁄I0) determines whether or not a transmitter can belong to a sub-active set. The sub-active
set associated with the best carrier is the same as the active set of a single-carrier user. For the other carriers, the uplink Ec⁄Nt
received by the best server on the best carrier and on the studied carrier determines whether or not a carrier can have a sub-
active set, and the transmitters in the sub-active sets depend on the mode supported by the terminal (locked mode or
unlocked mode):
• The Ec/Nt received by the best serving transmitter on the best carrier must exceed the minimum uplink Ec/Nt defined
in the properties of the transmitter.
• The Ec/Nt received by the best serving transmitter on the studied carrier must exceed the minimum uplink Ec/Nt
defined in the properties of the transmitter.
• When the locked mode is used, the serving transmitters must be the same in all sub-active sets. With the unlocked
mode, the serving transmitters may be different from one sub-active set to another.
5.4 Simulations
The simulation process is divided into two steps:
1. Obtaining a realistic user distribution
Atoll generates a user distribution using a Monte-Carlo algorithm, which requires traffic maps and data as input. The
resulting user distribution complies with the traffic database and maps provided to the algorithm.
358
the simulation and on the network interferences. A user may be either active or inactive. Both active and inactive
users consume radio resources and create interference.
Additionally, each 1xEV-DO Rev. 0 user is assigned a transition flag ("True" or "False") for each possible throughput
transition (from 9.6 to 19.2 kbps, 19.2 to 38.4 kbps, 38.4 to 76.8 kbps, and 76.8 to 153.6 kbps for throughput upgrading
and from 153.6 to 76.8 kbps, 76.8 to 38.4 kbps, 38.4 to 19.2 kbps, and 19.2 to 9.6 kbps for throughput downgrading).
These transition flags are based on the throughput downgrading and upgrading probabilities. If a transition flag is
"True," the user throughput can be downgraded or upgraded if necessary.
Then, Atoll randomly assigns a shadowing error to each user using the probability distribution that describes the
shadowing effect.
Finally, another random trial determines user positions in their respective traffic zone and whether they are indoors
or outdoors (according to the clutter weighting and the indoor ratio per clutter class defined for the traffic maps).
2. Modelling the network regulation mechanism
This algorithm depends on the network. Atoll uses a power control algorithm in case of CDMA2000 1xRTT networks
and a different algorithm, which mixes throughput control on downlink and power control on uplink, for CDMA2000
1xEV-DO networks.

5.4.1.1 Number of Users, User Activity Status and User Throughput
During the simulation, a first random trial is performed to determine the number of users and their activity status. The
determination of the number of users and the activity status allocation depend on the type of traffic cartography used.
[Simulation]
RandomTotalUsers=0
5.4.1.1.1 Simulations Based on User Profile Traffic Maps

user profiles with an associated mobility type and a given density (number of subscribers with the same profile per km²).
User profile traffic maps: Each polygon and line of the map is assigned a density of subscribers with given user profile and
mobility type. If the map is composed of points, each point is assigned a number of subscribers with given user profile and
mobility type.
The user profile models the behaviour of the different subscriber categories. Each user profile contains a list of services and
their associated parameters describing how these services are accessed by the subscriber.
From environment (or polygon) surface (S) and user profile density (D), a number of subscribers (X) per user profile is inferred.
X = SD
• In case of user profile traffic maps composed of lines, the number of subscribers (X)
per user profile is calculated from the line length (L) and the user profile density (D)
(nb of subscribers per km) as follows: X = L  D
• The number of subscribers (X) is an input when a user profile traffic map is composed
of points.
For each behaviour described in a user profile, according to the service, frequency use and exchange volume, Atoll calculates
the probability for the user being connected in uplink and in downlink at an instant t.
N call  d
p 0 = -------------------
-
3600
where N call is the number of calls per hour and d is the average call duration (in second).
359
Then, Atoll calculates the total number of users trying to access a certain service.
• Calculation of the number of users trying to access the service j ( n j ):
nj = X  p0
The next step determines the activity status of each user.

This steps depends on the type of service (Voice, 1xRTT data, 1xEV-DO data…).
• CDMA2000 1xRTT Services
Activity status of voice and data service users is determined as follows.
Users are always active on FCH in both directions, uplink and downlink. Therefore, we have:
Probability of being active on UL: p UL = 0
Probability of being active on DL: p DL = 0
Probability of being active both on UL and DL: p UL + DL = 1
Probability of being inactive: p inactive = 0
Thus, for voice and data services, we have:

Number of inactive users: n j  inactive  = n j  p inactive = 0
Number of users active on UL: n j  UL  = n j  p UL = 0
Number of users active on DL: n j  DL  = n j  p DL = 0
Number of users active on UL and DL both: n j  UL + DL  = n j  p UL + DL = n j
n j = n j  UL  + n j  DL  + n j  UL + DL  + n j  inactive  = n j  UL + DL 
• Voice Users
Voice users are active on uplink and downlink. However, the FCH can have inactivity periods on both links. This is modelled by
UL DL
the FCH activity factor, AF FCH and AF FCH . Therefore, all voice service users try to access the service with the following FCH
FCH UL FCH DL
throughputs, TP P – UL  AF FCH on uplink and TP P – DL  AF FCH on downlink.
FCH FCH
TP P – UL and TP P – DL are respectively the uplink and downlink FCH peak throughputs.
• Data Users
Data service users are active on uplink and downlink. FCH is always allocated but can have inactivity periods on both links; this
UL DL
is modelled by the FCH activity factor, AF FCH and AF FCH . SCH may be allocated with four possible throughputs (2x, 4x, 8x and
16xFCH peak throughput).
Therefore, data service users can access the service with different throughputs. Possible throughputs are detailed in the table
below:
SCH throughput factor Allocated throughputs
rk On UL On DL
FCH UL FCH DL
Only FCH is used - TP P – UL  AF FCH TP P – DL  AF FCH
FCH UL FCH DL
2x TP P – UL   AF FCH + 2  TPP – DL   AF FCH + 2 
FCH UL FCH DL
Both FCH and SCH are used
FCH UL FCH DL
FCH UL FCH DL
16x TP P – UL   AF FCH + 16  TP P – DL   AF FCH + 16 
FCH FCH
TP P – UL and TP P – DL are respectively the uplink and downlink FCH peak throughputs.
Then, Atoll determines the distribution of users between the different possible throughputs.
360
UL DL
In case of a data service, j, several data throughput probabilities, P k and P k , can be assigned to different throughput
factors, r k , for SCH channel.
For non-data services, these probabilities are 0.
For data service users, a random trial compliant with throughput probabilities is performed for each link in order to determine
the throughput for each user.
On uplink, we have:
rk
FCH UL
For each SCH throughput factor, r k , the number of users n j with the throughput TP P – UL   AF FCH + r k  is calculated as
follows,
rk
UL
nj = Pr  nj
k
FCH FCH UL
Therefore, the number of users n j with the throughput, TP P – UL  AF FCH , is:
rk
 nj
FCH
nj = nj –
r
k
On downlink, we have:
rk
FCH DL
For each SCH throughput factor, r k , the number of users, n j with the throughput, TP P – DL   AF FCH + r k  , is calculated as
follows,
rk
DL
nj = Pr  nj
k
FCH FCH DL
Therefore, the number of users n j with the throughput, TP P – DL  AF FCH , is:
r
 nj
FCH k
nj = nj –
rk
• CDMA2000 1xEV-DO Services

As power control is performed in the uplink only, 1xEV-DO data service users will be considered either active in the uplink or
inactive. 1xEV-DO data Rev. 0 service users can access the service with uplink throughputs of 9.6, 19.2, 38.4, 76.8 and 153.6
kbps. 1xEV-DO data Rev. A and Rev. B service users can access the service with uplink throughputs of 4.8, 9.6, 19.2, 38.4, 76.8,
115.2, 153.6, 230.4, 307.2, 460.8, 614.4, 921.6, 1,228.8 and 1,848.2 kbps.
UL UL
For each service, j, several throughput probabilities, P k , can be assigned to different throughputs TP k . The number of users
active on uplink ( n j  UL  ) and the number of inactive users ( n j  inactive  ) are calculated as follows:
 Pk
UL UL
Probability of being active on UL: p UL =  TP k 
UL
Rk
 Pk
UL UL
Probability of being inactive: p inactive = 1 –  TP k 
UL
Rk
Probability of being active on DL: p DL = 0
Probability of being active on UL and DL both: p UL + DL = 0
Therefore, we have:
Number of users active on UL: n j  UL  = n j  p UL
Number of inactive users: n j  inactive  = n j  p inactive
Number of users active on DL: n j  DL  = n j  p DL = 0
361
Number of users active on UL and DL both: n j  UL + DL  = n j  p UL + DL = 0
n j = n j  UL  + n j  DL  + n j  UL + DL  + n j  inactive  = n j  UL  + n j  inactive 
UL
Then, Atoll determines the distribution of users between the different possible throughputs, TP k . The number of users with
UL UL
the throughput TP k , n j  TP k  , is calculated as follows:
UL UL
n j  TP k  = P k  n j
Inactive users have a requested throughput equal to 0.
• The user distribution per service is an average distribution and the service of each
user is randomly drawn in each simulation. Therefore, if you compute several
simulations at once, the average number of users per service will correspond to the
calculated distribution. But if you check each simulation, the user distribution
between services is different in each of them.
• It is the same for the SCH throughput distribution between 1xRTT data service users
and the traffic throughput distribution between 1xEV-DO data service users.
• In calculations detailed above, we assume that the sum of throughput probabilities is
less than or equal to 1. If the sum of throughput probabilities exceeds 1, Atoll
 
considers normalised throughput probabilities values, P r    P r  , instead of
k k
 rk 
specified throughput probabilities P r .

k
5.4.1.1.2 Simulations Based on Sector Traffic Maps

Sector traffic maps can be based on live traffic data from OMC (Operation and Maintenance Centre). Traffic is spread over the
best server coverage area of each transmitter and each coverage area is assigned either the throughputs in the uplink and in
the downlink, or the number of users per activity status or the total number of users (including all activity statuses).
CDMA2000 1xRTT Services
• Voice Service (j)

For each transmitter, Txi, Atoll proceeds as follows:
UL
• When selecting Throughputs in Uplink and Downlink, you can input the throughput demands in UL ( TPD ) and
DL
DL ( TPD ) for each sector.
Atoll calculates the number of users active in UL and DL using the voice service in the Txi cell as follows:
UL DL
N UL = TPD
--------------- and N DL = TPD
---------------
UL DL
TP j TP j
Where,
UL
TPD is the number of kbits per second transmitted in UL in the Txi cell to provide the service j to the users (user-
defined value in the traffic map properties)
DL
TPD is the number of kbits per second transmitted in DL in the Txi cell to provide the service j to the users (user-
defined value in the traffic map properties).
UL DL
TP j and TP j correspond to the UL and DL throughputs of a user. FCH is always allocated to active users but
UL FCH UL FCH
can have inactivity periods on both links. Therefore, we have TP j = TP P – UL  AF FCH (where TP P – UL is the
UL
service FCH peak throughput on UL and AF FCH corresponds to the FCH activity factor on UL) and
DL FCH DL FCH DL
TP j = TPP – DL  AF FCH (where TP P – DL is the service FCH peak throughput on DL and AF FCH corresponds to the
FCH activity factor on DL).
Users are always active on FCH for both links. Therefore, we have following activity probabilities.
Probability of being active in UL: p UL = 0
362
Probability of being active in DL: p DL = 0
Probability of being active in UL and DL both: p UL + DL = 1
Number of users active in UL and DL both: n j  UL + DL  = max (N UL,N DL)
Number of users active in UL and inactive in DL: n j  UL  = 0
Number of users active in DL and inactive in UL: n j  DL  = 0
inactive
Number of inactive users in UL and DL: n j = 0
Therefore, all connected voice users ( n j ) are active in both links.
• When selecting Total Number of Users (All Activity Statuses), you can input the number of connected users for
each sector ( n j ).

inactive
Number of inactive users in UL and DL: n j = n j  p inactive = 0
Number of users active in UL and inactive in DL: n j  UL  = n j  p UL = 0
Number of users active in DL and inactive in UL: n j  DL  = n j  p DL = 0
Number of users active in UL and DL both: n j  UL + DL  = n j  p UL + DL = n j
Therefore, all connected users ( n j ) are active in both links.
• When selecting Number of Users per Activity Status, you can directly input the number of users active in the
uplink and downlink ( n j  UL + DL  ), for each sector.
FCH UL FCH DL
Voice service users try to access the service with the FCH throughputs, TPP – UL  AF FCH on uplink and TP P – DL  AF FCH on
downlink.
All user characteristics determined, a second random trial is performed to obtain their geographical positions.
• Data Service Users (j)
FCH is always allocated to active users but can have inactivity periods on both links. This is modelled by the FCH activity factors,
UL DL
AF FCH and AF FCH . SCH may be allocated with four possible throughputs (2x, 4x, 8x, 16xFCH peak throughput). Several
UL DL
throughput probabilities, P k and P k , can be assigned to different throughputs factor, r k , for SCH channel.
For non-data services, these probabilities are 0.
For each transmitter, Txi, Atoll proceeds as follows:

UL
• When selecting Throughputs in Uplink and Downlink, you can input the throughput demands in UL ( TPD ) and
DL
DL ( TPD ) for each sector.
Atoll calculates the number of users active in UL and DL using the service in the Txi cell as follows:
363
UL DL
N UL = TPD
--------------- and N DL = TPD
---------------
UL DL
TP j TP j
Where,
UL
TPD is the number of kbits per second transmitted in UL in the Txi cell to provide the service j to the users (user-
defined value in the traffic map properties)
DL
TPD is the number of kbits per second transmitted in DL in the Txi cell to provide the service j to the users (user-
UL DL
TP j and TP j correspond to uplink and downlink throughputs of a user.
 
  r k + AF FCH   R FCH  P r +  1 –  P r   TP P – UL  AF FC
UL UL UL UL UL FCH UL
Pj =
k  k
rk  rk 
 
  rk + AFFCH   RFCH  Prk +  1 –  Prk   TPP – DL  AFFC
DL DL DL DL DL FCH DL
Pj =
rk  rk 
FCH FCH
TP P – UL and TP P – DL are the uplink and downlink FCH peak throughputs respectively.
• In calculations detailed above, we assume that the sum of throughput probabilities is

less than or equal to 1. If the sum of throughput probabilities exceeds 1, Atoll
 
considers normalised throughput probabilities values, P r    P r  , instead of
k  k
 rk 
specified throughput probabilities P r .

k
Then, Atoll calculates the number of users per activity status and the total number of users:
Number of users active in UL and DL both: n j  UL + DL  = max (N UL,N DL)
Number of users active in UL and inactive in DL: n j  UL  = 0
Number of users active in DL and inactive in UL: n j  DL  = 0
inactive
Number of inactive users in UL and DL: n j = 0
364

inactive
Number of inactive users in UL and DL: n j = n j  p inactive = 0
Number of users active in UL and inactive in DL: n j  UL  = n j  p UL = 0
Number of users active in DL and inactive in UL: n j  DL  = n j  p DL = 0
Number of users active in UL and DL both: n j  UL + DL  = n j  p UL + DL = n j
• When selecting Number of Users per Activity Status, you can directly input the number of users active in the
uplink and downlink ( n i  UL + DL  ), for each sector.
As explained above, data service users can access the service with different throughputs. Possible throughputs are detailed in
the table below:
SCH throughput factor Allocated throughputs

rk On UL On DL
FCH UL FCH DL
Only FCH is used - TP P – UL  AF FCH TP P – DL  AF FCH
FCH UL FCH DL
2x TPP – UL   AF FCH + 2  TP P – DL   AF FCH + 2 
FCH UL FCH DL
Both FCH and SCH are used
FCH UL FCH DL
FCH UL FCH DL
Atoll determines the distribution of users with the different possible throughputs. A random trial compliant with throughput
probabilities is performed for each link in order to determine the throughput of each user.
On uplink, we have,
rk
FCH UL
For each SCH throughput factor, r k , the number of users n j with the throughput TP P – UL   AF FCH + r k  is calculated as
follows,
rk
UL
nj = Pr  nj
k
FCH FCH UL
Therefore, the number of users n j with the throughput, TP P – UL  AF FCH , is,
rk
 nj
FCH
nj = nj –
rk
On downlink, we have,
r
k FCH DL
For each SCH throughput factor, r k , the number of users, n j with the throughput, TP P – DL   AF FCH + r k  , is calculated as
follows,
k DL
nj = Pk  nj
FCH FCH DL
Therefore, the number of users n j with the throughput, TP P – DL  AF FCH , is,
rk
 nj
FCH
nj = nj –
rk
CDMA2000 1xEV-DO Services
As power control is performed in the uplink only, 1xEV-DO data service users will be considered either active in the uplink or
inactive. 1xEV-DO data Rev. 0 service users can access the service with uplink throughputs of 9.6, 19.2, 38.4, 76.8 and 153.6
kbps. 1xEV-DO data Rev. A and Rev. B service users can access the service with uplink throughputs of 4.8, 9.6, 19.2, 38.4, 76.8,
115.2, 153.6, 230.4, 307.2, 460.8, 614.4, 921.6, 1,228.8 and 1,848.2 kbps.
365
UL UL
For each service, j, several throughput probabilities, P k , can be assigned to different uplink throughputs TP k . The number
of users active in uplink ( n j  UL  ) and the number of inactive users ( n j  inactive  ) are calculated into several steps. First of
all, Atoll determines the number of users active in UL using the service j in the Txi cell.
For each transmitter, Txi, and each service j:
UL
• When selecting Throughputs in Uplink and Downlink, you can input the throughput demands in UL ( TPD ) for
each sector.
Atoll calculates the number of users active in UL using the service j in the Txi cell as follows:
UL
N UL = TPD
---------------
UL
TP j
UL
Where TPD is the number of kbits per second transmitted on UL in the Txi cell to provide the service j (user-
UL
TP j corresponds to the uplink throughput for a user.
 Pk
UL UL UL
TP j =  TP k
k
In the above calculations, we assume that the sum of throughput probabilities is less than
or equal to 1. If the sum of throughput probabilities exceeds 1, Atoll considers
 
normalised throughput probabilities values, P r    P r  , instead of specified
k  k
 rk 
throughput probabilities P r .
k
We have the following activity probabilities:
 Pk
UL UL
Probability of being active in UL: p UL =  TP k 
UL
Rk
 Pk
UL UL
UL
R
k
Therefore, we have:
Number of users active in UL: n j  UL  = N UL  p UL
Number of inactive users: n j  inactive  = N UL  p inactive
Number of users active in DL: n j  DL  = 0
Number of users active in UL and DL both: n j  UL + DL  = 0
Total number of connected users: n j = n j  UL  + n j  inactive 
We have the following activity probabilities:
 Pk
UL UL
Probability of being active in UL: p UL =  TP k 
UL
Rk
366
 Pk
UL UL
UL
R
k
Therefore, we have:
Number of users active in UL: n j  UL  = n j  p UL
Number of inactive users: n j  inactive  = n j  p inactive
Number of users active in DL: n j  DL  = 0
Number of users active in UL and DL both: n j  UL + DL  = 0
• When selecting Number of Users per Activity Status, you can directly input the number of inactive users
( n j  inactive  ) and the number of users active in the uplink ( n j  UL  ), for each sector.
The total number of connected users ( n j ) is calculated as follows
n j = n j  UL  + n j  inactive 
Then, Atoll determines the distribution of users with the different possible throughputs. The number of users with the
UL UL
throughput TP k , n j  TP k  , is calculated as follows:
UL UL
n j  TPk  = P k  n j
Inactive users have a requested throughput equal to 0.
The user distribution per service is an average distribution and the service of each user is
randomly drawn In each simulation. Therefore, if you compute several simulations at
once, the average number of users per service will correspond to the calculated
distribution. But if you check each simulation, the user distribution between services is
different in each of them.
It is the same for the SCH throughput distribution between 1xRTT data service users and
the traffic throughput distribution between 1xEV-DO data service users.
5.4.1.2 Transition Flags for 1xEV-DO Rev.0 User Throughputs

For 1xEV-DO Rev. 0 services supporting throughput downgrading, you can define the probability of the service being upgraded
UL UL UL UL UL
( P Upg – k  TP k  ) or downgraded ( P Downg – k  TP k  ) on the uplink (reverse link) for each throughput ( TP k ). The probabilities
are taken into account in order to determine if a user with a certain throughput can be upgraded or downgraded. User
throughput downgrading and upgrading occur during congestion control when the cell is over- or underloaded.
The following table shows the throughput changes that are possible when a throughput is upgraded or downgraded. The
probabilities are defined with a number from 1 to 255 for each throughput.
Possible Throughput Changes Possible Throughput Changes

During Upgrading During Downgrading
From To From To
9.6 kbps 19.2 kbps 153.6 kbps 76.8 kbps
During the generation of the user distribution, each 1xEV-DO Rev. 0 user is assigned a random number between 1 and 255 for
each possible throughput transition. When this number is lower or equal to the value of the probability, the transition flag for
this throughput transition is set to "True" meaning that this throughput transition can be performed if necessary.
367
UL
The number of 1xEV-DO Rev. 0 users with a certain throughput that can be downgraded ( n j  TP k  Downg ) and upgraded
UL
( n j  TP k  Upg ) are calculated as follows:
UL UL UL
UL P Upg – k  TP k   n j  TP k 
n j  TP k  Upg = ------------------------------------------------------------
-
255
And
UL UL UL
UL P Downg – k  TP k   n j  TP k 
n j  TP k  Downg = ------------------------------------------------------------------
-
255
The number of users with a certain throughput that can be downgraded or upgraded is
an average. Therefore, if you compute several simulations at once, the average number
of users with a certain throughput that can be downgraded or upgraded will correspond
to the calculated value. But if you check each simulation, this number is different in each
of them.
5.4.1.3 User Geographical Position

Once all the user characteristics determined, another random trial is performed to obtain their geographical positions and
whether they are indoors or outdoors according to the percentage of indoor users per clutter class defined for the traffic
maps.
5.4.2 Network Regulation Mechanism

5.4.2.1 CDMA2000 1xRTT Power Control Simulation Algorithm
CDMA2000 1xRTT network automatically regulates itself using traffic driven uplink and downlink power control on the
fundamental and supplemental channels (FCH and SCH respectively) in order to minimize interference and maximize capacity.
Atoll simulates this network regulation mechanism with an iterative algorithm and calculates, for each user distribution,
network parameters such as base station power, mobile terminal power, active set and handoff status for each terminal.
The power control simulation is based on an iterative algorithm, where in each iteration, all the mobiles selected during the
user distribution generation (1st step) try to connect to network active transmitters with a calculation area. The process is
repeated from iteration to iteration until convergence is achieved. The algorithm steps are detailed below.
Figure 5.1: CDMA2000 1xRTT Power Control Algorithm
368
5.4.2.1.1 Algorithm Initialization

Total power on carrier ic, P Tx  ic  , of base station Sj is initialised to P pilot  ic  + P sync  ic  + P paging  ic  .
UL UL
intra extra UL
Uplink received powers on carrier ic, I tot  ic  , I tot  ic  and I inter – carrier  ic  , at base station Sj are initialised to 0 W
(no connected mobile).
UL
UL I tot  S j ic 
 X k  S j ic  = -------------------------
- = 0
UL
N tot  S j ic 
5.4.2.1.2 Presentation of the Algorithm

UL
The algorithm is detailed for any iteration k. Xk is the value of the variable X at the iteration k. In the algorithm, all Q req and
DL
Q req thresholds depend on user mobility type and are defined in Service and Mobility parameters tables. All variables are
described in Definitions and formulas part.
The algorithm applies to single frequency band networks and to multi-band networks (dual-band and tri-band networks).
Multi-band terminals can have the following configurations:
• Configuration 1: The terminal can work on f1, f2 and f3 without any priority (select "All" as main frequency band
in the terminal property dialogue).
• Configuration 2: The terminal can work on f1, f2 and f3 but f1 has a higher priority (select "f1" as main frequency
band, "f2" as secondary frequency band and "f3" as third frequency band in the terminal property dialogue).
For each mobile (Mi), Atoll only considers the cells (Sj,ic) for which the pilot RSCP exceeds the minimum pilot RSCP:
P c  Sj M i ic   RSCP min  Sj ic  .
For each mobile Mi, we have the following steps:
Determination of Mi’s Best Serving Cell
For each transmitter Sj containing Mi in its calculation area and working on the main frequency band supported by the Mi’s
terminal (i.e. either f1 for a single frequency band network, or f1, f2 or f3 for a multi-band terminal with the configuration 1,
or f1 for a multi-band terminal with the configuration 2).
   BTS  P c  Sj M i ic 
Calculation of Q pilot  Sj ic M i  = ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
k DL DL DL DL Term
P tot  Sj ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0
Determination of the candidate cells, (SBS,ic).
For each carrier ic, selection of the transmitter with the highest Q pilot  Sj M i ic  ,  S BS ic   M i  .
k
Analysis of candidate cells, (SBS,ic).

For each pair (SBS,ic), calculation of the uplink load factor:
UL
UL I tot  S BS ic  UL
X k  S BS ic  = ----------------------------- + X
UL
N tot  S BS ic 
pilot
If Q pilot  S BS M i ic   Q req then (SBS,ic) is rejected by Mi
k
UL UL
If X k  S BS ic   X max , then (SBS,ic) is rejected by Mi
Else
Keep (SBS,ic) as good candidate cell
For multi-band terminals with the configuration 1 or terminals working on one frequency band only, if no good candidate cell
has been selected, Mi has failed to be connected to the network and is rejected.
For multi-band terminals with the configuration 2, if no good candidate cell has been selected, try to connect Mi to
transmitters txi containing Mi in their calculation area and working on the secondary frequency band supported by the Mi’s
terminal (i.e. f2). If no good candidate cell has been selected, try to connect Mi to transmitters txi containing Mi in their
369
calculation area and working on the third frequency band supported by the Mi’s terminal (i.e. f3). If no good candidate cell
Determination of the best carrier, icBS.
If a given carrier is specified for the service requested by Mi
ic BS  M i  is the carrier specified for the service
UL
ic BS  M i  is the cell with the lowest X k  S BS ic 

ic BS  M i  is the cell with the lowest P tx  S BS ic  k

ic BS  M i  is randomly selected

UL UL
ic BS  M i  is the first carrier where X k  S BS ic X max
Endif
Determination of the best serving cell, (SBS,icBS).
max
(S BS,ic BS) k  M i  is the best serving cell ( BestCell k  M i  ) and its pilot quality is Q pilot  M i  .
k
In the following lines, we will consider ic as the carrier used by the best serving cell.
Determination of the Active Set
For each station Sj containing Mi in its calculation area, using ic, and if neighbours are used, neighbour of BestCell k  M i 
   BTS  P c  M i S j 
Calculation of Q pilot  M i S j ic  = -------------------------------------------------
-
DL
k
I 0  ic 
Rejection of station Sj if the pilot is not received

pilot
If Q pilot  M i S j ic   Q min then Sj is rejected by Mi
k
Else Sj is included in the Mi active set

Rejection of Sj if the Mi active set is full

k
EndFor
req
Calculation of the required power for Mi, P term  M i ic  k
For each cell (Sj,ic) present in the Mi active set

Calculation of quality level on Mi traffic channel at (Sj,ic), with the minimum power allowed on traffic channel for the Mi
service
FCH – r eq SCH – r eq
FCH – U L P term  M i ic  k – 1 SCH – U L P term  M i ic  k – 1
Pb  M i S j ic  = ----------------------------------------------
- and P b  M i S j ic  = ----------------------------------------------
-
L T  M i S j  L T  M i S j 
FCH – U L
UL  term  P b  M i S j ic  – UL
-  G FCH
Q FCH  M i S j ic  k = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- p  Service 
UL FCH – U L SCH – U L
N tot  ic  –  1 – F MUD    term   P b  M i S j ic  + P b  M i S j ic  
370
SCH – U L
-  G SCH
Q SCH  M i S j ic  k = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- p  Service 
UL FCH – U L SCH – U L
N tot  ic  –  1 – F MUD    term   P b  M i S j ic  + P b  M i S j ic  
If the user selects the option “Total noise”

FCH – U L
-  G FCH
Q FCH  M i S j ic  k = -------------------------------------------------------------- p  Service 
UL
N tot  ic 
SCH –U L
-  G SCH
Q SCH  M i S j ic  k = -------------------------------------------------------------- p  Service 
UL
N tot  ic 
End For
If (Mi is not in handoff)
UL UL UL UL
Q FCH  M i  = Q FCH  M i S j ic  k and Q SCH  M i  = Q SCH  M i S j ic  k
k k

UL UL UL
Q FCH  M i  = f rake efficiency  Q FCH  M i S j ic  k
k
S j  ActiveSet

UL UL UL
Q SCH  M i  = f rake efficiency  Q SCH  M i S j ic  k
k
S  ActiveSet
j
Else if (Mi is in soft or softer/soft without MRC)

UL UL UL
Q FCH  M i  =  G macro – diversity  2 links  Max  Q FCH  M i S j ic  k 
k S j  ActiveSet
UL UL UL
Q SCH  M i  =  G macro – diversity  2 links  Max  Q SCH  M i S j ic  k 
k S j  ActiveSet
Else if (Mi is in soft/soft)

UL UL UL
Q FCH  M i  =  G macro – diversity  3 links  Max  Q FCH  M i S j ic  k 
k S j  ActiveSet
UL UL UL
Q SCH  M i  =  G macro – diversity  3 links  Max  Q SCH  M i S j ic  k 
k S j  ActiveSet
Else if (Mi is in softer/soft with MRC)
 
 
Q FCH  M i  = Max  f rake efficiency    ic    G macro – diversity  2 links
UL UL UL UL UL
Q FCH  ic  Q FCH
k  
 i AS  ActiveSet other site 
 
(same site)
 
 
Q SCH  M i  = Max  f rake efficiency    ic    G macro – diversity  2 links
UL UL UL UL UL
Q SCH  ic  Q SCH
k  
 i  ActiveSet other site 
 AS

(same site)
EndIf
UL
FCH – r eq  Q req  Service  M i  Term  M i  Mobility  M i    FCH – r eq
P term -  P FCH
 M i ic  k = --------------------------------------------------------------------------------------------------------------------------- term  M i ic  k – 1
UL
Q FCH  M i 
k
UL
SCH – r eq  Q req  Service  Mi  Term  M i  Mobility  M i  SCH_rate_multiple   SCH – r eq
P term -  P SCH
 M i ic  k = ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------- term  M i ic  k – 1
UL
Q SCH  M i 
k
req FCH – r eq SCH – r eq

P term  M i ic  k = P term  M i ic  k + P term  M i ic  k
371
req min
If P term  M i ic  k  P term  M i  then
min
FCH – r eq P term  M i S j  – r eq
P term -  P FCH
 M i ic  k = ------------------------------ term  M i ic  k
req
P term  M i  k
min
SCH – r eq P term  M i S j  – r eq
P term -  P SCH
 M i ic  k = ------------------------------ term  M i ic  k
req
P term  M i  k
EndIf
FCH – r eq max
If P term  M i ic  k  P term  M i  then Mi cannot select any station and its active set is cleared
req max
If P term  M i ic  k  P term  M i  and Mi uses SCH then:
Downgrading the service SCH throughput:

req max SCH FCH
While P term  M i ic  k  P term  M i  and TP P – UL  Service  M i    TPP – UL  Service  M i    2
SCH
SCH TP P – UL  Service  M i  
TP P – UL  Service  M i    -----------------------------------------------------
-
2
SCH – r eq
SCH – r eq P term  M i ic  UL
 Q req  Service  M i  Term  M i  Mobility  M i  TP P – UL  Service  Mi     SCH
SCH
P term  M i ic  k = -----------------------------------------k  ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
2 UL
 Q  Service  M  Term  M  Mobility  M  TP
SCH
 Service  M   2   
req i i i P – UL i SCH

EndWhile
req max
If P term  M i ic  k  P term  M i  then Mi will not use SCH
Endif
Endif
If the required number of channel elements exceeds the available quantity in the site of Sj (Best server of Mi) and Mi uses SCH
then:
Downgrading the service SCH throughput:
Max SCH FCH
While N CE –U L  M i   N CE –U L  S j  and TP P – UL  Service  M i    TP P – UL  Service  M i    2
SCH
SCH TP P – UL  Service  M i  
TP P – UL  Service  M i    -----------------------------------------------------
-
2
SCH
SCH N CE –U L  M i  k
N CE –U L  M i  k = -----------------------------
-
2
SCH – r eq SCH – UL
SCH – r eq P term  M i ic  Q req  Service  M i  Term  M i  Mobility  M i  TP P – UL  Service  Mi   
SCH
P term  M i ic  k = -----------------------------------------k  --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
SCH – UL
2 Q  Service  M  Term  M  Mobility  M  TP
SCH
 Service  M   2  
req i i i P – UL i

SCH FCH
N CE –U L  M i  k = N CE –U L  M i  k + N CE –U L  M i  k
EndWhile
Endif
Downlink Power Control
If Mi uses an SCH on the downlink

For each cell (Sj,ic) in Mi FCH active set
Calculation of quality level on (Sj,ic) FCH at Mi, with the minimum power allowed on FCH for the Mi service
372
min
FCH – D L P FCH  Service  M i  
Pb  M i S j ic  = ----------------------------------------------
L T  M i S j 
FCH – D L
DL  BTS  P b  M i S j  – DL
Q FCH  M i -  G FCH
S j ic  k = ------------------------------------------------------------------------------------------------------------- p  Service  M i  
DL DL
N tot  ic  –  1 – F ortho    BTS  P b  M i S j ic 

FCH – D L
DL  BTS  P b  M i S j 
Q FCH  M i S j ic  k = -----------------------------------------------------
DL
N tot  ic 
If cell (Sj,ic) in Mi SCH active set

Calculation of quality level on (Sj,ic) SCH at Mi, with the minimum power allowed on SCH for the Mi service
min
SCH – D L P SCH  Service  M i  
Pb  M i S j ic  = ----------------------------------------------
L T  M i S j 
SCH – D L
DL  BTS  P b  M i S j  – DL
-  G SCH
Q SCH  M i S j ic  k = ------------------------------------------------------------------------------------------------------------- p  Service  Mi  
DL DL
N tot  ic  –  1 – F ortho    BTS  P b  M i S j ic 

SCH – D L
DL  BTS  P b  M i S j 
Q SCH  M i S j ic  k = -----------------------------------------------------
DL
N tot  ic 
EndIf
End For
Recombination of the first f active set links (f is the number of fingers of the Mi terminal): only quality levels from the first f
cells (Sf,ic) of active set are recombined.

DL DL DL
Q FCH  M i  = f rake efficiency  Q FCH  M i S j ic  k
k
S f  ActiveSet  FCH 

DL DL DL
Q SCH  M i  = f rake efficiency  Q SCH  M i S j ic  k
k
S f  ActiveSet  SCH 
Do
For each cell (Sj,ic) in Mi FCH active set
Calculation of the required power for DL traffic channel between (Sj,ic) and Mi:
DL FCH
req  Q req  Service  M i  Term  M i  Mobility  M i  TP P – DL  Service  M i     FCH
-  P min
P FCH  M i S j ic  k = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ FCH  Service  M i  
DL
Q FCH  M i 
k
req max
If P FCH  M i S j ic  k  P FCH  Service  M i   then  S j ic  is excluded from Mi active set
DL
Recalculation of a decreased Q req
If cell (Sj,ic) in Mi SCH active set

Calculation of the required power for DL traffic channel between (Sj,ic) and Mi:
DL SCH
req  Q req  Service  M i  Term  M i  Mobility  M i  TP P – DL  Service  M i     SCH
-  P min
P SCH  M i S j ic  k = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ SCH  Service  M i  
DL
Q SCH  M i 
k
Downgrading the service SCH throughput (only for (Sj,ic) best server cell of Mi):
req max SCH
While P SCH  M i S j ic  k  P SCH  Service  M i  TP P – DL  Service  M i   
req SCH FCH

Or P tx  S j ic  k + P tch  M i S j ic  k  P max  S j ic  and TPP – DL  Service  M i    TP P – DL  Service  M i    2
373
SCH
SCH TP P – DL  Service  M i  
TP P – DL  Service  M i   = -----------------------------------------------------
2
req DL SCH
req P SCH  M i S j ic  k  Q req  Service  M i  Term  M i  Mobility  M i  TPP – DL  Service  M i     SCH
P SCH  M i S j ic  k = --------------------------------------  ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
2 DL
SCH
 Service  M   2   
req i i i P – DL i SCH
req req req

P tch  M i S j ic  k = P SCH  M i S j ic  k + P FCH  M i S j ic  k
EndWhile
req max req
If P SCH  M i S j ic  k  P SCH  Service  M i   or P tx  S j ic  k + P tch  M i S j ic  k  P max  S j ic  then Mi will not use SCH
Endif
Max SCH FCH
While N CE –D L  M i   N CE –D L  S j  and TP P – DL  Service  M i    TP P – DL  Service  Mi    2
SCH
TP P – DL  Service  M i   = -----------------------------------------------------
2
SCH
SCH N CE –D L  M i  k
N CE –D L  M i  k = -----------------------------
-
2
req DL SCH
P SCH  M i S j ic  k = --------------------------------------  ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
2 DL
SCH
 Service  M   2   
req req req

FCH SCH
N CE –D L  M i  k = N CE –D L  M i  k + N CE –D L  M i  k
EndWhile
Max
If N CE –D L  M i   N CE –D L  S j  then Mi will not use SCH
Endif
Max SCH FCH
While N Codes  M i   N Codes  S j ic  and TP P – DL  Service  M i    TP P – DL  Service  Mi    2
SCH
TP P – DL  Service  M i   = -----------------------------------------------------
2
SCH
SCH N Codes  M i  k
N Codes  M i  k = ----------------------------
-
2
req DL SCH
P SCH  M i S j ic  k = --------------------------------------  ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
2 DL
SCH
 Service  M   2   
req req req

FCH SCH
N Codes  M i  k = N Codes  M i  k + N Codes  M i  k
EndWhile
Max
If N Codes  M i   N Codes  S j ic  then Mi will not use SCH
Endif
Endif
EndFor
Recombination of the first f active set links (f is the number of fingers of the Mi terminal): only quality levels from the first f
cells (Sf,ic) of active set are recombined.

DL DL DL
Q FCH  M i  = f rake efficiency  Q FCH  M i S f ic  k
k
S f  ActiveSet  FCH 
374

DL DL DL
Q SCH  M i  = f rake efficiency  Q SCH  M i S f ic  k
k
S  ActiveSet  SCH 
f
DL DL
While Q k  M i   Q req  Service  M i  Mobility  M i   and Mi FCH active set is not empty
DL DL
And Q k  M i   Q req  Service  M i  Mobility  M i   (if SCH active set is not empty)
Endif
Uplink and Downlink Interference Updates
Update of interference on active mobiles only (old contributions of mobiles and stations are replaced by the new ones)
For each cell (Sj,ic)
UL
Update of N tot  S j ic 
EndFor
For each mobile Mi
DL
Update of N tot  ic 
EndFor
Control of Radio Resource Limits (Walsh Codes, Cell Power and Site Channel Elements)
For each cell (Sj,ic) on a site Nl
P tx  S j ic  DL
While -------------------------k  %Power max
P max
req
Rejection of mobile with highest P tch  S j M b ic  k for the lowest service priority
EndWhile
EndFor
For each site Nl
The list of rejected mobiles for the site Nl is L rejected  N l 
If the equipment installed on Nl supports power pooling between transmitters

Activation of power pooling between transmitters for each cell (Sj,ic) containing rejected users
Control of the available power for the other cells (Si,ic) of the site where power pooling between transmitters is not activated

DL
If  %Power max  P max – P tx  S i ic  k   0
 S i ic 
Si  Nl
Then, the power unused by the cells (Si,ic) of the site can be allocated to cells (Sj,ic)
Sort of all the rejected mobiles by priority in a descending order and by simulation rank in a descending order
For the first mobile Mb of the list ( M b  L rejected  N l  )
req DL
If P tx  S j ic  k + P tch  S j M b ic  k  %Power max  P max + M Pooling  S j ic 
Mb is reconnected
EndIf
EndFor
EndIf
EndFor
Max
While N Codes  S j ic  k  N Codes  S j ic 
375
Rejection of last admitted mobile

EndFor
For each site (Node B) Nl
Max
While N CE – DL  N I  k  N CE – DL  N I 
req
Rejection of mobile with highest P tch  M i S j  k for the lowest service priority
Max
While N CE – UL  N I  k  N CE – UL  N I 
req
Rejection of mobile with highest P term  M i ic  k for the lowest service priority
EndFor
UL UL
For each cell (Sj,ic) with X  S j ic   X max
Rejection of a mobile with the lowest service priority

EndFor
UL UL
While at least one cell with X  S j ic   X max exists
5.4.2.1.3 Convergence Criterion

The convergence criteria are evaluated at each iteration, and can be written as follow:
  P tx  ic  k – P tx  ic  k – 1   DL DL
N user  ic  k – N user  ic  k – 1 
 DL = max  int  ma x -------------------------------------------------  100  int  ma x ------------------------------------------------------------  100 
  Stations P tx  ic  k   Stations
N
DL
 ic  
user k
  UL UL
I tot  ic  k – I tot  ic  k – 1   UL UL
 UL = max  int  ma x --------------------------------------------------  100  int  ma x ------------------------------------------------------------  100 
  Stations UL
I  ic    Stations N
UL
 ic  
tot k user k
1st case: Between two successive iterations,  UL and  DL are lower (  ) than their respective thresholds (defined when
creating a simulation).
Example: Let us assume that the maximum number of iterations is 100, UL and DL convergence thresholds are set to 5. If
 UL  5 and  DL  5 between the 4th and the 5th iteration, Atoll stops the algorithm after the 5th iteration. Convergence has
been achieved.
2nd case: After 30 iterations,  UL or/and  DL are still higher than their respective thresholds and from the 30th iteration,  UL
or/and  DL do not decrease during the next 15 successive iterations.

Examples: Let us assume that the maximum number of iterations is 100, UL and DL convergence thresholds are set to 5.
1. After the 30th iteration,  UL and/or  DL equal 100 and do not decrease during the next 15 successive iterations: Atoll stops
the algorithm at the 46th iteration. Convergence has not been achieved.
2. After the 30th iteration,  UL and/or  DL equal 80, they start decreasing slowly until the 40th iteration (without going under
the thresholds) and then do not change during the next 15 successive iterations: Atoll stops the algorithm at the 56th iteration
without achieving convergence.

If  UL and/or  DL are still strictly higher than their respective thresholds, the simulation has not converged (specific
divergence symbol).
If  UL and  DL are lower than their respective thresholds, the simulation has converged.
376
5.4.2.2 CDMA2000 1xEV-DO Power/Data Rate Control Simulation Algorithm
Figure 5.2: CDMA2000 1xEVDO Power Control Algorithm
In a CDMA2000 1xEV-DO system, power control is performed in the uplink only. In the downlink, the transmitter transmits at
the full power (Pmax) when a connection is established. Instead of power control, there is a data rate control based on the C/
I ratio calculated at the mobile. For each distribution of users, Atoll simulates the power control mechanism for the UL and
the data rate control for the DL.
The simulation uses an iterative algorithm, where in each iteration, all the 1xEV-DO data service users selected during the user
distribution generation (1st step) try to connect to network active transmitters with a calculation area. Atoll considers the
guaranteed bit rate service users first, in the order established during the generation of the user distribution, and then, it
processes the variable bit rate service users, in the order established during the generation of the user distribution.
The process is repeated from iteration to iteration until convergence is achieved. The algorithm steps are detailed below.
5.4.2.2.1 Algorithm Initialization

UL UL
intra extra UL
Uplink received powers on carrier ic, I tot  ic  , I tot  ic  and I inter – carrier  ic  , at base station Sj are initialised to 0 W
(no connected mobile).
UL
UL I tot  S j ic 
 X k  S j ic  = -------------------------
- = 0
UL
5.4.2.2.2 Presentation of the Algorithm

The algorithm is detailed for any iteration k. Xk is the value of the variable X at the iteration k.
Ec UL
In the algorithm,  ----- is the minimum pilot quality level required in the uplink to operate 1xEV-DO Rev. 0. This
N t min – Rev0
threshold depends on the user mobility type and is defined in the Mobility parameters table.
Ec
 ----
UL
-
 N t min – RevB is the minimum pilot quality level required in the uplink to operate EV-DO multi-carrier. This threshold is
defined in the Transmitter properties dialogue.
E UL
For 1xEV-DO Rev. A and Rev. B users, the value of  ----c- depends on the user requested throughput. This throughput can
 N t min
be obtained by using a certain uplink 1xEV-DO radio bearer ( Index UL – Bearer ) in a certain number of subframes ( n SF ).
Ec
 ----
UL
- is the value defined in the 1xEV-DO Radio Bearer Selection (Uplink) table for the combination (radio bearer Index,
 N t min
377
mobility and number of subframe) providing the user requested throughput. Two values are available for this parameter, one
when the service uplink mode is "Low Latency" and another one for high capacity services.
All variables are described in Definitions and formulas part (see "Definitions and Formulas" on page 340).
The algorithm applies to single frequency band networks and to multi-band networks (dual-band and tri-band networks).
Multi-band terminals can have the following configurations:
• Configuration 1: The terminal can work on f1, f2 and f3 without any priority (select "All" as main frequency band
in the terminal property dialogue).
• Configuration 2: The terminal can work on f1, f2 and f3 but f1 has a higher priority (select "f1" as main frequency
band, "f2" as secondary frequency band and "f3" as third frequency band in the terminal property dialogue).
For each mobile (Mi), Atoll only considers the cells (Sj,ic) for which the pilot RSCP exceeds the minimum pilot RSCP:
P c  Sj M i ic b pilot   RSCP min  Sj ic  .
For each mobile Mi, we have the following steps:
Determination of Mi’s Best Serving Cell
For each transmitter Sj containing Mi in its calculation area and working on the main frequency band supported by the Mi’s
terminal (i.e. either f1 for a single frequency band network, or f1, f2 or f3 for a multi-band terminal with the configuration 1,
or f1 for a multi-band terminal with the configuration 2).
   BTS  P c  Sj M i ic ,b pilot 
Calculation of Q pilot  Sj ic M i  = ------------------------------------------------------------------------------------------------------------------------------------------------------------------
k DL DL DL Term
P tot  Sj ic ,b pilot  + I extra  ic ,b pilot  + I inter – carrier  ic ,b pilot  + N 0
Determination of the candidate cells, (SBS,ic).
For each carrier ic, selection of the transmitter with the highest Q pilot  Sj M i ic  ,  S BS ic   M i  .
k
Analysis of candidate cells, (SBS,ic).

For each pair (SBS,ic), calculation of the uplink load factor:
UL
UL I tot  S BS ic  UL
X k  S BS ic  = ----------------------------- + X
UL
N tot  S BS ic 
pilot
If Q pilot  S BS M i ic   Q req then (SBS,ic) is rejected by Mi
k
UL UL
If X k  S BS ic   X max , then (SBS,ic) is rejected by Mi
Else
Keep (SBS,ic) as good candidate cell
For multi-band terminals with the configuration 1 or terminals working on one frequency band only, if no good candidate cell
For multi-band terminals with the configuration 2, if no good candidate cell has been selected, try to connect Mi to
transmitters txi containing Mi in their calculation area and working on the secondary frequency band supported by the Mi’s
terminal (i.e. f2). If no good candidate cell has been selected, try to connect Mi to transmitters txi containing Mi in their
calculation area and working on the third frequency band supported by the Mi’s terminal (i.e. f3). If no good candidate cell
Determination of the best carrier, icBS.
If a given carrier is specified for the service requested by Mi
ic BS  M i  is the carrier specified for the service
UL
ic BS  M i  is the cell with the lowest X k  S BS ic 
378
ic BS  M i  is the cell with the lowest P tx  S BS ic  k

ic BS  M i  is randomly selected

UL UL
ic BS  M i  is the first carrier where X k  S BS ic   X max
Endif
Determination of the best serving cell, (SBS,icBS).
max
(S BS,ic BS) k  M i  is the best serving cell ( BestCell k  M i  ) and its pilot quality is Q pilot  M i  .
k
In the following lines, we will consider ic as the carrier used by the best serving cell.
Determination of the Active Set
For each station Sj containing Mi in its calculation area, using ic, and if neighbours are used, neighbour of SBS(Mi)
DL
 BTS    P tot  M i S j ic b pilot 
Calculation of Q pilot  M i S j ic  = ----------------------------------------------------------------------------
-
DL
k
I 0  ic b pilot 

min
If Q pilot  M i S j ic   Q pilot then Sj is rejected by Mi
k
Else Sj is included in the Mi active set

Rejection of Sj if the Mi active set is full

k
EndFor
Determination of the Sub-active Sets of a EVDO Multi-carrier User
For multi-carrier EV-DO Rev.B service users with a 1xEV-DO Rev. B capable terminal, calculation of the quality level received
by the best serving cell (SBS,ic)
max
Ec
 ----
UL  term  P term  M i 
-
 N t  S BS ic  = ------------------------------------------
-
UL
L T  N tot  S BS ic 
E c UL E c UL
If  -----  S BS ic    -----  S BS  then EV-DO multi-carrier is not activated.
Nt N t min
For each transmitter Sj containing Mi in its calculation area and using other EV-DO carriers, icn (either icn belongs to f1 for a
single frequency band network, or it belongs to f1, f2 or f3 for a multi-band terminal)
Calculation of Q pilot  Sj ic n M i 
k
Ranking of carriers, icn,according to Q pilot  Sj ic n M i  , from the highest to the lowest value.
k
For each received carrier, icn, in the defined order:

carriers
While n max  M i  is not exceeded
Determination of the best transmitter of the sub-active set, based on the received pilot quality, Q pilot  Sj ic n M i  .
k
Determination of the other transmitters of the sub-active set, based on the received pilot quality, Q pilot  Sj ic n M i  .
k
Calculation of the quality level received by the best serving cell (SBS,icn)
max
Ec
 ----
UL  term  P term  M i 
-  S BS ic n  = ------------------------------------------
-
 N t UL
L T  N tot  S BS ic n 
379
E UL E UL
If  ----c-  S BS ic n    ----c-  S BS  , then no sub-active set is associated with icn
 N t  N t min
If the user terminal supports the ’Locked’ mode, analysis of the sub-active set
If a transmitter of the studied sub-active set does not belong to the sub-active set associated with the best carrier, then it is
removed.
If the studied sub-active set does not contain the same transmitters as the sub-active set associated with the best carrier, then
the studied sub-active set is removed.
EndIf
Endwhile
EndFor
req
Calculation of the required power for Mi, P term  M i ic  k
For each cell (Sj,ic) present in the Mi active set or sub-active set
Calculation of quality level on Mi traffic channel at (Sj,ic), with the minimum power allowed on traffic channel for the Mi
service
req
UL P term  M i ic k – 1
P b  M i S j ic  = ---------------------------------------
-
L T  M i S j 
UL
UL  term  P b  M i S j ic 
-  G UL
Q  M i S j ic  k = -------------------------------------------------------------------------------------------------------------- p  Service 
UL Tx UL
N tot  ic  –  1 – F MUD    term  P b  M i S j ic 

UL
UL  term  P b  M i S j ic 
-  G UL
Q  M i S j ic  k = ---------------------------------------------------- p  Service 
UL
N tot  ic 
End For
If (Mi is not in handoff)
UL UL
Q total  M i  = Q  M i S j ic 
k

UL UL UL
Q total  M i  = f rake efficiency  Q  M i S j ic  k
k
S j  ActiveSet
Else if (Mi is in soft or softer/soft without MRC)

UL UL UL
Q total  M i  = Max  Q  M i S j ic  k    G macro – diversity  2 links
k I AS  ActiveSet
Else if (Mi is in soft/soft)

UL UL UL
Q total  M i  = Max  Q  M i S j ic  k    G macro – diversity  3 links
k I  ActiveSet
AS
Else if (Mi is in softer/soft with MRC)
 
 
Q total  M i  = Max  f rake efficiency      G UL
UL UL UL UL
Q  M i S j ic  k Q  M i S j ic  k macro – diversity  2 links
k 
 i AS  ActiveSet othersite 
 
(same site)
EndIf
UL
req Q req  Service  M i  Term  M i  Mobility  M i  
-  P req
P term  M i ic  k = --------------------------------------------------------------------------------------------------------------- term  M i ic  k – 1
UL
Q total  M i 
k
380
If the service of Mi uses Transmission Control Protocol (TCP)

For the best server cell (Sk,ic) of Mi
Calculation of the Mi downlink application throughput
DL
Calculation of N tot  ic b traffic 

DL
 Ptot  Sj i c btraffic  + ----------------------------------------------------------------

N tot  ic b traffic  = - + N0
j j  k
Calculation of the maximum throughput supplied to Mi, TP max – DL  M i S k 
Calculation of pilot quality level at Mi

DL
E P tot  M i S k ic b pilot 
----c-  M i S k ic b pilot  = --------------------------------------------------
-
Nt DL
N tot  ic b pilot 
If Mi is a 1xEV-DO Rev. 0 service user, determination of the peak throughput from the graph (Peak throughput=f(C/I)) specified
for the mobility type of Mi
E
TP max – DL  M i S k  = f  ----c-  M i S k ic b pilot 
 Nt 
If Mi is a 1xEV-DO Rev. A service user, selection of the downlink 1xEV-DO radio bearer ( Index DL – Bearer ): Index DL – Bearer
Ec Ec DL
where -----  M i S k ic b pilot    -----  Index DL – Bearer 
Nt Nt min
If Mi is a 1xEV-DO Rev. B service user, selection of the downlink 1xEV-DO radio bearer ( Index DL – Bearer ): Index DL – Bearer
E E DL
where ----c-  M i S k ic b pilot    ----c-  Index DL – Bearer  and the modulation scheme is supported by the terminal.
Nt  Nt  min
DL
TP P – R LC  Index DL – Bearer 
Determination of the peak throughput: TP max – DL  M i S k  = ------------------------------------------------------------
-
n TS  Index DL – Bearer 
DL
TP A  M i S k  = TP max – DL  M i S k   f TP – Scaling  Service  Mi   – TP Offset  Service  Mi  
UL
Determination of the uplink throughput due to TCP acknowledgements, TP TCP – ACK  M i S k  from the graph (UL Thr due to
TCP=f(DL Thr) specified for the service of Mi
UL DL
TP TCP – ACK  M i S k  = f  TP A  M i S k  
UL UL UL
Determination of the nearest lower and higher supported throughputs ( TP low and TPhigh ) for TP TCP – ACK  M i S k 
UL UL UL UL UL UL
For TPlow and TP high , calculation of CI req  TP low  and CI req  TP high 
E UL
CI req =  ----c-
UL
  1 + G DRC + G TCH  for DO Rev.0 terminals
N t min
And
E UL
CI req =  ----c-
UL
  1 + G DRC + G TCH + G RRI + G Auxiliary – pilot  for DO Rev.A and DO Rev.B terminals
 N t min
EndFor
UL UL UL UL UL UL
Linear interpolation of CI req  TP TCP – ACK  between CI req  TP low  and CI req  TP high 
UL UL UL UL UL
CI req = CI req  TP  + CI req  TP TCP – ACK 
UL UL W
Q req = CI req  -----------------------------------------------
UL UL
-
 TP + TPTCP – ACK 
EndIf
381
req req min

P term  M i ic  k = Max  P term  M i ic k P term  M i S j  
For DO Rev.0 and DO Rev.A users

req max
If P term  M i ic  k  P term  M i  then:
Downgrading the traffic channel throughput

req max
While P term  M i ic  k  P term  M i 
And
UL
TP  Service  M i    9.6kbps for 1xEV-DO Rev. 0 users,
UL
TP  Service  M i    4.8kbps for (1xEV-DO Rev. A - Variable bit rate) service users,
UL
TP  Service  M i    4.8kbps for single-carrier 1xEV-DO Rev. B service users,
UL
TP  Service  M i    TPD min – UL  Service  M i   for (1xEV-DO Rev. A - Guaranteed bit rate) service users,
req
req P term  M i ic  k
-  TP UL
P term  M i ic  k = ---------------------------------------------- low  Service  M i  
UL
( TP low  Service  M i   is the nearest lower supported
UL
TP  Service  M i  
throughput)
For 1xEV-DO Rev. 0, (1xEV-DO Rev. A - Variable bit rate) and single-carrier 1xEV-DO Rev. B service users,
UL UL
TP  Service  M i   = TP low  Service  M i  
UL
For (1xEV-DO Rev. A - Guaranteed bit rate) service users, TP  Service  M i   = TPD min – UL  Service  M i  
EndWhile
req max
If P term  M i ic  k  P term  M i  then Mi is rejected
req
P term  M i ic  = P term  M i ic  k
req
For (1xEV-DO Rev. A - Guaranteed bit rate) service users, P term  M i ic  = P term  M i ic  k  C UL – Bearer
Endif
Endif
For multi-carrier 1xEV-DO Rev. B service users, load balancing between carriers is performed. The available terminal power is
shared between each carrier as follows:
The maximum terminal power is allocated to the best carrier ( ic 1 ).
UL
Calculation of the traffic channel throughput ( TP  Service   M i  ic 1   )
UL
Downgrading the traffic channel throughput ( TP  Service   M i  ic 1   )
req max UL
While P term  M i ic 1  k  P term  M i  and TP  Service   M i  ic 1    153 6kbps
req
req P term  M i ic 1  k
-  TP UL
P term  M i ic 1  k = -------------------------------------------------------------
UL
low  Service  M i   ( TP low  Service  M i   is the nearest lower supported
UL
TP  Service   M i  ic 1  
throughput)
UL UL
TP  Service   M i  ic 1   = TP low  Service  M i  
EndWhile
req max
If P term  M i ic 1  k  P term  M i  , then Mi is not connected to cells of the sub-active set associated with ic 1 .
382
Else the remaining terminal power is allocated to the second best carrier ( ic 2 ) and the traffic channel throughput
UL
TP  Service   M i  ic 2   is calculated.
The same process is repeated for the other carriers in Mi ’s active set as long as the remaining terminal power is sufficient to
obtain the lowest bearer allowed.
If no sub-active set can be used, then Mi is rejected.
Endif
UL
Calculation of TP  Service  M i   for each combination of carriers
 TP
UL UL
TP  Service  M i   =  Service   M i  ic   where n corresponds to the number of carriers in the combination.
ic = 1
UL
Selection of the configuration providing the highest throughput, Max  TP  Service  M i    .
UL UL UL
If Max  TP  Service  M i     TP high  Service  M i   ( TP high  Service  M i   is the nearest supported throughput higher than
the requested throughput)
Downgrading the traffic channel throughput
UL UL UL
While Max   TP  Service  M i     TP high  Service  M i    and TP  Service   M i  ic    153 6kbps
EndWhile
EndIf
Endfor
Uplink Interference Updates
Update of interference on active mobiles only (old contributions of mobiles and stations are replaced by the new ones)
UL
EndFor
Control of Radio Resource Limits (Number of EVDO users, MAC Indices and Site Channel Elements)

Max
While n EVDO  S j ic   n EVDO  S j ic 
Rejection of the last admitted mobile

EndFor
Max
While N MacIndexes  S j ic   N MacIndexes  S j ic 

EndFor
For each site (Node B) Nl
Max
While N EVDO – CE  N I  k  N EVDO – CE  N I 

EndFor
UL UL UL
For each cell (Sj,ic) with NR  S j ic   NR threshold  S j ic  + NRthreshold  S j ic 
UL UL UL
While NR  S j ic   NR threshold  S j ic  + NRthreshold  S j ic  and there is at least one mobile that can be downgraded
383
Downgrading the traffic channel throughput for all 1xEV-DO Rev. 0 mobiles for which the throughput transition flag is set to
"True".
UL
Endwhile
UL UL UL
For each cell (Sj,ic) with NR  S j ic   NR threshold  S j ic  – NRthreshold  S j ic 
UL UL UL
While NR  S j ic   NR threshold  S j ic  – NRthreshold  S j ic  and there is at least one mobile that can be upgraded
Upgrading the traffic channel throughput for all 1xEV-DO Rev. 0 mobiles for which the throughput transition flag is set to
"True". (only 1xEV-DO Rev. 0 mobiles which have not been downgraded can be upgraded. In addition, the upgraded
throughput cannot exceed the initial user throughput drawn by the Monte-Carlo algorithm. This means that only mobiles
downgraded during the uplink power control step can be upgraded).
UL
Endwhile
UL UL
For each cell (Sj,ic) with X  S j ic   X max
Rejection of a mobile with the lowest service priority

EndFor
UL UL
While at least one cell with X  S j ic   X max exists
Downlink Data Rate Control
For each mobile Mi connected to a cell (Sk,ic)

DL
Calculation of N tot  ic b traffic 
For each cell (Sj,ic) ( k  j )
Determination of the number of mobiles connected to the cell (Sj,ic), N mobiles  S j ic 
If N mobiles  S j ic  = 0 then, P tx  S j ic b traffic  = G idle – power  P max  S j ic 
Else P tx  S j ic b traffic  = P max  S j ic 
EndFor
 Ptot  Sj ic btraffic  + N0

DL DL term
N tot  ic b traffic  =
j j  k
EndFor
Calculation of the maximum throughput supplied to Mi, TP max – DL
For the Mi’s best server cell (Sk,ic) (in the active set or each sub-active set)
Calculation of pilot quality level at Mi
DL
E P tot  M i S k ic b pilot 
----c-  M i S k ic b pilot  = --------------------------------------------------
-
Nt DL
N tot  ic b pilot 
If Mi is a 1xEV-DO Rev. 0 service user, determination of the peak throughput from the graph (Peak throughput=f(C/I)) specified
for the mobility type of Mi
E
TP max – DL  M i S k  = f  ----c-  M i S k ic b pilot 
Nt
If Mi is a 1xEV-DO Rev. A service user, selection of the downlink 1xEV-DO radio bearer ( Index DL – Bearer ) for which
E E DL
----c-  M i S k ic b pilot    ----c-  Index DL – Bearer 
Nt  Nt  min
384
If Mi is a 1xEV-DO Rev. B service user, selection of the downlink 1xEV-DO radio bearer ( Index DL – Bearer ) for which
E E DL
----c-  M i S k ic b pilot    ----c-  Index DL – Bearer  and the modulation is supported by Mi’s terminal.
Nt  Nt  min
DL
If Mi is a (1xEV-DO Rev. A - Guaranteed bit rate) service user and TP P – R LC  Index DL – Bearer   TPD min – DL  Service  M i   , Mi is
rejected.
DL
TP P –R LC  Index DL – Bearer 
Determination of the peak throughput: TP max – DL  M i S k ic  = ------------------------------------------------------------
-
n TS
TP max – DL  M i  = TP max – DL  M i S k ic 
For (1xEV-DO Rev. A - Guaranteed bit rate) service users, TP max – DL  M i  = TPD min – DL  Service  M i  

DL
For multi-carrier 1xEV-DO Rev. B service users, TP max – DL  M i  = TP max – DL max  M i S k ic 
 S k ic 
For (1xEV-DO Rev. A - Guaranteed bit rate) service users, calculation of C DL – Bearer
EndFor
DL
Calculation of the average cell throughput, TP av
G MU  N mobiles  S j ic   
 TP max – DL  M i S j ic 
    
 Mi  NVBR –m obiles  Sj ic   
 ----------------------------------------------------------------------------------------
-   1 –  C DL – Bearer  M k S j ic 
 N VBR –m obiles  S j ic   
  M k  N GBR – m obiles  Sj ic  
 
DL
TP av  S j ic  = 
 TPD min – DL  M k 
  
 Mk  NGBR – m obiles  Sj ic  
+  --------------------------------------------------------------------------------------------- 
 N GBR – m obiles j  S  ic    C DL – Bearer  M k S j ic 
  M  N  S  ic 
k GBR – m obiles j
 
N mobiles
 1 –  ER  S  ic      1 – TS
 DRC j  BCMCS  S j ic  – TS EVDO – CCH  S j ic   + TP BCMCS  S j ic   TS BCMCS  S j ic 
If N mobiles  S j ic  = 1 , then G MU = 1
Else if N mobiles  S j ic   1 , G MU is determined from the graph (MUG table=f(nb users)) specified for (Sj,ic). If the transmitter
supports the multi-carrier EV-DO mode, G MU is determined from the graph (MUG table=f(nb users)) specified for Sj.
EndIf
EndFor
5.4.2.2.3 Convergence Criterion

The algorithm convergence is studied on uplink only. The uplink convergence criterion is evaluated at each iteration, and can
be written as follow:
  UL UL
I tot  ic  k – I tot  ic  k – 1   UL UL
 UL = max  int  ma x --------------------------------------------------  100  int  ma x ------------------------------------------------------------  100 
  Stations UL
I  ic    Stations
N
UL
 ic  
tot k user k
1st case: Between two successive iterations,  UL is lower (  ) than the threshold (defined when creating a simulation).

Example: Let us assume that the maximum number of iterations is 100, UL convergence threshold is set to 5. If  UL  5
between the 4th and the 5th iteration, Atoll stops the algorithm after the 5th iteration. Convergence has been achieved.
385
2nd case: After 30 iterations,  UL is still higher than the threshold and from the 30th iteration,  UL does not decrease during
the next 15 successive iterations.
Examples: Let us assume that the maximum number of iterations is 100, UL convergence threshold is set to 5.
1. After the 30th iteration,  UL equals 100 and do not decrease during the next 15 successive iterations: Atoll stops the
algorithm at the 46th iteration. Convergence has not been achieved.
2. After the 30th iteration,  UL equals 80, it starts decreasing slowly until the 40th iteration (without going under the
threshold) and then does not change during the next 15 successive iterations: Atoll stops the algorithm at the 56th iteration
without achieving convergence.

If  UL is still strictly higher than the threshold, the simulation has not converged (specific divergence symbol).
If  UL is lower than the threshold, the simulation has converged.
5.4.3 Appendices
5.4.3.1 Admission Control
During admission control, Atoll calculates the uplink load factor of a considered cell assuming the mobile concerned is
connected with it. Here, activity status assigned to users is not taken into account. So even if the mobile is not active on UL,
it can be rejected due to cell load saturation. To calculate the cell UL load factor, either Atoll takes into account the mobile
power determined during power control if mobile was connected in previous iteration, or it estimates a load rise due to the
UL
mobile and adds it to the current load. The load rise ( X ) is calculated as follows:
UL 1
X = -------------------------------------
-
W
1 + --------------------------- -
UL UL
Q req  TP
In case of CDMA2000 1xRTT networks, we have:

UL UL UL UL FCH SCH
Q req =  Q req  FCH +  Q req  SCH and TP = TP P – UL + TP P – UL
5.4.3.2 Resources Management

5.4.3.2.1 Walsh Code Management
Walsh codes are managed in the downlink during the simulation in case of CDMA2000 1xRTT networks. Atoll performs Walsh
code allocation during the radio resource control step.
Walsh codes form a binary tree with codes of a longer length generated from codes of a shorter length. Length-k Walsh codes
are generated from length-k/2 Walsh codes. Therefore, if a channel needs 1 length-k/2 Walsh code, it is equivalent to using 2
length-k Walsh codes, or 4 length-2k Walsh codes and so on.
Figure 5.3: Walsh Code Tree Indices (Not Walsh Code Numbers)
128 128-bit-length Walsh codes per cell are available in CDMA2000 documents.
386
During the resource control, Atoll determines the number of 128-bit-length Walsh codes that will be consumed by each cell.
Therefore, it allocates :
• A code with the longest length (i.e. a 128 bit-length code) per common channel for each cell. The number of common
channels per cell corresponds to the value defined for the DL overhead resources for common channels per cell
parameter available in the site equipment properties.
• Two 128 bit-length codes per cell-receiver link for FCH in RC1, RC2, RC3 or RC5 and only one for FCH in RC4.
• The number of 128 bit-length codes to be allocated per cell-receiver link for SCH (in case SCH is supported by the user
Walsh codes
radio configuration), N 128 bits , is determined as follows:
Walsh codes SCH

N 128 bits = TPF DL  2 for RC1, RC2, RC3 and RC5,
And
Walsh codes SCH
N 128 bits = TPF DL for RC4.
Where
SCH
TPF DL is the SCH throughput factor.
The Walsh code allocation follows the “Buddy” algorithm, which guarantees that:
• If a k-length Walsh code is used, all of its children with lengths 2k, 4k, …, cannot be used as they are not orthogonal.
• If a k-length Walsh code is used, all of its ancestors with lengths k/2, k/4, …, cannot be used as they are not orthogonal.
• The Walsh code allocation follows the mobile connection order (mobile order in the
Mobiles tab).
• The Walsh code and channel element management is dealt with differently in case of
“softer” handoff. Atoll allocates Walsh codes for each transmitter-receiver link while
it assigns channel elements globally to a site.
5.4.3.2.2 Channel Element Management

Channel elements are controlled in the simulation.
CDMA2000 1xRTT networks
Atoll checks the availability of this resource on uplink and downlink.

On uplink, Atoll consumes N CE – UL  j  channel elements for each cell j on a site NI. This figure includes:
Overhead
• N CE – UL channel elements for control channels (Pilot channel),
FCH SCH
• N CE – UL   1 + TPF UL  per cell-receiver link, for TCH (TCH correspond to Traffic channels i.e. FCH and SCH).
Therefore, the number of channel elements required on uplink at the site level, N CE – UL  N I  , is:
N CE – UL  N I  =  NCE – UL  j 
j  NI
In the downlink, Atoll consumes N CE – DL  j  channel elements for each cell j on a site NI. This figure includes:
Overhead
• N CE – DL channel elements for control channels (Pilot channel, Synchronisation channel, Paging channel),
FCH SCH
• N CE – DL   1 + TPF DL  per cell-receiver link, for TCH (TCH correspond to Traffic channels i.e. FCH and SCH).
Therefore, the number of channel elements required on downlink at the site level, N CE – DL  N I  , is:
N CE – DL  N I  =  NCE – DL  j 
j  NI
In case of “softer” handover (the mobile has several links with co-site cells), Atoll
allocates channel elements for the best serving cell-mobile link only.
387
CDMA2000 1xEV-DO networks
In the uplink, Atoll consumes N CE – UL  j  channel elements for each cell j on a site NI. This figure includes:
• 2 channel elements for control channels (Pilot channel, Data Rate Control channel, etc ). This value is fixed and hard-
coded.
TCH
• N CE – UL per cell-receiver link, for (EV-DO - Variable bit rate) service users.
TCH
• N CE – UL  C UL – Bearer per cell-receiver link, for (EV-DO - Guaranteed bit rate) service users.
Therefore, the number of channel elements required on uplink at the site level, N CE – UL  N I  , is:
N CE – UL  N I  =  NCE – UL  j 
j  NI
In the downlink, only one user can be served by a cell at a time, so this resource is not limited.
5.4.3.3 Downlink Load Factor Calculation

Atoll calculates the downlink load factor for each cell (available in the Cells tab of any given simulation results) and each
connected mobile (available in the Mobiles tab of any given simulation results).
5.4.3.3.1 Downlink Load Factor per Cell

The downlink load factor is calculated for each CDMA2000 1xRTT cell.
Approach for downlink load factor evaluation is highly inspired by the downlink load factor defined in the book “WCDMA for
UMTS by Harry Holma and Antti Toskala”.
DL – FCH DL – SCH
Q req Q req
Let CI req = -------------------- + -------------------- be the required quality.
DL – FCH DL – SCH
Gp Gp
FCH SCH
So, we have CI req = CI req + CI req
In case of soft handoff, required quality is limited to the effective contribution of the transmitter.
 Ptch  ic 
DL ortho
P tx  ic  = P pilot  ic  + P sync  ic  + P paging  ic  + P SCH  ic  + P FCH  ic  = P CCH  ic  +
tch
where
ortho
P CCH  ic  = P pilot  ic  + P sync  ic  + P paging  ic 
 Ptch  ic  = P SCH  ic  + P FCH  ic 

tch
At mobile level, we have a required power, Ptch:

term
P tch  ic  = CI req   I extra  ic  + I intra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0   LT
 I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  

 
P tch  ic  = CI req    P tx  ic  – P tch  ic 
DL  L
  T
 - + N term
+  1 – F ortho   BTS    ----------------------------------------- 0 
  LT  
DL term
 I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic    L T +  1 – F ortho   BTS   P tx  ic  + N 0  L T
P tch  ic  = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
1
----------- +  1 – F ortho   BTS 
CIreq
where
DL
I intra  ic  is the total power received at receiver from the cell to which it is connected.
DL
I extra  ic  is the total power received at receiver from other cells.
I inter – carrier  ic  is the inter-carrier interference received at receiver.
388
I inter – techno log y  ic  is the inter-technology interference received at receiver.
 I  ic  + I inter – carrier  ic  + I inter – techno log y  ic    L T  

  extra 
 +  –    
DL
  +
term


 1 F ortho BTS P tx ic N 0 L T 

DL ortho
P tx  ic  = P CCH  ic  +  --------------------------------------------------------------------------------------------------------------------------------------------------
 1
----------- +  1 – F ortho   BTS  
tch  
CI req
 
 
We have:
 I 
extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic    L T  
  ---------------------------------------------------------------------------------------------------------------------------------------
 DL
P tx  ic  
 
 DL term 
  +  1 – F ortho   BTS   P tx  ic  + N 0  L T 

DL ortho
P tx  ic  = P CCH  ic  +  -------------------------------------------------------------------------------------------------------------------------------------------------
-
 1 - + 1 – F
----------    
tch  ortho BTS 
CI req
 
 
 
 
  ---------------------------------------------------------------------------------------------------------------------------------------
 I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic    L T
+ 1 – F ortho   BTS 

   DL
P tx  ic  
DL  tch DL
P tx  ic  –  ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------  P tx  ic 
 1 
----------- +  1 – F ortho   BTS 
 CI req 
 
 term

 ortho N0  LT 
=  P CCH  ic  +
  ---------------------------------------------------------------
1 
 tch ----------- +  1 – F ortho   BTS 
CI req
 term

 ortho N0  LT 
 P CCH  ic  +
  -
--------------------------------------------------------------
1 
 tch ----------- +  1 – F ortho   BTS 
DL
CI req
P tx  ic  = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
  I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic    L T
--------------------------------------------------------------------------------------------------------------------------------------- + 1 – F ortho   BTS
 DL
P tx  ic  
1 – -
 
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1
----------- +  1 – F ortho   BTS 

 tch 
 CI req 
Therefore, the downlink load factor can be expressed as:
 ---------------------------------------------------------------------------------------------------------------------------------------
I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic    L T
+ 1 – F ortho   BTS
 P
DL
 ic  
 --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

DL tx
X =
 1 - + 1 – F 
tch  ---------- ortho   BTS  
 CI req 
The downlink load factor represents the signal degradation in relative to the reference interference (thermal noise).
5.4.3.3.2 Downlink Load Factor per Mobile

Atoll evaluates the downlink load factor for any connected mobile (CDMA2000 1xRTT 1xEV-DO user) as follows,
DL
DL I tot  ic 
X = ------------------
-
DL
N tot  ic 
5.4.3.4 Best Server Determination in Monte Carlo Simulations - Old Method

Before Atoll 2.8.0, best server determination used to be performed by selecting the best carrier within transmitters according
to the selected method (site equipment) and then the best transmitter using the best carrier. To switch back to this method,
add the following lines in the Atoll.ini file:
[CDMA]
389
MultiBandSimu = 0
The method is described below:

For each station Sj containing Mi in its calculation area and using the main frequency band supported by the Mi’s terminal (i.e.
either f1 for a single frequency band network, or f1, f2 or f3 for a multi-band terminal without any priority on frequency bands,
or f1 for a multi-band terminal with f1 as main frequency band).
Determination of BestCarrier k  S j M i  .
If a given carrier is specified for the service requested by Mi and if it is used by Sj
BestCarrier k  S j M i  is the carrier specified for the service.
Else the carrier selection mode defined for Sj is considered.

For each carrier ic used by Sj, we calculate current loading factor:
UL
UL I tot  S j ic  UL
X k  S j ic  = -------------------------
- + X
UL
EndFor
UL
BestCarrier k  S j M i  is the carrier with the lowest X k  S j ic 

BestCarrier k  S j M i  is the carrier with the lowest P tx  S j ic  k

BestCarrier k  S j M i  is randomly selected

UL UL
BestCarrier k  S j M i  is the first carrier so that X k  S j ic   X max
   BTS  P c  M i S j BestCarrier 
Calculation of Q pilot  M i S j BestCarrier  = -----------------------------------------------------------------------------------
-
DL
k
I 0  BestCarrier k  S j M i  

pilot
If Q pilot  M i S j BestCarrier   Q req then Sj is rejected by Mi
k
max
If Q pilot  M i S j BestCarrier   Q pilot  M i 
k k
Admission control (If simulation respects a load factor constraint and Mi was not connected in previous iteration).
UL UL
If X k  S j BestCarrier k  S j M i    X max , then Sj is rejected by Mi
Else
max
Q pilot  M i  = Q pilot  M i S j BestCarrier 
k k
S BS  M i  = S j
Endif
EndFor
If no SBS has been selected and Mi’s terminal can work on one frequency band only, Mi has failed to be connected to the
network and is rejected.
If no SBS has been selected and Mi’s terminal can work on another frequency band.
Determination of BestCarrier k  Sj M i  for each station txj containing Mi in its calculation area and using another frequency
band supported by the Mi’s terminal (i.e. f1, f2 or f3 for a multi-band terminal without any priority on frequency bands, or f2
for a multi-band terminal with f2 as secondary frequency band)
390

UL
- + X
X k  S j ic  = -------------------------
UL
EndFor
UL



UL UL
-
DL
k

pilot
k
max
k k
UL UL
Else
max
k k
S BS  M i  = S j
Endif
EndFor
If no SBS has been selected and Mi’s terminal can work on two frequency bands only, Mi has failed to be connected to the
network and is rejected.
If no SBS has been selected and Mi’s terminal can work on another frequency band.
Determination of BestCarrier k  Sj M i  for each station txj containing Mi in its calculation area and using another frequency
band supported by the Mi’s terminal (i.e. f1, f2 or f3 for a multi-band terminal without any priority on frequency bands, or f3
for a multi-band terminal with f3 as third frequency band)

UL
X k  S j ic  = -------------------------
- + X
UL
391
EndFor
UL



UL UL
-
DL
k

pilot
k
max
k k
UL UL
Else
max
k k
S BS  M i  = S j
Endif
EndFor
If no SBS has been selected, Mi has failed to be connected to the network and is rejected.
5.4.3.5 Radio Bearer Allocation Algorithm for Multi-carrier EVDO Rev.B - Old
Method
Before Atoll 3.2.1, radio bearer allocation for multi-carrier EVDO Rev.B used to be performed by equally sharing the available
terminal power between the carriers.
To switch back to this method, add the following lines in the Atoll.ini file:
[CDMA]
SharingEquallyPower = 1
UsingPreviousIterationPowerWeight = 1
5.5 CDMA2000 Prediction Studies

5.5.1 Point Analysis: The AS Analysis Tab
Let us assume a receiver with a terminal, a mobility type and a service with certain UL and DL throughputs. This receiver does
not create any interference. You can make the prediction for a specific carrier or for the best 1xRTT or 1xEV-DO carrier. The
type of carrier and the carriers you can select depend on the service and on the frequency band(s) supported by the terminal.
The analysis is based on the uplink load percentage and the downlink total power of cells. These parameters can be either
outputs of a given simulation, average values calculated from a group of simulations, or user-defined cell inputs.
Results are displayed for any point of the map where the pilot signal level exceeds the defined minimum RSCP.
392
5.5.1.1 Bar Graph and Pilot Sub-Menu

We can consider the following cases:

The carrier that can be used by transmitters is fixed. In this case, for each transmitter i containing the receiver in its calculation
area and using the selected carrier, Atoll calculates the pilot quality at the receiver on this carrier. Then, it determines the best
serving transmitter using the selected carrier ic.
2nd case: Analysis based on the best carrier of all frequency bands
Atoll determines the best carrier for each transmitter i which contains the receiver in its calculation area and uses a frequency
band supported by the receiver’s terminal. The best carrier selection depends on the option selected for the site equipment
(UL minimum noise, DL minimum power, random, sequential). Then, Atoll calculates the pilot quality at the receiver from
these transmitters on their best carriers (ic) and defines the best server (on its best carrier).
3rd case: Analysis based on the best carrier of any frequency band (for multi-band terminals with priority defined on frequency
bands only)
The frequency band that can be used is fixed. Atoll determines the best carrier for each transmitter i containing the receiver
in its calculation area and using the selected frequency band. The best carrier selection depends on the option selected for
the site equipment (UL minimum noise, DL minimum power, random, sequential). Then, Atoll calculates the pilot quality at
the receiver from these transmitters on their best carriers (ic) and defines the best server (on its best carrier).
Atoll provides the same outputs in the bar graph and pilot sub-menu whichever the studied network, CDMA2000 1xRTT or
1xEV-DO.
• Ec/I0 (or Q pilot  ic  ) evaluation
We assume that ic is the best carrier of a transmitter i containing the receiver in its calculation radius.
For CDMA2000 1xRTT users we have,
 BTS    P c  i ic 
Q pilot  i ic  = ---------------------------------------------
DL
I 0  ic 
DL DL DL DL DL term
with I 0  ic  = P tot  i ic  + I extra  ic  + I inter – carrier  ic  + I inter – techno log y  ic  + N 0
For CDMA2000 1xEV-DO users, we have,

DL
 BTS    P tot  i ic b pilot 
Q pilot  i ic  = ----------------------------------------------------------------
-
DL
I 0  ic b pilot 
DL DL DL DL DL term
With I 0  ic b pilot  = P tot  i ic b pilot  + I extra  ic b pilot  + I inter – carrier  ic b pilot  + I inter – techno log y  ic  + N 0
The calculation of Q pilot  i ic  can be divided into 6 steps explained in the table below.
CDMA2000 1xRTT users CDMA2000 1xEV-DO users

DL
P tot  i ic b pilot  calculation for each cell (i,ic)
P c  i ic  calculation for each cell (i,ic) DL
P tot  i ic b pilot  is the pilot burst from the transmitter i on the
P c  i ic  is the pilot power from a transmitter i on the carrier ic at carrier ic at the receiver.
the receiver.
P tx  i ic b pilot 
1st step
DL
P pilot  i ic  P tot  i ic b pilot  = -----------------------------------
-
P c  i ic  = -------------------------
- LT
I
LT
I and
P tx  i ic b pilot  = P max  i ic 

L T is the total loss between the transmitter i and the receiver: -------------------------------------------------------------------------------------------------------------------------------------
I G Tx  G term
393

DL DL DL
I extra  ic  , I inter – carrier  ic  and I inter – techno log y  ic  calculation
We have,
 Ptot  j ic 
DL DL
I extra  ic  = DL DL
I extra  ic b pilot  and I inter – carrier  ic b pilot  calculation
j j  i
DL
For each transmitter of the network, P tot  j ic  is the total power We have,
received at the receiver from the transmitter j on the best carrier ic
of the transmitter i.
 Ptot  j ic bpilot 
DL DL
I extra  ic b pilot  =
DL P tx  j ic 
P tot  j ic  = -------------------- j j  i
2nd step
LT
 Ptot  j icadj bpilot 

DL
P tx  j ic  is the total power transmitted by the transmitter j on the
DL j j
best carrier of the transmitter i. I inter – carrier  ic b pilot  = ----------------------------------------------------
Finally, we have, and
Tx
DL

DL
Tx Tx m
DL  j
I inter – carrier  ic  = j------------------------------------
- ni
i
and
Tx

DL
Tx Tx m
n
i i
term
3rd step
N0 calculation
Tx DL
NF Term  K  T  W  NR inter – techno log y
4th step
DL
I 0  ic  and Q pilot  i ic  evaluation based on formulas defined above
DL
G macro – diversity calculation
DL
The macro-diversity gain, G macro – diversity , models the decrease in shadowing margin due to the fact there are several pilot signals at the
mobile.
5th step
DL npaths
G macro – diversity = M Shadowing – Ec  Io – M Shadowing –Ec  Io
npaths
M Shadowing – Ec  Io is the shadowing margin for the mobile receiving n pilot signals (not necessarily from transmitters belonging to the
mobile active set).
Note: This parameter is determined from the fixed cell edge coverage probability and the model standard deviation. When the model
standard deviation is set to 0, the macro-diversity gain equals 0.
394

Determination of active set
Atoll takes the transmitter i with the highest Q pilot  i ic  and calculates the best pilot quality received with a fixed cell edge coverage
Resulting
probability Q pilot  ic  .
Resulting DL
Q pilot  ic  = G macro – diversity  max  Q pilot  i ic  
Resulting req Resulting
Q pilot  Q pilot means that the pilot quality at the receiver exceeds Q pilot  ic  x% of times (x is the fixed cell edge coverage
probability). The cell with the highest Q pilot  i ic  enters the active set as best server ( Q pilot  BS ic  ) and the best carrier (icBS) of the
best server
6th step
BS will be the one used by other transmitters of active set (when active set size is greater than 1). Pilot is available.
Resulting req
If Q pilot  ic   Q pilot , no cell (i,ic) can enter the active set. Pilot is unavailable.
Then, pilot qualities at the receiver from transmitters i (other than the best server) on the best carrier of the best server, icBS, are
recalculated to determine the entire receiver active set (when active set is greater than 1). Same formulas and calculation method are
used to update
DL
I 0  ic BS  and determine Q pilot  i ic BS  .
Other cells (i,icBS) in active set must fulfill the following criteria:
pilot
Q pilot  i ic BS   Q min
 i ic BS   neighbour list  BS ic BS  (optional)
For multi-carrier 1xEV-DO Rev.B service users, these results are detailed for each sub-active set. For each carrier, Atoll displays
the thermal noise, I0 (Best server), the pilot quality from the best server and from the other servers of the sub-active set, and
the downlink macro-diversity gain. They are calculated as described above.
• Number of cells in active set
This is a user-defined input in the terminal properties. It corresponds to the active set size.
• Number of fingers
The number of fingers, f, of the rake receiver. This parameter is defined in the terminal properties. It is relevant in CDMA2000
1xRTT only11. This is the maximum number of active set links that the terminal (rake) can combine.
• Thermal noise
This parameter is calculated as described above (3rd step).

• I0 (Best server)
I0 (Best server) is the total noise received at the receiver on icBS.
• Downlink macro-diversity gain
This parameter is calculated as described above (5th step).
5.5.1.2 Downlink Sub-Menu

Outputs calculated by Atoll depend on the studied network (CDMA2000 1xRTT or CDMA2000 1xEV-DO).
5.5.1.2.1 CDMA2000 1xRTT

Let mFCH and mSCH respectively denote the number of cells in the receiver active set for the fundamental channel (FCH) and
the supplemental channel (SCH) and f be the number of rake fingers defined for the terminal. We assume that f is less than
or equal to mFCH and mSCH.
Among the mFCH cells of the receiver active set, only the first f cells will be considered in order to determine the FCH
availability on downlink. In the same way, only the first f cells among the mSCH cells of the receiver active set will be considered
in order to determine the SCH availability on downlink. Each of these cells is noted (k,icBS).
Atoll calculates the traffic channel quality on FCH from each cell (k,icBS). No power control is performed as in simulations.
Here, Atoll determines the downlink traffic channel quality on FCH at the receiver for the maximum traffic channel power per
transmitter allowed on FCH. Then, after combination, the total downlink traffic channel quality on FCH is evaluated and
compared with the specified target quality.
11. CDMA2000 1xEV-DO systems do not support soft handover on downlink.
395
Atoll calculates the traffic channel quality on SCH from each cell (k,icBS). No power control is performed as in simulations.
Here, Atoll determines the downlink traffic channel quality on SCH at the receiver for the maximum traffic channel power per
transmitter allowed on SCH. This value depends on the downlink throughput specified in the analysis. Then, after
combination, the total downlink traffic channel quality on SCH is evaluated and compared with the specified target quality.
• Eb/Nt target on FCH and Eb/Nt target on SCH
DL
Eb/Nt target on FCH (  Q req  FCH ) is the downlink traffic data quality target on the fundamental channel (FCH). This value is
user-defined for a given service and terminal.
DL
Eb/Nt target on SCH (  Q req  SCH ) is the downlink traffic data quality target on the supplemental channel (SCH). This value is
specified for a given service, terminal and SCH throughput.
• Required transmitter powers on FCH and SCH
req req
The calculation of the required transmitter powers on FCH and SCH ( P FCH and P SCH ) may be divided into three steps.
1st step: Eb/Nt max for the first f (number of fingers) cells of active set
DL DL
Let us assume the following notations: Eb/Nt max on FCH and SCH respectively correspond to  Q max  FCH and  Q max  SCH .

DL – FCH
DL  BTS  P b –max  k ic BS  – FCH
 Q max  k ic BS   FCH -  G DL
= -------------------------------------------------------- p
DL
N tot  ic BS 
And
DL – SCH
DL  BTS  P b –max  k ic BS  – SCH
-  G DL
 Q max  k ic BS   SCH = -------------------------------------------------------- p
DL
N tot  ic BS 
max max
DL – FCH P FCH DL – SCH P SCH
With P b  k ic BS  = ----------
- , P b –max  k ic BS  = ----------
-
LT LT
k k
DL DL DL DL DL term
And N tot  ic BS  = I intra  ic BS  + I extra  ic BS  + I inter – carrier  ic BS  + I inter – techno log y  ic BS  + N 0
Where
max
P FCH is the maximum power allowed on FCH. This parameter is user-defined in the Services table for a certain terminal.
max
P SCH is the maximum power allowed on SCH for the specified downlink throughput. This parameter is user-defined in the
Services table for a certain terminal and SCH throughput.
L T is the total loss between the transmitter i and the receiver.
k

DL
L T = ----------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
DL
N tot  ic BS  is the total noise at the receiver on the best carrier of the best server.
With
DL
I intra  ic BS  =  1 –  BTS  F ortho   P DL  k ic 
tot BS
And
DL DL
j j  k
DL
For each transmitter in the network, P tot  ic BS  is the total power received at the receiver from this transmitter on icBS.
DL
I inter – carrier  ic BS  is the inter-carrier interference at the receiver on the best carrier of the best server.
396
DL
DL  j
----------------------------------------
-
DL
I inter – techno log y  ic BS  is the inter-technology interference at the receiver on the best carrier of the best server.
Tx

DL
Tx Tx m
ni
i BS
th
Tx m
i BS
2nd step: Calculation of the total traffic channel quality on FCH and SCH
DL
 Q MAX  FCH is the traffic channel quality on FCH at the receiver on icBS after combining the signal from each cell (k,icBS).

DL DL
 Q MAX  ic BS  FCH =  Q max  k ic BS  FCH
  Qmax  k icBS  FCH

DL DL DL
 Q MAX  ic BS  FCH = f rake efficiency 
k
Where
DL
DL
 Q MAX  SCH is the traffic channel quality on SCH at the receiver on icBS after combining the signal from each cell (k,icBS).

DL DL
 Q MAX  ic BS  SCH =  Q max  k ic BS   SCH
  Qmax  k icBS  SCH

DL DL DL
 Q MAX  ic BS  SCH = f rake efficiency 
k
req req
3rd step: P FCH and P SCH calculation
DL
req  Q req  FCH
-  P max
P FCH = --------------------------------------- FCH
DL
 Q MAX  ic BS   FCH
DL
req  Q req  SCH
-  P max
P SCH = --------------------------------------- SCH
DL
 Q MAX  ic BS   SCH
• Eb/Nt max on FCH for the first f (number of fingers) cells of active set
DL
Let us assume the following notation: Eb/Nt max on FCH corresponds to  Q max  FCH .

DL – FCH
DL  BTS  P b –max  k ic BS  – FCH
-  G DL
 Q max  k ic BS   FCH = -------------------------------------------------------- p
DL
N tot  ic BS 
397
max
DL – FCH P FCH DL DL DL DL term
With P b –max  k ic BS  = ----------
- and N tot  ic BS  = I intra  ic BS  + I extra  ic BS  + I inter – carrier  ic BS  + N 0
LT
k
Where
max
P FCH is the maximum power allowed on FCH. This parameter is user-defined in the Services table for a certain terminal.

k

DL
L T = ----------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
DL
With
max req
DL
I intra  ic BS  =  1 –  BTS  F ortho   P DL P FCH – P FCH
tot  k ic BS  –  1 –  BTS   max (----------------------------,0)
LT
k
And
DL DL
j j  k
DL
For each transmitter in the network, P tot  ic BS  is the total power received at the receiver from the transmitter on icBS.
DL
DL
DL txj j
I inter – carrier  ic BS  = ----------------------------------------
-
DL
Tx

DL
Tx Tx m
ni
i BS
th
Tx m
i BS
• Eb/Nt max on SCH for the first f (number of fingers) cells of active set
DL
Let us assume the following notation: Eb/Nt max on SCH corresponds to  Q max  SCH .

DL – SCH
DL  BTS  P b –max  k ic BS  – SCH
-  G DL
 Q max  k ic BS   SCH = -------------------------------------------------------- p
DL
N tot  ic BS 
max
DL – SCH P SCH
With P b –max  k ic BS  = ----------
-
LT
k
DL DL DL DL DL term
and N tot  ic BS  = I intra  ic BS  + I extra  ic BS  + I inter – carrier  ic BS  + I inter – techno log y  ic BS  + N 0
Where
398
max
P SCH is the maximum power allowed on SCH for the specified downlink throughput. This parameter is user-defined in the
Services table for a certain terminal and SCH throughput.
k

DL
L T = ----------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
DL
With
max req
DL
I intra  ic BS  =  1 –  BTS  F ortho   P DL P SCH – P SCH
tot  k ic BS  –  1 –  BTS   max (----------------------------,0)
LT
k
And
DL DL
j j  k
DL
For each transmitter in the network, P tot  ic BS  is the total power received at the receiver from the transmitter on icBS.
DL
DL
DL  j
----------------------------------------
-
DL
Tx

DL
Tx Tx m
ni
i BS
th
Tx m
i BS
• Eb/Nt max on FCH and Eb/Nt max on SCH

DL
 Q MAX  FCH is the traffic channel quality on FCH at the receiver on icBS after combining the signal from each cell (k,icBS).

DL DL
 Q MAX  ic BS  FCH =  Q max  k ic BS  FCH
  Qmax  k icBS  FCH

DL DL DL
 Q MAX  ic BS  FCH = f rake efficiency 
k
Where
DL
DL
 Q MAX  SCH is the traffic channel quality on SCH at the receiver on icBS after combining the signal from each cell (k,icBS).
399
DL DL
 Q MAX  ic BS   SCH =  Q max  k ic BS   SCH
  Qmax  k icBS  SCH

DL DL DL
 Q MAX  ic BS   SCH = f rake efficiency 
k
DL DL
Therefore, the service on the downlink traffic channel is available if  Q MAX  ic BS   FCH   Q req  FCH and
DL DL
 Q MAX  ic BS   SCH   Q req  SCH .
• Effective Eb/Nt on FCH and Eb/Nt on SCH

DL DL
 Q eff  FCH and  Q eff  SCH are respectively effective traffic channel qualities at the receiver on icBS supplied on FCH and SCH.
DL DL DL
 Q eff  FCH = min   Q MAX  FCH  Q req  FCH 
And
DL DL DL
 Q eff  SCH = min   Q MAX  SCH  Q req  SCH 
• Downlink soft handover gain on FCH and downlink soft handover gain on SCH
DL DL
 G SHO  FCH and  G SHO  SCH respectively correspond to DL soft handover gains on FCH and SCH.
DL
DL  Q MAX  ic BS  FCH
 G SHO  FCH = --------------------------------------------------------------
-
DL
max   Q max  k ic BS   FCH 
k
And
DL
DL  Q MAX  ic BS   SCH
 G SHO  SCH = --------------------------------------------------------------
-
DL
max   Q max  k ic BS   SCH 
k
DL DL
k
5.5.1.2.2 CDMA2000 1xEV-DO

Atoll calculates the effective pilot quality level at the receiver and compares this value with the required quality level.
1xEV-DO Rev.0 and 1xEV-DO Rev. A Service Users
For 1xEV-DO Rev.0 and 1xEV-DO Rev. A users, Atoll displays the following results:
• Required throughput
DL
The required throughput, TP req , is the downlink throughput selected for the analysis.
• Required C/I
For 1xEV-DO Rev. 0 users, the required C/I (  --- ) is determined from the graph “Peak throughput=f(C/I)” defined for the
C
I req
mobility type selected in the analysis. It corresponds to the value read in the graph “Peak throughput=f(C/I) (Rev0)” for the
DL
specified required throughput, TP req .
DL
For 1xEV-DO Rev. A users, the required throughput ( TP req ) is obtained by using a certain downlink transmission format (i.e.
a 1xEV-DO radio bearer ( Index DL – Bearer ) with a certain number of timeslots ( n TS )). It is calculated as follows:
DL
DL TP P – R LC  Index DL – Bearer 
TP req = ------------------------------------------------------------
-
n TS
400
C
--- is the value defined in the 1xEV-DO Radio Bearer Selection (Downlink) table for this downlink transmission format
 I  req
(radio bearer Index, mobility and number of timeslots). It corresponds to the C/I required to obtain the defined required
DL
throughput, TP req .
• Effective C/I
Ec
Let -----  ic BS b pilot  be the effective C/I at the receiver on icBS.
Nt
For the best cell (BS,icBS) of the receiver active set, we have:
 
E  1 
----c-  ic BS b pilo t  =  ------------------------------
-
Nt 1
 ---------------------- – 1
 Q pilot 
resulting
Where
pilot DL
Q resulting  ic BS  = G macro – diversity  Q pilot  ic BS 
BS
• Obtained throughput
DL
For 1xEV-DO Rev. 0 users, the obtained throughput, TP , is determined from the graph “Peak throughput=f(C/I) (Rev0)”
DL
defined for the mobility type selected in the analysis. TP is the value read in the graph “Peak throughput=f(C/I) (Rev0)” for
E
the calculated effective C/I, ----c-  ic BS b pilot  .
Nt
DL
For 1xEV-DO Rev. A users, the obtained throughput ( TP ) on downlink depends on the downlink transmission format, i.e the
radio bearer index ( Index DL – Bearer ) with the number of timeslots ( n TS ). For the defined mobility type, Atoll selects the
E
downlink transmission format where ----c-  ic BS b pilot    --- . Then, it determines the downlink obtained throughput as
C
Nt  I  req
follows:
DL
DL TP P –R LC  Index DL – Bearer 
TP = ------------------------------------------------------------
-
n TS
DL DL
The traffic data channel in downlink is available if TP  TP req .
• Bearer Consumption
For (1xEV-DO Rev. A - Guaranteed bit rate) service users, Atoll calculates the 1xEV-DO bearer consumption.
TPD min – DL
C DL – Bearer = ------------------------------------------------------------
-
DL
TPP –R LC  Index DL – Bearer 
Where TPD min – DL corresponds to the minimum bit rate required by the service in the downlink.
1xEV-DO Rev. B Service Users
For single-carrier and multi-carrier 1xEV-DO Rev. B users, Atoll displays the following results:
DL
The required throughput, TP req , is the downlink throughput selected for the analysis.
The obtained throughput corresponds to the sum of the obtained throughputs on each carrier.
 TP
DL DL
TP =  ic 
ic
DL DL
The traffic data channel on downlink is available if TP  TP req .
• For each sub-active set, Atoll indicates the effective C/I and the obtained throughput:
401
E
Let ----c-  ic b pilot  be the effective C/I at the receiver on ic, the carrier associated with the sub-active set.
Nt
For the best cell (BS,ic) of the receiver sub-active set, we have:
pilot
E   Q resulting  ic 
----c-  ic b pilot  = -----------------------------------------
Nt pilot
 – Q resulting  ic 
Where
pilot DL
Q resulting  ic  = G macro – diversity  Q pilot  ic 
BS
DL
The obtained throughput ( TP  ic  ) on downlink depends on the downlink transmission format, i.e the radio bearer index
( Index DL – Bearer ) with the number of timeslots ( n TS ). For the defined mobility type, Atoll selects the downlink transmission
Ec
format where -----  ic b pilot    ---
C
and whose modulation scheme is supported by the terminal.
Nt I req
C
---
 I  req is the value defined in the 1xEV-DO Radio Bearer Selection (Downlink) table for this downlink transmission format
(radio bearer Index, mobility and number of timeslots). It corresponds to the C/I required to obtain the defined required
DL
throughput, R req .
The downlink obtained throughput is determined as follows:

DL
DL TP P – RLC  Index DL – Bearer 
TP  ic  = --------------------------------------------------------------
n TS
5.5.1.3 Uplink Sub-Menu

Outputs calculated by Atoll depend on the studied network (CDMA2000 1xRTT or CDMA2000 1xEV-DO).
5.5.1.3.1 CDMA2000 1xRTT

For each cell (i,icBS) in the receiver active set, Atoll calculates the uplink traffic channel quality on FCH and SCH from the
receiver. No power control is performed as in simulations. Here, Atoll determines the uplink traffic channel quality on FCH at
the cell for the maximum terminal power allowed on FCH. In the same way, it evaluates the uplink traffic channel quality on
SCH at the cell for the maximum terminal power allowed on SCH. Then, total uplink traffic channel qualities on FCH and SCH
are evaluated with respect to the receiver handover status. From these values, Atoll deduces required terminal powers on
FCH and SCH, calculates the total terminal power required and compares this value with the maximum terminal power
allowed.
• Max terminal power on FCH and SCH
max
The Max terminal power parameter ( P term ) is user-defined for each terminal. It corresponds to the maximum terminal power
allowed. On uplink, the terminal power is shared between pilot, FCH and SCH channels. So, we may write:
max max max max
P term =  P term  pilot +  P term  FCH +  P term  SCH
We have:
max max
 P term  pilot = p  P term
Where p is the percentage of the terminal power dedicated to pilot. This parameter is user-defined in the terminal properties.
And
max UL FCH UL
 P term  FCH  Q req  FCH TP P – UL  AF FCH
-  --------------------------------------
------------------------- = ----------------------
max UL SCH
 P term  SCH  Q req  SCH TP P – UL
Therefore,
max
max  1 – p   P term
 P term  FCH = -----------------------------------------------------------------------------
-
UL SCH
 Q req  SCH  TP P – UL
1 + --------------------------------------------------------------------
UL FCH UL
 Q req  FCH  TP P – UL  AF FCH
And
402
max
max  1 – p   P term
 P term  SCH = -----------------------------------------------------------------------------
-
UL FCH UL
 Q req  FCH  TP P – UL  AF FCH
1 + --------------------------------------------------------------------
UL SCH
 Q req  SCH  TP P – UL
• Required terminal power on FCH and SCH

req req
The required terminal powers on FCH and SCH, respectively  P term  FCH and  P term  SCH , are calculated as follows:
UL UL
1st step: Evaluation of uplink traffic channel qualities on FCH and SCH,  Q max  ic BS   and  Q max  ic BS   , for each cell
i FCH i SCH
of active set.
For each cell (i,icBS), we have:
UL – FCH
UL  term  P b –max  i ic BS  – FCH
-  G UL
 Q max  i ic BS   FCH = --------------------------------------------------------- p
UL
N tot  i ic BS 
And
UL – SCH
UL  term  P b –max  i ic BS  – SCH
 Q max  i ic BS   SCH -  G UL
= --------------------------------------------------------- p
UL
max max
UL – FCH  P term  FCH UL – SCH  P term  SCH
With P b –max  i ic BS  = ------------------------
- and P b –max  i ic BS  = ------------------------
-
LT LT
i i

i

UL
L T = ----------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
UL
N tot  i ic BS  is the total noise at the transmitter on the best carrier of the best server. This value is deduced from the cell
UL
uplink load factor X  i ic BS  .
tx
UL N0
N tot  i ic BS  = -----------------------------------
-
UL
1 – X  i ic BS 
tx
UL UL
2nd step: Calculation of FCH and SCH total traffic channel qualities at the transmitter on icBS,  Q MAX  FCH and  Q max  SCH ,
based on the receiver handover status.
If there is no handoff, we have:
UL UL UL UL
 Q MAX  ic BS  FCH =  Q max  i ic BS   FCH and  Q MAX  ic BS   SCH =  Q max  i ic BS   SCH
For soft handover, we have:

UL UL UL
 Q MAX  ic BS  FCH =  G macro – diversity  2 links  max   Q max  i ic BS   FCH 
i
And
UL UL UL
 Q MAX  ic BS  SCH =  G macro – diversity  2 links  max   Q max  i ic BS   SCH 
i
UL
UL UL
max  Q max  i ic BS   corresponds to the highest Q max  i ic BS  value.
i
For soft-soft handover, we have:

UL UL UL
i
403
And
UL UL UL
 Q MAX  ic BS   SCH =  G macro – diversity  3 links  max   Q max  i ic BS   SCH 
i
UL
 G macro – diversity  3 links is the uplink macro-diversity gain.This parameter is determined from the fixed cell edge coverage
For softer and softer-softer handovers, we have:
  Qmax  i icBS  FCH

UL UL UL
 Q MAX  ic BS   FCH = f rake efficiency 
i
  Qmax  i icBS  SCH

UL UL UL
And  Q MAX  ic BS   SCH = f rake efficiency 
i
For softer-soft handover, there are two possibilities. If the MRC option is selected (option available in Global parameters), we
have:
 UL 

UL UL UL UL
 Q MAX  ic BS   FCH =  G macro – diversity  2 links  max  f rake efficiency   Q max  i ic BS   FCH  Q max  i ic BS   
 i on the other site FCH
i on the same site
And
 UL 

UL UL UL UL
 Q MAX  ic BS   SCH =  G macro – diversity  2 links  max  f rake efficiency   Q max  i ic BS   SCH  Q max  i ic BS   
 i on the other site SCH
i on the same site
otherwise,
UL UL UL
 Q MAX  ic BS   FCH =  G macro – diversity  2 links  max   Q max  i ic BS   FCH 
i
And
UL UL UL
i
req req
3rd step: Calculation of  P term  FCH and  P term  SCH
UL UL
req  Q req  FCH  Q req  SCH
-   P max
 P term  FCH = ---------------------------------------
req
-   P max
term  FCH and  P term  SCH = --------------------------------------- term  SCH
UL UL
 Q MAX  ic BS   FCH  Q MAX  ic BS   SCH
Where
UL
 Q req  FCH is the user-defined uplink data traffic quality target on FCH for a given service and a terminal. This parameter is
available in the Services table.
UL
 Q req  SCH is the user-defined uplink data traffic quality target on SCH for a given service, terminal and SCH throughput. This
parameter is available in the Services table.
req
Then, from the required terminal power on FCH and SCH, Atoll determines the total terminal power required ( P term ).
req req req req

P term =  P term  FCH +  P term  SCH +  P term  pilot
req req
As  P term  pilot = p  P term , we have:
req req
req  P term  FCH +  P term  SCH
P term = --------------------------------------------------------
-
1–p
req max
Therefore, the service on the uplink data traffic channel is available if P term  P term .
• Eb/Nt max on FCH for each cell in active set

UL – FCH
UL  term  P b –max  i ic BS  – FCH
-  G UL
 Q max  i ic BS   FCH = --------------------------------------------------------- p
UL
404
max
UL – FCH  P term  FCH
With P b –max  i ic BS  = ------------------------
-
LT
i

i

UL
L T = ----------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
UL
UL
tx max req
UL N0 P FCH – P FCH
- +  1 –  term   max (---------------------------
N tot  i ic BS  = ----------------------------------- -,0)
UL LT
1 – X  i ic BS  i
tx
• Eb/Nt max on SCH for each cell in active set

UL – SCH
UL  term  P b –max  i ic BS  – SCH
-  G UL
 Q max  i ic BS   SCH = --------------------------------------------------------- p
UL
max
UL – SCH  P term  SCH
With P b –max  i ic BS  = ------------------------
-
LT
i

i

UL
L T = ----------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
UL
UL
tx max req
UL N0 P SCH – P SCH
N tot  i ic BS  = -----------------------------------
- +  1 –  term   max (---------------------------
-,0)
UL
1 – X  i ic BS  LT
i
tx
• Eb/Nt max on FCH and SCH

UL UL
 Q MAX  ic BS  FCH and  Q MAX  ic BS   SCH are respectively the traffic channel qualities on FCH and SCH at the transmitter on icBS
after signal combination of all the transmitters of the active set.
UL UL UL UL
 Q MAX  ic BS  FCH =  Q max  i ic BS   FCH and  Q MAX  ic BS   SCH =  Q max  i ic BS   SCH

UL UL UL
i
And
UL UL UL
 Q MAX  ic BS  SCH =  G macro – diversity  2 links  max   Q max  i ic BS   SCH 
i
UL
UL UL
i
405

UL UL UL
i
And
UL UL UL
i
UL
  Qmax  i icBS  FCH

UL UL UL
 Q MAX  ic BS   FCH = f rake efficiency 
i
  Qmax  i icBS  SCH

UL UL UL
And  Q MAX  ic BS   SCH = f rake efficiency 
i
have:
 UL 

UL UL UL UL
 Q MAX  ic BS   FCH =  G macro – diversity  2 links  max  f rake efficiency   Q max  i ic BS   FCH  Q max  i ic BS   
 i on the other site FCH
i on the same site
And
 UL 

UL UL UL UL
 Q MAX  ic BS   SCH =  G macro – diversity  2 links  max  f rake efficiency   Q max  i ic BS   SCH  Q max  i ic BS   
 i on the other site SCH
i on the same site
otherwise,
UL UL UL
i
And
UL UL UL
i
• Effective Eb/Nt on FCH and SCH

UL
 Q eff  FCH is the uplink effective traffic channel quality on FCH at the receiver on icBS.
UL
 Q eff  SCH is the uplink effective traffic channel quality on SCH at the receiver on icBS.
UL UL UL UL UL UL
 Q eff  FCH = min   Q MAX  FCH  Q req  FCH  and  Q eff  SCH = min   Q MAX  SCH  Q req  SCH 
• Uplink soft handover gain FCH and SCH

UL
 G SHO  FCH corresponds to the UL soft handover gain on FCH.
UL
 G SHO  SCH corresponds to the UL soft handover gain on SCH.
UL UL
UL  Q MAX  ic BS   FCH UL  Q MAX  ic BS   SCH
 G SHO  FCH = ------------------------------------------------------------
- and  G SHO  SCH = ------------------------------------------------------------
-
UL UL
max   Q max  i ic BS  FCH  max   Q max  i ic BS   SCH 
I I
UL UL
I
5.5.1.3.2 CDMA2000 1xEV-DO

1xEV-DO Rev.0 and 1xEV-DO Rev. A Service Users
For each cell (l,icBS) in the receiver active set, Atoll calculates the uplink quality level from the receiver. No power control is
performed as in simulations. Here, Atoll determines the uplink quality level at the cell for the maximum terminal power
406
allowed. Then, the total uplink quality level is evaluated with respect to the receiver handover status. From this value, Atoll
calculates the required terminal power and compares it with the maximum terminal power allowed.
• Max terminal power
max
allowed.
• Required terminal power with ACK
req
The required terminal power ( P term ) calculation may be divided into four steps:
UL
1st step: Evaluation of the uplink quality, Q max  i ic BS  , for each cell of active set

UL
UL  term  P b – max  i ic BS 
-  G UL
Q max  i ic BS  = ----------------------------------------------------- p
UL
max
UL P term
With P b –max  i ic BS  = ------------
LT
i

i

UL
L T = ----------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
UL
UL
tx max req
UL N0 P term – P term
- +  1 –  term   max (------------------------------
N tot  i ic BS  = ----------------------------------- -,0)
UL LT
1 – X  i ic BS  i
tx
UL
2nd step: Calculation of the total quality at the transmitter on icBS ( Q MAX ) based on the receiver handover status.

UL UL
Q MAX  ic BS  = Q max  i ic BS 

UL UL UL
Q MAX  ic BS  =  G macro – diversity  2 links  max  Q max  i ic BS  
i
UL
UL UL
i

UL UL UL
i
UL
 Qmax  i icBS 
UL UL UL
i
407
have:
 UL 
 Qmax  i icBS  Qmaxi on the other site  i icBS 
UL UL UL UL
Q MAX  ic BS  =  G macro – diversity  2 links  max  f rake efficiency 

i on the same site
otherwise,
UL UL UL
i
UL
3rd step: Evaluation of the required quality level on uplink, Q req
In case of a 1xEV-DO Rev. 0 capable terminal, we have:

E UL
Q req =  ----c-
UL UL
 G p   1 + G ACK + G DRC + G TCH 
 N t min
Where
Ec
 ----
UL
- is the minimum pilot quality level on uplink. This parameter is available in the Mobility types table.
 N t min
G ACK , G DRC and G TCH are respectively acknowledgement, data rate control and traffic data gains relative to the pilot. They
are defined in the terminal properties (1xEV-DO Rev. 0 tab).
In case of a 1xEV-DO Rev. A capable terminal, we have:
E UL
Q req =  ----c-
UL UL
 G p   1 + G ACK + G DRC + G TCH + G RRI + G Auxiliary – pilot 
 N t min
Where
Ec
 ----
UL
UL
- is the minimum pilot quality level required on uplink to obtain the defined throughput, TP req . The required
 N t min
UL
throughput, TPreq (i.e. the uplink throughput selected for the analysis) is obtained by using a certain uplink transmission
format (i.e. 1xEV-DO radio bearer ( Index UL – Bearer ) with a certain number of subframes ( n SF )) and calculated as follows:
UL
UL TP P – R LC  Index UL – Bearer 
TP req = ------------------------------------------------------------
-
n SF
Ec
 ----
UL
-
 N t min is the value defined in the 1xEV-DO Radio Bearer Selection (Uplink) table for this uplink transmission format (radio
bearer Index, mobility and number of subframe). Two values are available for this parameter, one when the service uplink
mode is "Low Latency" and another one for high capacity services.
G ACK , G DRC , G TCH , G RRI and G Auxiliary – pilot are respectively acknowledgement, data rate control, traffic data channel,
reverse rate indicator and auxiliary pilot channel gains relative to the pilot. They are defined in the terminal properties (1xEV-
DO Rev. A tab). Two values of G TCH are available, one when the service uplink mode is "Low Latency" and another one for
high capacity services.
req
4th step: Calculation of P term
UL
req Q req
-  P max
P term = -------------------------- term
UL
Q MAX  ic BS 
req max
Therefore, the service on the uplink traffic data channel is available if P term  P term .
• Required terminal power without ACK

Atoll also calculates the required terminal power without taking into account the ACK channel contribution. Calculations are
quite similar to those detailed in the previous paragraph, only the evaluation of the required quality on uplink is different.
E UL
 Q req  withoutACK =  ----c-
UL UL
 G p   1 + G DRC + G TCH  for 1xEV-DO Rev. 0 capable terminals
N t min
408
And
E UL
 Q req  withoutACK =  ----c-
UL UL
 G p   1 + G DRC + G TCH + G RRI + G Auxiliary – pilot  for 1xEV-DO Rev. A capable terminals
 N t min
And then,
UL
req  Q req  withoutACK
-  P max
 P term  withoutACK = -------------------------------------- term
UL
Q MAX  ic BS 
• UL SHO gain
UL
1st step: Evaluation of the uplink quality, Q max  i ic BS  , for each cell of active set.

UL
UL  term  P b – max  i ic BS 
-  G UL
Q max  i ic BS  = ----------------------------------------------------- p
UL
max
UL P term
With P b –max  i ic BS  = ------------
LT
i

i

UL
L T = ----------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
UL
UL
tx max req
UL N0 P term – P term
N tot  i ic BS  = -----------------------------------
- +  1 –  term   max (------------------------------
-,0)
UL
1 – X  i ic BS  LT
i
tx
UL
2nd step: Calculation of the total quality at the transmitter on icBS ( Q MAX ) based on the receiver handover status.
UL
Q MAX  ic BS  is the traffic channel quality at the transmitter on icBS after signal combination of all the transmitters of the active
set.
UL UL
Q MAX  ic BS  = Q max  i ic BS 

UL UL UL
i
UL
UL UL
i

UL UL UL
i
UL
409
 Qmax  i icBS 
UL UL UL
i
have:
 UL 
 Qmax  i icBS  Qmaxi on the other site  i icBS 
UL UL UL UL
Q MAX  ic BS  =  G macro – diversity  2 links  max  f rake efficiency 

i on the same site
otherwise,
UL UL UL
i
3rd step: Calculation of the UL SHO gain

UL
G SHO corresponds to the uplink soft handover gain.
UL
UL Q MAX  ic BS 
G SHO = -----------------------------------------------
-
UL
max  Q max  i ic BS  
i
• Bearer Consumption
For (1xEV-DO Rev. A - Guaranteed bit rate) service users, Atoll calculates the 1xEV-DO bearer consumption.
TPD min – UL
C UL – Bearer = --------------------------------------------------------------
UL
TP P – RLC  Index UL – Bearer 
Where TPD min – UL corresponds to the minimum bit rate required by the service in the uplink.
1xEV-DO Rev. B Service Users
For multi-carrier 1xEV-DO Rev. B users, Atoll models load balancing between carriers. Atoll shares the available terminal
power between the carriers and determines the uplink 1xEV-DO radio bearer obtained on each carrier. It starts allocating the
maximum terminal power to the best carrier and selects the highest 1xEV-DO radio bearer. If it remains terminal power after
serving the first carrier, then Atoll continues allocating the available terminal power to the second carrier, and so on for the
other carriers of the active set as long as the remaining terminal power is sufficient to obtain the lowest bearer.
The following results are displayed:
• For each carrier used in the selected configuration, Atoll indicates the UL SHO Gain, the obtained throughput and the
required power.
The calculations can be divided into four steps:
UL
1st step: Evaluation of the uplink quality, Q max  i ic  , for each cell of the sub-active set
For each cell (i,ic), we have:

UL
UL  term  P b –max  i ic 
-  G UL
Q max  i ic  = ------------------------------------------------ p
UL
N tot  i ic 
max
UL P term  ic 
With P b –max  i ic  = ---------------------
-
LT
i
max
P term  ic  is the terminal power available for the carrier (ic).

i

UL
L T = ----------------------------------------------------------------------------------------------------------------------------------------------
-
G Tx  G term
UL UL
N tot  i ic  is the total noise at the transmitter on the carrier ic. This value is deduced from the cell uplink load factor X  i ic  .
tx max req
UL N0 P term  ic  – P term
- +  1 –  term   max (----------------------------------------
N tot  i ic  = ------------------------------ ,0)
UL LT
1 – X  i ic  i
410
tx
UL
2nd step: Calculation of the total quality at the transmitter on ic ( Q MAX ) based on the receiver handover status.

UL UL
Q MAX  ic  = Q max  i ic 

UL UL UL
Q MAX  ic  =  G macro – diversity  2 links  max  Q max  i ic  
i
UL
UL UL
max  Q max  i ic   corresponds to the highest Q max  i ic  value.
i

UL UL UL
i
UL
 Qmax  i ic 
UL UL UL
Q MAX  ic  = f rake efficiency 
i
have:
 UL 
 Qmax  i ic  Qmaxi on the other site  i ic 
UL UL UL UL
Q MAX  ic  =  G macro – diversity  2 links  max  f rake efficiency 

i on the same site
otherwise,
UL UL UL
i
UL
3rd step: Calculation of the UL SHO gain ( G SHO )
UL
UL Q MAX  ic 
G SHO = -------------------------------------------
UL
max  Q max  i ic  
i
4th step: Selection of the uplink 1xEV-DO radio bearer

UL req
Atoll evaluates of the required quality level in the uplink ( Q req ) and the required terminal power ( P term  ic  ) for each 1xEV-
DO radio bearer.
E UL
Q req =  ----c-
UL UL
N t min
Where
Ec
 ----
UL
- is the minimum pilot quality level required in the uplink to obtain the 1xEV-DO radio bearer. The values are defined
 N t min
in the 1xEV-DO Radio Bearer Selection (Uplink) table for each uplink transmission format (radio bearer Index, mobility and
number of subframe). Two values are available, one when the service uplink mode is "Low Latency" and another one for high
capacity services.
411
And
UL
req Q req
-  P max
P term  ic  = --------------------- term  ic  
UL
Q MAX  ic 
Then, Atoll selects the best 1xEV-DO radio bearer. This is the 1xEV-DO radio bearer ( Index UL – Bearer ) with the highest
UL
UL TPP – RLC  Index UL – Bearer 
obtained throughput ( TP  ic  = -------------------------------------------------------------- ) where:
n SF  Index UL – Bearer 
req max
• P term  ic   P term  ic   ,
• And the required modulation scheme is supported by the terminal.
n SF is the number of subframes associated with the 1xEV-DO radio bearer ( Index UL – Bearer ).
• Max terminal power

max
allowed.
UL
The required throughput, R req , is the uplink throughput selected for the analysis.
Atoll calculates the throughput for all combinations of carriers.
m
 TP
UL UL
TP total =  ic  where m corresponds to the number of carriers in the combination.
ic = 1
UL
The obtained throughput ( TP ) corresponds to the best configuration among all combinations of carriers, i.e., the
UL
combination which provides the highest throughput, Max  TP total  .
UL UL
The traffic data channel is available in uplink if TP  TP req .
• Required terminal power

m
 Pterm  ic 
req req
P term =
ic = 1

Atoll calculates CDMA-specific coverage studies on each pixel where the pilot signal level exceeds the minimum RSCP
threshold. Let us assume each pixel of the map corresponds to a probe receiver with a terminal, a mobility type and a service.
This receiver does not create any interference. You can make the coverage prediction for a specific carrier or for the best
1xRTT or 1xEV-DO carrier. The type of carrier and the carriers you can select depend on the service and on the frequency
band(s) supported by the terminal. Coverage studies are based on the uplink load percentage and the downlink total power
of cells. These parameters can either be either simulation results, or average values calculated from a group of simulations,
or user-defined cell inputs.
5.5.2.1 Pilot Quality Analysis

For further details on calculation formulas, see "Definitions and Formulas" on page 340. For further details on calculations,
see "Bar Graph and Pilot Sub-Menu" on page 392

The carrier that can be used by transmitters is fixed. In this case, for each transmitter i containing the receiver in its calculation
area and using the selected carrier, Atoll calculates pilot quality at the receiver on this carrier icgiven. Then, it determines the
best serving transmitter BS using the carrier icgiven ( Q pilot  ic given  ) and deduces the best pilot quality received with a fixed
BS
Resulting
cell edge coverage probability, Q pilot  ic given  .
412
2nd case: Analysis based on the best carrier of all frequency bands
Atoll proceeds as in point analysis. It determines the best carrier of each transmitter i containing the receiver in its calculation
area and using a frequency band supported by the receiver’s terminal. The best carrier selection depends on the option
selected for the site equipment (UL minimum noise, DL minimum power, random, sequential) and is based on the UL load
percentage and the downlink total power of cells (simulation results or cell properties). Atoll calculates the pilot quality at the
receiver from these transmitters on their best carrier and determines the best serving transmitter BS on its best carrier icBS
Resulting
( Q pilot  ic BS  ). Then, it deduces the best pilot quality received with a fixed cell edge coverage probability, Q pilot  ic BS  .
BS
3rd case: Analysis based on the best carrier of any frequency band (for multi-band terminals with priority defined on frequency
bands only)
The frequency band that can be used is fixed. Atoll determines the best carrier of each transmitter i containing the receiver
in its calculation area and using the selected frequency band. The best carrier selection depends on the option selected for
the site equipment (UL minimum noise, DL minimum power, random, sequential) and is based on the UL load percentage and
the downlink total power of cells (simulation results or cell properties). Then, Atoll calculates the pilot quality at the receiver
from these transmitters on their best carrier and determines the best serving transmitter BS on its best carrier icBS
Resulting
( Q pilot  ic BS  ). Then, it calculates the best pilot quality received with a fixed cell edge coverage probability, Q pilot  ic BS  .
BS
• Single colour
Resulting req
Atoll displays a coverage if Q pilot  ic   Q pilot . Coverage consists of a single layer with a unique colour.
ic = ic BS or ic given
• Colour per transmitter

Resulting req
Atoll displays a coverage if Q pilot  ic   Q pilot ( ic = ic BS or ic given ). Coverage consists of several layers with associated
colours. There is a layer per transmitter with no intersection between layers. Layer colour is the colour assigned to the best
serving transmitter BS.
• Colour per mobility
In this case, the receiver is not completely defined and no mobility assigned. Coverage consists of several layers with a layer
per user-defined mobility type defined in the Mobility Types sub-folder. For each layer, area is covered if
Resulting req
Q pilot  ic   Q pilot ( ic = ic BS or ic given ). Each layer is assigned a colour and displayed with intersections between layers.
• Colour per probability

and the “All” option in the Condition tab of prediction properties).
Coverage consists of several layers with a layer per user-defined probability level defined in the Display tab (Prediction
Resulting req
properties). For each layer, area is covered if Q pilot  ic   Q pilot ( ic = ic BS or ic given ) in the required number of
simulations. Each layer is assigned a colour and displayed with intersections between layers.
• Colour per cell edge coverage probability
Coverage consists of several layers with a layer per user-defined cell edge coverage probability, p, defined in the Display tab
Resulting req
(Prediction properties). For each layer, area is covered if Q pilot  ic p   Q pilot ( ic = ic BS or ic given ). Each layer is assigned
a colour and displayed with intersections between layers.
• Colour per quality level (Ec/I0)
Coverage consists of several layers with a layer per user-defined quality threshold defined in the Display tab (Prediction
Resulting
properties). For each layer, area is covered if Q pilot  ic    Q pilot  threshold ( ic = ic BS or ic given ). Each layer is assigned a
colour and displayed with intersections between layers.
• Colour per quality margin (Ec/I0 margin)
Coverage consists of several layers with a layer per user-defined quality margin defined in the Display tab (Prediction
Resulting req
properties). For each layer, area is covered if Q pilot  ic  – Q pilot   Q pilot  m arg in ( ic = ic BS or ic given ). Each layer is
assigned a colour and displayed with intersections between layers.
413
Colour per pilot signal level (Ec)
Coverage consists of several layers with a layer per user-defined pilot signal level defined in the Display tab (Prediction
Resulting
properties). For each layer, area is covered if Q pilot  ic    Q pilot  threshold ( ic = ic BS or ic given ). Each layer is assigned a
5.5.2.2 Downlink Service Area Analysis

The downlink service area analysis depends on the studied network (CDMA2000 1xRTT or CDMA2000 1xEV-DO). Several
display options are available when calculating this study, some of which are dedicated to CDMA2000 1xRTT networks while
others are relevant when analysing CDMA2000 1xEV-DO systems only.
5.5.2.2.1 CDMA2000 1xRTT

As in point analysis, Atoll calculates downlink quality on FCH at the receiver for each cell (k,ic) (with ic=icBS or icgiven) (these
cells are the first f cells in the receiver’s active set and f is the number of fingers defined for the terminal). No power control
is performed as in simulations. Here, Atoll determines the downlink quality on FCH at the receiver for a maximum traffic
channel power per transmitter allowed on the fundamental channel (FCH). Then, the total downlink quality on FCH
DL
(  Q MAX  ic   FCH ) is evaluated after recombination.
Best server and active set determination is performed as in point prediction.
Atoll displays total traffic channel quality at the receiver on the carrier ic ( ic BS or ic given ).
For further details on formulas, see "Definitions and Formulas" on page 340. For further details on calculation, see "Downlink
Sub-Menu" on page 395.
You may choose following display options:
• Single colour
DL DL DL
Atoll displays a coverage with a unique colour if  Q MAX  ic   FCH   Q req  FCH .  Q req  FCH is the downlink traffic data quality
target on the fundamental channel (FCH). This parameter is user-defined for a given service and a terminal in the Services sub-
folder.
DL DL
Atoll displays a coverage if  Q MAX  ic   FCH   Q req  FCH . Coverage consists of several layers with associated colours. There is
a layer per transmitter with no intersection between layers. Layer colour is the colour assigned to best serving transmitter.
In this case, receiver is not completely defined and no mobility is assigned. Coverage consists of several layers with a layer per
DL DL
user-defined mobility defined in Mobility sub-folder. For each layer, area is covered if  Q MAX  ic   FCH   Q req  FCH . Each layer
is assigned a colour and displayed with intersections between layers.
• Colour per service
In this case, receiver is not completely defined and no service is assigned. Coverage consists of several layers with a layer per
DL DL
user-defined service defined in Services sub-folder. For each layer, area is covered if  Q MAX  ic   FCH   Q req  FCH . Each layer
and the “All” option in the Condition tab of prediction properties).
Coverage consists of several layers with a layer per user-defined probability level defined in the Display tab (Prediction
DL DL
properties). For each layer, area is covered if  Q MAX  ic   FCH   Q req  FCH in the required number of simulations. Each layer is
414
DL DL
(Prediction properties). For each layer, area is covered if  Q MAX  ic p   FCH   Q req  FCH . Each layer is assigned a colour and
displayed with intersections between layers.
• Colour per maximum quality level (max Eb/Nt)
DL
properties). For each layer, area is covered if  Q MAX  ic   FCH  Threshold . Each layer is assigned a colour and displayed with
intersections between layers.
• Colour per effective quality level (Effective Eb/Nt)
DL
properties). For each layer, area is covered if  Q eff  ic   FCH  Threshold . Each layer is assigned a colour and displayed with
• Colour per quality margin (Eb/Nt margin)
DL DL
properties). For each layer, area is covered if  Q MAX  ic   FCH –  Q req  FCH  M arg in . Each layer is assigned a colour and
• Colour per required power
req
Atoll calculates the downlink power required on FCH, P FCH  ic  , as follows:
DL
req  Q req FCH
-  P max
P FCH  ic  = ---------------------- FCH
DL
Q MAX  ic 
max
Where P FCH is a user-defined input for a given service and terminal. It corresponds to the maximum traffic data power
allowed on FCH for a transmitter.
Coverage consists of several layers with a layer per user-defined required power threshold defined in the Display tab
req
(Prediction properties). For each layer, area is covered if P FCH  ic   Threshold . Each layer is assigned a colour and displayed
with intersections between layers.
• Colour per required power margin
Coverage consists of several layers with a layer per user-defined power margin defined in the Display tab (Prediction
max req
properties). For each layer, area is covered if P FCH – P FCH  ic   M arg in . Each layer is assigned a colour and displayed with
• Colour per throughput
DL FCH DL
This display option is relevant for CDMA2000 1xRTT data services only. For each possible throughput, TP ( TP P – DL  AF FCH ,
FCH DL FCH DL FCH DL FCH DL
TP P – DL   AF FCH + 2  , TP P – DL   AF FCH + 4  , TPP – DL   AF FCH + 8  , TP P – DL   AF FCH + 16  ), Atoll calculates traffic
channel quality at the receiver for each cell (k,ic) (with ic=icBS or icgiven). Downlink traffic channel quality at the receiver is
evaluated from a maximum traffic channel power per transmitter allowed for the corresponding throughput. Then, the total
DL DL
downlink traffic channel quality ( Q MAX  ic TP  ) is calculated after recombination.
DL
Coverage consists of several layers with a layer per possible throughput, TP . For each layer, area is covered if
DL DL DL DL
Q MAX  ic TP   Q req  TP  . Each layer is assigned a colour and displayed with intersections between layers.
DL DL DL
Q req  TP  is the downlink traffic data quality target for the throughput, TP . This parameter is user-defined for a given
service, terminal and throughput in the Services sub-folder.
5.5.2.2.2 CDMA2000 1xEV-DO

E
As in point analysis, Atoll calculates the effective pilot quality level at the receiver from the best server cell, ----c-  ic b pilot  . Best
Nt
server and active set determination is performed as in point prediction (AS analysis). Then, from this value, it determines the
DL
effective downlink throughput received, TP .
For further details on formulas, see "Definitions and Formulas" on page 340. For further details on calculations, see "Downlink
415
1xEV-DO Rev. 0 Users
For 1xEV-DO Rev. 0 users (users with EV-DO Rev. 0-capable terminals and EV-DO Rev. 0 services), the obtained throughput
DL
( TP ) on downlink is determined from the graph “Peak throughput=f(C/I) (Rev0)” defined for the mobility type selected in
DL
the Condition tab (Prediction properties). TP is the value read in the graph “Peak throughput=f(C/I) (Rev0)” for the
E
calculated effective pilot quality level, ----c-  ic BS b pilot  .
Nt
1xEV-DO Rev. A Users
For 1xEV-DO Rev. A users (users with EV-DO Rev. A-capable terminals and EV-DO Rev. A services), the obtained throughput
DL
( TP ) on downlink depends on the downlink transmission format, i.e the radio bearer index ( Index DL – Bearer ) with the
Ec
number of timeslots ( n TS ). Atoll selects the downlink transmission format where -----  ic BS b pilot    --- . Then, it
C
Nt I req
determines the downlink obtained throughput as follows:
DL
TP = -------------------------------------------------------------- .
n TS
The obtained throughput corresponds to the guaranteed throughput after a certain number of retransmissions (i.e. the
number of timeslots, n TS ).
DL
throughput is an average throughput ( TP av ) calculated as follows:
DL
TP av = ---------------------------------------------------------------------
-
DL
 n Rtx (Index DL – Bearer,n TS)  av
DL
The average number of retransmissions (  n Rtx  av ) is determined from early termination probabilities defined for the selected
downlink transmission format. The Early Termination Probability graph shows the probability of early termination ( p ) as a
DL DL
function of the number of retransmissions ( n Rtx ). Atoll calculates the average number of retransmissions (  n Rtx av ) as
follows:
 n DL 
 Rtx max

DL DL DL
 p  n Rtx  – p  n Rtx – 1    n Rtx
DL
DL n =1
 n Rtx  av = --------------------------------------------------------------------------------------------
Rtx -
DL
p   n Rtx  max 
1xEV-DO Rev. B Users
Single-carrier EV-DO Rev. B service users are managed as 1xEV-DO Rev. A service users.
DL
For multi-carrier EV-DO Rev. B service users, the obtained throughput ( TP ) in the downlink corresponds to the sum of the
obtained throughputs on each carrier.
DL
The obtained throughput ( TP  ic  ) on a carrier depends on the downlink transmission format, i.e the radio bearer index
( Index DL – Bearer ) with the number of timeslots ( n TS ). Atoll selects the downlink transmission format where
E
----c-  ic b pilot    C
---
Nt  I  req and whose modulation scheme is supported by the terminal.
The downlink obtained throughput corresponds to the guaranteed throughput after a certain number of retransmissions (i.e.
the number of timeslots, n TS ). It is determined as follows:
DL
TP  ic  = --------------------------------------------------------------
n TS
DL
throughput on a carrier is an average throughput ( TP av  ic  ) calculated as follows:
416
DL
TP av  ic  = ---------------------------------------------------------------------
-
DL
 n Rtx (Index DL – Bearer,n TS)  av
DL
The average number of retransmissions (  n Rtx av ) is determined from early termination probabilities defined for the selected
downlink transmission format. The Early Termination Probability graph shows the probability of early termination ( p ) as a
DL DL
function of the number of retransmissions ( n Rtx ). Atoll calculates the average number of retransmissions (  n Rtx  av ) as
follows:
 n DL 
 Rtx max

DL DL DL
 p  n Rtx  – p  n Rtx – 1    n Rtx
DL
DL n =1
 n Rtx av = --------------------------------------------------------------------------------------------
Rtx -
DL
DL
The average throughput ( TP av ) provided on downlink corresponds to the sum of the average throughputs obtained on each
carrier.
Display Options
You may choose the following display options:

• Colour per C/I
Coverage consists of several layers with a layer per quality threshold defined in the Display tab (Prediction properties). For
E
each layer, area is covered if ----c-  ic b pilot   Threshold . Each layer is assigned a colour and displayed with intersections
Nt
between layers.
DL
Coverage consists of several layers with a layer per possible throughput ( TP ). For each layer, area is covered if the
DL
throughput, TP , can be obtained. Each layer is assigned a colour and displayed with intersections between layers.
• Colour per average throughput
This display option is available for 1xEV-DO Rev. A and 1xEV-DO Rev. B users only. It enables you to view the obtained downlink
DL
throughput when HARQ is used. Coverage consists of several layers with a layer per possible average throughput ( TP av ). For
DL
each layer, area is covered if the average throughput, TP av , can be obtained. Each layer is assigned a colour and displayed
5.5.2.3 Uplink Service Area Analysis

The results displayed when calculating the uplink service area analysis depend on the studied network (CDMA2000 1xRTT or
CDMA2000 1xEV-DO).
5.5.2.3.1 CDMA2000 1xRTT

As in point analysis, Atoll calculates uplink quality on FCH from receiver for each cell (l,ic) (with ic=icBS or icgiven) in receiver
active set. No power control simulation is performed. Atoll determines uplink quality on FCH at the transmitter for the
UL
maximum terminal power. Then, the total uplink traffic channel quality (  Q MAX  ic   FCH ) is evaluated with respect to the
receiver handover status.
Best server and active set determination is performed as in point prediction (AS analysis).
Atoll displays uplink quality on FCH at transmitters in active set on the carrier ic ( ic BS or ic given ) received from the receiver.
For further details on formulas, see "Definitions and Formulas" on page 340. For further details on calculations, see "Uplink
• Single colour
417
UL UL UL
Atoll displays a coverage if  Q MAX  ic   FCH   Q req  FCH . Coverage colour is unique.  Q req  FCH is the uplink data traffic quality
target on the fundamental channel (FCH). This parameter is user-defined for a given service and a terminal in the Services sub-
folder.
UL UL
Atoll displays a coverage if  Q MAX  ic   FCH   Q req  FCH . Coverage consists of several layers with associated colours. There is
a layer per transmitter with no intersection between layers. Layer colour is the colour assigned to best server transmitter.
UL UL
user-defined mobility defined in Mobility sub-folder. For each layer, area is covered if  Q MAX  ic   FCH   Q req  FCH . Each layer
UL UL
user-defined service defined in Services sub-folder. For each layer, area is covered if  Q MAX  ic   FCH   Q req  FCH . Each layer
and the “All” option in the Condition tab of prediction properties). Coverage consists of several layers with a layer per user-
defined probability level defined in the Display tab (Prediction properties). For each layer, area is covered if
UL UL
 Q MAX  ic   FCH   Q req  FCH in the required number of simulations. Each layer is assigned a colour and displayed with
UL UL
(Prediction properties). For each layer, area is covered if  Q MAX  ic p   FCH   Q req  FCH . Each layer is assigned a colour and
• Colour per maximum quality level (Max Eb/Nt)
UL
properties). For each layer, area is covered if  Q MAX  ic   FCH  Threshold . Each layer is assigned a colour and displayed with
UL
properties). For each layer, area is covered if  Q effective  ic   FCH  Threshold . Each layer is assigned a colour and displayed
UL UL
properties). For each layer, area is covered if  Q MAX  ic   FCH –  Q req  FCH  M arg in . Each layer is assigned a colour and
FCH – re q
The required terminal power, P term , is calculated as described in the Point analysis – AS analysis tab – Uplink sub-menu
part. Coverage consists of several layers with a layer per user-defined power threshold defined in the Display tab (Prediction
FCH –re q
properties). For each layer, area is covered if P term  ic   Threshold . Each layer is assigned a colour and displayed with
max FCH –re q
properties). For each layer, area is covered if P term – P term  ic   M arg in . Each layer is assigned a colour and displayed
418
UL FCH UL
This display option is relevant for CDMA2000 1xRTT data services only. For each possible throughput, TP ( TP P – UL  AF FCH ,
FCH UL FCH UL FCH UL FCH UL
TP P – UL   AF FCH + 2  , TP P – UL   AF FCH + 4  , TP P – UL   AF FCH + 8  , TP P – UL   AF FCH + 16  ), Atoll calculates the total
UL UL
uplink traffic channel quality ( Q MAX  ic TP  ). Coverage consists of several layers with a layer per possible throughput,
UL UL UL UL UL
TP . For each layer, area is covered if Q MAX  ic TP   Q req  TP  . Each layer is assigned a colour and displayed with
UL UL UL
intersections between layers. Q req  TP  is the uplink traffic data quality target for the throughput, TP . This parameter is
user-defined for the service, a given terminal and throughput in the service properties.
5.5.2.3.2 CDMA2000 1xEV-DO

As in point analysis, Atoll calculates the uplink quality from receiver for each cell (l,ic) (with ic=icBS or icgiven) in receiver active
set. No power control simulation is performed. For 1xEV-DO Rev. 0 users, Atoll determines the uplink quality at the
transmitter for the maximum terminal power allowed and an uplink channel throughput of 9.6 kbps. For 1xEV-DO Rev. A and
1xEV-DO Rev. B users, Atoll determines the uplink quality at the transmitter for the maximum terminal power allowed and an
UL
uplink channel throughput of 4.8 kbps. Then, the total uplink quality ( Q MAX  ic  ) is evaluated with respect to the receiver
handover status.
Best server and active set determination is performed as in point prediction (AS analysis).
Atoll displays the uplink quality at transmitters in active set on the carrier ic ( ic BS or ic given ) received from the receiver. For
multi-carrier EV-DO users, Atoll considers the best sub-active set.
For further details on formulas, see "Definitions and Formulas" on page 340. For further details on calculations, see "Uplink
• Single colour
UL UL UL
Atoll displays a coverage if Q MAX  ic   Q req . Coverage colour is unique. For 1xEV-DO Rev. 0 users, Q req is the quality required
UL
on uplink for a 9.6 kbps channel throughput. For 1xEV-DO Rev. A and 1xEV-DO Rev. B users, Q req is the quality required on
uplink for a 4.8 kbps channel throughput. This parameter is calculated from the minimum uplink pilot quality and gains on the
different uplink channels.
We have:
E UL
Q req =  ----c-
UL UL
 G p   1 + G ACK + G DRC + G TCH  for 1xEV-DO Rev. 0 terminals,
 N t min
And
E UL
Q req =  ----c-
UL UL
 G p   1 + G ACK + G RRI + G DRC + G TCH + G Auxiliary – Pilot  for 1xEV-DO Rev. A and 1xEV-DO Rev. B terminals.
N t min

UL UL
Atoll displays a coverage if Q MAX  ic   Q req . Coverage consists of several layers with associated colours. There is a layer per
transmitter with no intersection between layers. Layer colour is the colour assigned to best server transmitter.
UL UL
user-defined mobility defined in Mobility sub-folder. For each layer, area is covered if Q MAX  ic   Q req . Each layer is assigned
UL UL
user-defined service defined in Services sub-folder. For each layer, area is covered if Q MAX  ic   Q req . Each layer is assigned
and the “All” option in the Condition tab of prediction properties). Coverage consists of several layers with a layer per user-
419
UL UL
defined probability level defined in the Display tab (Prediction properties). For each layer, area is covered if Q MAX  ic   Q req
in the required number of simulations. Each layer is assigned a colour and displayed with intersections between layers.
UL UL
(Prediction properties). For each layer, area is covered if Q MAX  ic p   Q req . Each layer is assigned a colour and displayed
• Colour per maximum quality level (Max Eb/Nt)
UL
Here, Atoll calculates the total uplink traffic channel quality (  Q MAX  ic   TCH ).
UL
UL  Q req  TCH
-  P max
 Q MAX  ic   TCH = ---------------------- term
req
P term
With
E UL
 Q req  TCH =  ----c-
UL UL
 G p  G TCH
 N t min
UL
properties). For each layer, area is covered if  Q MAX  ic   TCH  Threshold . Each layer is assigned a colour and displayed with
UL
properties). For each layer, area is covered if  Q effective  ic   TCH  Threshold . Each layer is assigned a colour and displayed
UL UL UL
 Q effective  ic   TCH = min   Q MAX  ic   TCH  Q req  TCH 

UL UL
properties). For each layer, area is covered if Q MAX  ic  – Q req  M arg in . Each layer is assigned a colour and displayed with
1xEV-DO Rev. 0, 1xEV-DO Rev.A and single-carrier 1xEV-DO Rev. B service users
Coverage consists of several layers with a layer per user-defined power threshold defined in the Display tab (Prediction
TCH – re q
properties). For each layer, area is covered if P term  ic   Threshold . Each layer is assigned a colour and displayed with
TCH –re q
The required terminal power on traffic data channel, P term , is calculated as described in the Point analysis – AS analysis
tab – Uplink sub-menu part.
req
TCH – re q P term
P term = ---------------------------------------------------------  G TCH for 1xEV-DO Rev. 0 terminals,
1 + G ACK + G DRC + G TCH
And
req
TCH – re q P term
P term -  G TCH for 1xEV-DO Rev. A terminals.
= -------------------------------------------------------------------------------------------------------------------
1 + G ACK + G RRI + G DRC + G TCH + G Auxiliary – Pilot
Multi-carrier 1xEV-DO Rev. B service users

For multi-carrier EV-DO users, the coverage consists of several layers with a layer per user-defined power threshold defined
TCH –re q
in the Display tab (Prediction properties). For each layer, area is covered if P term  Threshold . Each layer is assigned a
TCH – re q
For the selected configuration (i.e., the combination of carriers which provides the highest throughput), P term corresponds
to the sum of the terminal powers required on each carrier of the configuration.
420
max req
properties). For each layer, area is covered if P term – P term  ic   M arg in . Each layer is assigned a colour and displayed with
1xEV-DO Rev. 0 service users
UL UL UL
For each possible throughput, TP , Atoll calculates the total uplink quality ( Q MAX  ic TP  ). Coverage consists of several
UL UL UL UL
layers with a layer per possible throughput. For each layer, area is covered if Q MAX  ic TP   Q req  TP  . Each layer is
UL UL UL
Q req  TP  is the uplink quality required to obtain the throughput, TP .
UL
The possible throughputs on uplink, TP , are: 9.6, 19.2, 38.4, 76.8 and 153.6 kbps
E UL
Q req =  ----c-
UL UL
 G p   1 + G ACK + G DRC + G TCH 
 N t min
Where
Ec
 ----
UL
-
 N t min is the minimum pilot quality level on uplink. This parameter is available in the Mobility types table.
G ACK , G DRC and G TCH are respectively acknowledgement, data rate control and traffic data gains relative to the pilot. They
are defined in the terminal properties (1xEV-DO Rev. 0 tab).
1xEV-DO Rev. A and single-carrier 1xEV-DO Rev. B service users
UL UL UL
For each possible throughput, TP , Atoll calculates the total uplink quality ( Q MAX  ic TP  ). Coverage consists of several
UL UL UL UL
layers with a layer per possible throughput. For each layer, area is covered if Q MAX  ic TP   Q req  R v  . Each layer is
UL UL UL
Q req  TP  is the uplink quality required to obtain the throughput, TP .
UL
The throughput, TP is obtained when a certain uplink transmission format (i.e. 1xEV-DO radio bearer ( Index UL – Bearer )
with a certain number of subframes ( n SF )) is used. It is calculated as follows:
UL
UL TP P – RLC  Index UL – Bearer 
TP req = --------------------------------------------------------------
n SF
E UL
Q req =  ----c-
UL UL
 N t min
Where
Ec
 ----
UL
UL
-
 N t min is the minimum pilot quality level required on uplink to obtain the throughput, TP . The value is defined in the
1xEV-DO Radio Bearer Selection (Uplink) table for the uplink transmission format (radio bearer Index, mobility and number of
subframe). Two values are available for this parameter, one when the service uplink mode is "Low Latency" and another one
for high capacity services.
For multi-carrier 1xEV-DO Rev. B users, Atoll models load balancing between carriers. Atoll allocates the available terminal
power to carriers sequentially and determines the uplink 1xEV-DO radio bearer obtained on each carrier. Then, it selects the
best configuration among all combinations of carriers, i.e., the combination which provides the highest throughput.
UL UL
Coverage consists of several layers with a layer per possible throughput. For each layer, area is covered if TP  TP req . Each
layer is assigned a colour and displayed with intersections between layers.
421
UL
TP req is the uplink throughput associated with the layer.
UL
TP corresponds to the throughput of the best configuration, i.e., the combination which provides the highest throughput.
• Colour per average throughput
This display option is available for 1xEV-DO Rev. A and 1xEV-DO Rev. B users only. When HARQ (Hybrid Automatic Repeat
Request) is used, the required average number of retransmissions is smaller and the throughput is an average throughput
UL
( TP av ) calculated as follows:
UL
UL TP P – RLC  Index UL – Bearer 
TP av = -----------------------------------------------------------------------
-
UL
 n Rtx  Index UL – Bearer n SF   av
UL
The average number of retransmissions (  n Rtx  av ) is determined from early termination probabilities defined for the selected
uplink transmission format (i.e. the radio bearer index ( Index UL – Bearer ) with the number of subframes ( n SF )). The Early
Termination Probability graph shows the probability of early termination ( p ) as a function of the number of retransmissions
UL UL
( n Rtx ). Atoll calculates the average number of retransmissions (  n Rtx  av ) as follows:
 n UL 
 Rtx max

UL UL UL
 p  n Rtx  – p  n Rtx – 1    n Rtx
UL
UL n =1
 n Rtx  av = --------------------------------------------------------------------------------------------
Rtx -
UL
1xEV-DO Rev. A and single-carrier 1xEV-DO Rev. B service users

UL UL UL
For each possible average throughput, TP av , Atoll calculates the total uplink quality ( Q MAX  ic TP av  ). Coverage consists of
UL UL UL UL
several layers with a layer per possible average throughput. For each layer, area is covered if Q MAX  ic TP av   Q req  TP av  .
Each layer is assigned a colour and displayed with intersections between layers.
UL UL UL
Q req  TP av  is the uplink quality required to obtain the average throughput, TP av .

For multi-carrier 1xEV-DO Rev. B users, the coverage consists of several layers with a layer per possible throughput. For each
UL UL
layer, area is covered if TP av  TP req . Each layer is assigned a colour and displayed with intersections between layers.
UL
TP req is the uplink throughput associated with the layer.
UL
For the selected configuration (i.e., the combination of carriers which provides the highest throughput), TP av corresponds to
the sum of the average throughputs obtained on each carrier of the configuration.
5.5.2.4 Downlink Total Noise Analysis

Atoll determines downlink total noise generated by cells.
For CDMA2000 1xRTT systems, we have:
DL
 Ptot  ic  + ------------------------------------
N tot  ic  = + N0
txj j
For CDMA2000 1xEV-DO systems, we have:
 Ptot  icadj bpilot 

DL
 Ptot  ic bpilot  + ---------------------------------------------------

N tot  ic  = - + N0
txj j
term
DL  N0 
Downlink noise rise, NR DL  ic  , is calculated from the downlink total noise, N tot , as: NR DL  ic  = – 10 log  ------------
-
 N DL tot 
422
5.5.2.4.1 Analysis on the Best Carrier

If the best carrier is selected, Atoll determines DL total noise for the best carrier. Then, allows the user to choose different
displays.
• Colour per minimum noise level
Coverage consists of several layers with a layer per user-defined noise level defined in the Display tab (Prediction properties).
DL
For each layer, area is covered if min NR tot  ic   Threshold . Each layer is assigned a colour and displayed with
ic
• Colour per maximum noise level
DL
For each layer, area is covered if max NR tot  ic   Threshold . Each layer is assigned a colour and displayed with intersections
ic
between layers.
• Colour per average noise level
DL
For each layer, area is covered if average NRtot  ic   Threshold . Each layer is assigned a colour and displayed with
ic
• Colour per minimum noise rise
Atoll displays bins where min NR DL  ic   Threshold . Coverage consists of several areas with an area per user-defined noise
ic
rise threshold defined in the Display tab. Each area is assigned a colour with intersections between areas.
• Colour per maximum noise rise
Atoll displays bins where max NR DL  ic   Threshold . Coverage consists of several areas with an area per user-defined noise
ic
rise threshold defined in the Display tab. Each area is assigned a colour with intersections between areas.
• Colour per average noise rise
Atoll displays bins where average NRDL  ic   Threshold . Coverage consists of several areas with an area per user-defined
ic
noise rise threshold defined in the Display tab. Each area is assigned a colour with intersections between areas.
5.5.2.4.2 Analysis on a Specific Carrier

When only one carrier is analysed, Atoll determines DL total noise or DL noise rise on this carrier. In this case, the displayed
coverage is the same for any selected display per noise level (average, minimum or maximum) or any display per noise rise
(average, minimum or maximum).
• Colour per noise level
DL
For each layer, area is covered if N tot  ic   Threshold . Each layer is assigned a colour and displayed with intersections
between layers.
• Colour per noise rise
Atoll displays bins where NRDL  ic   Threshold . Coverage consists of several areas with an area per user-defined noise rise
threshold defined in the Display tab. Each area is assigned a colour with intersections between areas.

Atoll permits the automatic allocation of intra-technology neighbours in the current network. Two allocation algorithms are
available, one dedicated to intra-carrier neighbours and the other for inter-carrier neighbours.
The intra-technology neighbour allocation algorithms take into account all the cells of TBC transmitters. It means that all the
cells of TBC transmitters of your .atl document are potential neighbours.
The cells to be allocated will be called TBA cells. They must fulfill the following conditions:
423
If no focus zone exists in the ATL document, Atoll takes into account the computation
zone.

• "Neighbour Allocation for all Transmitters" on page 424.
5.6.1 Neighbour Allocation for all Transmitters

We assume that we have a reference cell A and a candidate neighbour, cell B. When automatic allocation starts, Atoll checks
• The distance between both cells must be less than the user-definable maximum inter-site distance. If the distance
discarded.
• Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. You may
choose one or more carriers. Atoll will allocate neighbours to cells using the selected carriers.
• Force co-site cells as neighbours: This option enables you to force cells located on the reference cell site in the
candidate neighbour list. This constraints can be weighted among the others and ranks the neighbours through
the importance field (see after).
• Force adjacent cells as neighbours (only for intra-carrier neighbours): This option enables you to force cells
geographically adjacent to the reference cell in the candidate neighbour list.This constraints can be weighted
among the others and ranks the neighbours through the importance field (see after).
reference cell is a candidate neighbour of another cell, this one will be considered as candidate neighbour of the
reference cell.
[Neighbours]
Let CellB be a candidate neighbour cell of CellA. CellB is considered adjacent to CellA if
there exists at least one pixel in the CellA Best Server coverage area where CellB is Best
Server (if several cells have the same best server value) or CellB is the second best server
that enters the Active Set (respecting the T_Drop of the allocation).
When the Force adjacent cells as neighbours check box is selected, adjacent cells are
sorted and listed from the most adjacent to the least, depending on the above criterion.
Adjacence is relative to the number of pixels satisfying the criterion.
424
• Intra-carrier neighbours: intra-carrier handover is a soft handover.

The reference cell A and the candidate cell B are located inside a continuous layer of cells with carrier c1 (c1 is the selected
carrier on which you run the allocation).
SA is the area where the cell A is the best serving cell. It means that the cell A is the first one in the active set.
• The pilot signal received from the cell A is greater than the minimum pilot signal level.
• The pilot quality from A exceeds Min. Ec/I0.
• The pilot quality from A is the best.
SB is the area where the cell B can enter the active set.
• The pilot signal received from the cell B is greater than the minimum pilot signal level.
• The pilot quality from B is greater than T_Drop.
• Inter-carrier neighbours: inter-frequency handover is a hard handover. It is needed in a multi-carrier (1xRTT and 1xEV-
DO carriers) CDMA network:
• To balance loading between carriers and layers (1st case),
• To make a coverage reason handover from micro cell frequency to macro cells (2nd case).
1st case: the reference cell A is located inside a continuous layer of cells with carrier c1 (c1 is the selected carrier on which you
run the allocation) and the candidate cell B belongs to a layer of cells with carrier c2.
SA is the area where:
• The pilot signal from A is not the highest one. It is strictly lower than the best pilot signal received and higher than
the best pilot signal minus the margin.
SB is the area where:
• The pilot signal from B is the highest one.
Figure 5.4: Overlapping Zones - 1st Case
2nd case: the reference cell A is located on the border of a layer with carrier c1 (c1 is the selected carrier on which you run the
allocation) and the candidate cell B belongs to a layer of cells with carrier c2.
SA is the area where:
• The pilot signal from A is the highest one
• The pilot signal from A is lower than the minimum pilot signal level plus the margin.
SB is the area where:
• The pilot signal from B is the highest one.
425
Figure 5.5: Overlapping Zones - 2nd Case
Two ways enable you to determine the I0 value:

• Global Value: A percentage of the cell maximum power is considered. If the % of
maximum power is too low, i.e. if %  Pmax  P pilot , Atoll takes into account the
pilot power of the cell. Then, I0 represents the sum of values calculated for each cell.
• Defined per Cell: Atoll takes into account the total downlink power defined per cell.
I0 represents the sum of total transmitted powers.
SA  SB
-  100 ) and compares this value to the % minimum covered area. If
SA
The coverage condition can be weighted among the others and ranks the neighbours through the importance field (see after).
allocated to each transmitter is exceeded. If we consider the case for which there are 15 candidate neighbours and the
maximum number of neighbours to be allocated to the reference cell is 8. Among these 15 candidate neighbours, only 8
(having the highest importance values) will be allocated to the reference cell. Note that specific maximum numbers of
neighbours (maximum number of intra-carrier neighbours, maximum number of inter-carrier neighbours) can be defined at
the cell level (property dialogue or cell table). If defined there, this value is taken into account instead of the default one
available in the Neighbour Allocation dialogue.
neighbour may be marked as exceptional pair, co-site, adjacent, coverage or symmetric. For neighbours accepted for co-site,
adjacency and coverage reasons, Atoll displays the percentage of area meeting the coverage conditions and the
corresponding surface area (km2), the percentage of area meeting the adjacency conditions and the corresponding surface
area (km2). Finally, if cells have previous allocations in the list, neighbours are marked as existing.
426
[Neighbours]
[Neighbours]
• No simulation or prediction study is needed to perform an automatic neighbour

allocation. When starting an automatic neighbour allocation, Atoll automatically
calculates the path loss matrices if not found.
• Even if no specific terminal, mobility or service is selected in the automatic allocation,
it is interesting to know that the algorithm works such as finding the maximum
number of neighbours by selection the multi-service traffic data as follows:
• Service: selection of the one with the lowest body loss.
• Mobility: no impact on the allocation, no specific selection.
• Terminal: selection of the one with the greatest (Gain - Loss) value, and, if equal, the
one with the lowest noise figure.
• The neighbour lists may be optionally used in the power control simulations to
determine the mobile's active set.
• The force neighbour symmetry option enables the users to consider the reciprocity of
a neighbourhood link. This reciprocity is allowed only if the neighbour list is not
already full. Thus, if the cell B is a neighbour of the cell A while the cell A is not a
neighbour of the cell B, two cases are possible:
• 1st case: There is space in the cell B neighbour list: the cell A will be added to the list.
It will be the last one.
• 2nd case: The cell B neighbour list is full: Atoll will not include cell A in the list and will
cancel the link by deleting cell B from the cell A neighbour list.
• When the options “Force exceptional pairs” and “Force symmetry” are selected, Atoll
considers the constraints between exceptional pairs in both directions so as to
respect symmetry condition. On the other hand, if neighbourhood relationship is
forced in one direction and forbidden in the other one, symmetry cannot be
respected. In this case, Atoll displays a warning in the Event viewer.

Transmitter
• TBA cells,
• Neighbours of TBA cells marked as exceptional pair, adjacent and symmetric,
• Neighbours of TBA cells that satisfy coverage conditions.
Automatic neighbour allocation parameters are described in "Neighbour Allocation for all Transmitters" on page 424.
427

Importance values are used by the allocation algorithm to rank the neighbours according to the allocation reason, and to
quantify the neighbour importance.
5.6.3.1 Importance of Intra-carrier Neighbours


selected
selected
relationship
transmitter.
d-
 Di  = 1 – ----------
d max
page 430.

Neighbourhood cause Resulting IF using the default values

Importance Function
Co-site Adjacent from the table above
No No 10%+20%{10%(Di)+90%(O)}+1%+9%(Di)
No Yes
Yes Yes
Where:
428
overlapping.
causes.
5.6.3.2 Importance of Inter-carrier Neighbours


Co-site cell If the Force co-site cells as neighbours option is selected Importance Function (IF)
If the Force neighbour symmetry option is selected Importance Function (IF)
relationship
transmitter.
d-
 Di  = 1 – ----------
d max
page 430.


Co-site Neighbourhood Resulting IF using the default
IF
cause values from the table above
Yes 60%+40%{1/7%(Di)+6/7%(O)}
Where
429
causes.
5.6.4 Appendices
5.6.4.1 Calculation of the Inter-Transmitter Distance
d = D   1 + x  cos  – x  cos  
5.7 PN Offset Allocation

PN offset is used to identify a cell. It is a time offset used by a cell to shift a Pseudo Noise sequence. Mobile processes the
strongest received PN sequence and reads its phase that identifies the cell. There are a maximum of 512 PN offsets numbered
from 0 to 511.
The cells to which Atoll allocates PN offsets are referred to as the TBA cells (cells to be allocated). TBA cells fulfil following
conditions:
• They belong to the folder on which allocation has been executed. This folder can be either the Transmitters folder
or a group of transmitters or a single transmitter.
zone.
430

5.7.1.1 Options and Constraints
The PN offset allocation algorithm can take into account following constraints and options:
• PILOT_INC parameter,
Atoll uses this parameter to determine the pool of possible PN offsets (512 divided by PILOT_INC value). The first PN
offset is PILOT_INC and other ones are multiples of this value.
For example: When PILOT_INC is set to 4, the pool of possible PN offsets consists of PN offsets from 4 to 508 with a
separation interval of 4 (i.e. [4,8,12,16,...508]).
• Neighbourhood between cells,
You may consider:
• In the context of the PN offset allocation, the term "neighbours" refers to intra-carrier
neighbours.
• In 3GPP2 multi-RAT documents, Atoll also tries to allocate different PN offsets to
CDMA cells that are neighbours of a common LTE cell.
• Cells fulfilling a criterion on Ec/I0 (option “Additional Ec/I0 conditions”),

Atoll reuses the intra-carrier neighbour allocation algorithm to determine the list of cells which cannot be allocated the same
scrambling code, and to calculate their importance.
For a reference cell “A”, Atoll considers all the cells “B” that can enter the active set on the area where the reference cell is
the best server (area where (Ec/I0)A exceeds Min. Ec/I0 and is the highest one and (Ec/I0)B exceeds T_Drop).
Atoll considers either a percentage of the cell maximum powers or the total downlink
power used by the cells in order to evaluate I0. In this case, I0 equals the sum of total
transmitted powers. When this parameter is not specified in the cell properties, Atoll
uses 50% of the maximum power.
• Co-PN Reuse distance,

Reuse distance is a constraint on the allocation of PN offsets. A PN offset cannot be reused at a site that is not at least as far
away as the reuse distance from the site allocated with the particular PN offset.
PN offset reuse distance can be defined at cell level. If this value is not defined, then Atoll
will use the default reuse distance defined in the PN offset Automatic Allocation
dialogue.
• PN-cluster size. Within the context of PN offset allocation, the term "PN-cluster" refers to a sub-group of PN offsets.
• Exceptional pairs,
• Domains of PN offsets,
When no domain is assigned to cells, Atoll considers the PILOT_INC parameter only to
determine available PN offsets (e.g., If PILOT_INC is set to 4, all PN offsets from 4 to 508
with a separation interval of 4 can be allocated).
• The carrier on which the allocation is run: It can be a given carrier or all of them. In this case, either Atoll independently
plans PN offsets for the different carriers, or it allocates the same PN offset to each carrier of a transmitter if the option
"Allocate carriers identically" is selected.
• The possibility to use a maximum of PN offsets (option "Use a Maximum of PN Offsets"): Atoll will try to spread the
PN offset spectrum the most.
• The "Delete All Codes" option: When selecting this option, Atoll deletes all the current PN offsets and carries out a
new PN offset allocation. If not selected, the existing PN offsets are kept.
431
In addition, it depends on the selected allocation strategy. Allocation strategies can be:
• PN offset per cell: The purpose of this strategy is to reduce the spectrum of allocated PN offsets the maximum
possible. Atoll will allocate the first possible PN offsets in the domain.
• Adjacent PN-Clusters per site: This strategy consists of allocating one cluster of adjacent PN offsets to each site, then,
one PN offset of the cluster to each cell of each transmitter according to its azimuth. When all the clusters have been
allocated and there are still sites remaining to be allocated, Atoll reuses the clusters at another site.
• Distributed PN-clusters per site: This strategy consists of allocating one cluster of PN offsets to each site in the
network, then, one PN offset of the cluster to each cell of each transmitter according to its azimuth. With this strategy,
the cluster is made of PN offsets separated as much as possible. When all the clusters have been allocated and there
are still sites remaining to be allocated, Atoll reuses the clusters at another site.
In the Results table, Atoll only displays PN offsets allocated to TBA cells.

TBA cell may be:
"First Order"),
• The cells that fulfil Ec/I0 condition (option “Additional Ec/I0 conditions”),
One additional constraint is considered in 3GPP2 multi-RAT documents:
• The cell and its near cells are neighbours of the same LTE cell.
and the cost of the PN offset plan. During the allocation, Atoll tries to assign different PN offsets to the TBA cell and its near
cells. If it respects all the constraints, the cost of the PN offset plan is 0. When a cell has too many constraints and there are
not anymore PN offsets available, Atoll breaks the constraint with the lowest cost so as to generate the PN offset plan with
the lowest cost. For information on the cost generated by each constraint, see "Cell Priority" on page 433.

Strategy: PN offset per cell
Atoll processes TBA cells according to their priority. It allocates PN offsets starting with the highest priority cell and its near
cells, and continuing with the lowest priority cells not allocated yet and their near cells. For information on calculating cell
priority, see "Cell Priority" on page 433.
Strategy: Adjacent PN-Clusters per site
Atoll assigns a PN-cluster of adjacent PN offsets to each site, starting with the highest priority site and its near sites, and
continuing with the lowest priority sites not allocated yet and their near sites. When all the clusters have been allocated and
there are still sites remaining to be allocated, Atoll reuses the clusters at another site. When the Co-PN Reuse Distance option
is selected, the algorithm reuses the clusters as soon as the Co-PN reuse distance is exceeded. Otherwise, when the option is
not selected, the algorithm tries to assign reused clusters as spaced out as possible.
Then, Atoll allocates a PN offset from the cluster to each cell of each transmitter located on the sites according to the
transmitter azimuth. It starts with the highest priority cell and its near cells and goes on with the lowest priority cells not
allocated yet and their near cells.
Strategy: Distributed PN-Clusters per site
Atoll assigns one cluster to each site, starting with the highest priority site and its near sites, and continuing with the lowest
priority sites not allocated yet and their near sites. When all the clusters have been allocated and there are still sites remaining
to be allocated, Atoll reuses the clusters at another site. When the Co-PN Reuse Distance option is selected, the algorithm
reuses the clusters as soon as the Co-PN reuse distance is exceeded. Otherwise, when the option is not selected, the algorithm
tries to assign reused clusters as spaced out as possible.
432
Then, Atoll assigns a PN offset from the cluster to each cell of each transmitter located on the sites according to the
transmitter azimuth. It starts with the highest priority cell and its near cells and goes on with the lowest priority cells not
allocated yet and their near cells.

In case you have a multi-carrier network and you run the PN offset allocation on all the carriers, the allocation process
depends on wether the option "Allocate Carriers Identically" is selected or not.
When the option is not selected, algorithm works for each strategy, as explained above. On the other hand, when the option
is selected, allocation order changes. It is no longer based on the cell priority but depends on the transmitter priority. All
transmitters which have constraints with the studied transmitter will be referred to as near transmitters.
In case of a "Per cell" strategy (PN offset per cell), Atoll starts PN offset allocation with the highest priority transmitter and its
near transmitters and continues with the lowest priority transmitters not allocated yet and their near transmitters. The same
PN offset is assigned to each cell of the transmitter.
In case of a "Per site" strategy (Adjacent and Distributed PN-clusters per site strategies), Atoll assigns a cluster to each site
and then, allocates a PN offset to each transmitter. It starts with the highest priority transmitter and its near transmitters and
continues with the lowest priority transmitters not allocated yet and their near transmitters. The same PN offset is assigned
to each cell of the transmitter.
For information on calculating cell priority, see "Cell Priority" on page 433. For information on calculating transmitter priority,
see "Transmitter Priority" on page 435.
alphanumeric order.
5.7.1.2.3 Difference between Adjacent and Distributed PN-Clusters

The following example explains the difference between "Adjacent PN-clusters" and "Distributed PN-clusters". The PILOT_INC
has been set to 4 and the PN-cluster size to 3. There are:
• 128 PN offsets that can be allocated: they are all PN offsets from 4 to 508 with a separation interval of 4.
• Each PN-cluster consists of three PN offsets. So, there are 42 PN-clusters available.
If you select "Adjacent PN-cluster per site" as allocation strategy, Atoll will consider PN-clusters consisted of adjacent PN
offsets (e.g. {4,8,12}, {16,20,24}, ...,{496,500,504}).
If you select "Distributed PN-cluster per site" as allocation strategy, Atoll will consider PN-clusters consisted of PN offsets
separated as much as possible (e.g. {4,172,340}, {8,176,344}, ...,{168,336,504}).

PN offset allocation algorithm in Atoll allots priorities to cells before performing the actual allocation. Priorities assigned to
cells depend upon how much constrained each cell is and the cost defined for each constraint. A cell without any constraint
has a default cost, C , equal to 0. The higher the cost on a cell, the higher the priority it has for the PN offset allocation process.
There are five criteria employed to determine the cell priority:
PN Offset Domain Criterion
The cost due to the domain constraint, C i  Dom  , depends on the number of PN offsets available for the allocation. The
domain constraint is mandatory and cannot be broken.
When no domain is assigned to cells, 512 PN offsets are available and we have:
C i  Dom  = 0
When domains of PN offsets are assigned to cells, each unavailable PN offset generates a cost. The higher the number of codes
available in the domain, the less will be the cost due to this criterion. The cost is given as:
C i  Dom  = 512 – Number of PN Offsets in the domain
433
Distance Criterion
C i  Dist  =  Cj  Dist  i  
j
Where

C i  EP  =  cEP  i – j 
j
Where
Neighbourhood Criterion

     
Ci  N  = 

C j  N1  i   = I j  c N1
Where
Because two first order neighbours must not have the same PN offset, Atoll considers the cost created by two first order
neighbours to be each other.
C j  N1  i   + C j  N1  i  
C j – j  N1  i   = -----------------------------------------------------
-
2
434
Where
Because two second order neighbours must not have the same PN offset, Atoll considers the cost created by two second order
C k  N2  i   + C k  N2  i  
C k – k  N2  i   = -------------------------------------------------------
-
2
C l  N3  i   = Max  j   c N3
Where
Because two third order neighbours must not have the same PN offset, Atoll considers the cost created by two third order
C l  N3  i   + C l  N3  i  
C l – l  N3  i   = -----------------------------------------------------
-
2
.
And
LTE Neighbour Criterion
This criterion is considered in 3GPP2 multi-RAT documents. If the cell i is neighbour of an LTE cell, the cell constraint level
depends on how many cells j are neighbours of the same LTE cell. The total cost due to LTE neighbour constraint is given as:
C i  N LTE  =  cNLTE  j – TxLTE 

j
Where
cN is the cost of the LTE neighbour constraint. This value can be defined in the Constraint Cost dialogue.
LTE
Therefore, the total cost due to constraints on any cell i is defined as:
C i = C i  Dom  + C i  U 
With
C i  U  = C i  Dist  + C i  EP  + C i  N  + C i  N 2G 

In case you have a multi-carrier network and you run PN offset allocation on "all" the carriers with the option "allocate carriers
identically", algorithm in Atoll allots priorities to transmitters. Priorities assigned to transmitters depend on how much
constrained each transmitter is and the cost defined for each constraint. The higher the cost on a transmitter, the higher the
priority it has for the PN offset allocation process.
435
With C Tx  U  = Max  C  U   and C  Dom  = 512 – Number of PN offsets in the domain

i  Tx i Tx

In case of "Per Site" allocation strategies (Adjacent PN-clusters per site and Distributed PN-clusters per site), algorithm in Atoll
allots priorities to sites. Priorities assigned to sites depend on how much constrained each site is. The higher the constraint on
a site, the higher the priority it has for the PN offset allocation process.
Let us consider a site S with three transmitters; each of them has two cells using carriers 0 and 1. The site constraint is given as:
C S = C S  U  + C S  Dom 
With, C S  U  =  CTx  U  , and CS  Dom  = 512 – Number of PN offsets in the domain

Tx
Here, the domain considered for the site is the intersection of domains available for transmitters of the site.
5.7.2 Allocation Examples

In order to understand the differences between the different allocation strategies and the behaviour of the algorithm when
using a maximum of PN offsets or not, let us consider the following sample scenario:
Figure 5.8: PN Offset Allocation
Let Site0, Site1, Site2 and Site3 be four sites with 3 cells using carrier 0 whom PN offsets have to be allocated. The PILOT_INC
parameter has been set to 4 and the PN Cluster Size is 3. Therefore, all PN offsets from 4 to 508 with a separation interval of
4 can be allocated. The reuse distance is supposed to be lower than the inter-site distance. Only co-site neighbours exist and
all of them have the same importance.
The following section lists the results of each combination of options with explanation where necessary.
5.7.2.1 Strategy: PN Offset per Cell

Since the restrictions of neighbourhood only apply to co-sites with the same importance and inter-site distances are greater
than reuse distances, every cell has the same priority. Then, the PN offset allocation to cells is performed in an alphanumeric
order.
436
Without ‘Use a Maximum of PN Offsets’ With ‘Use a Maximum of PN Offsets’
Atoll allocates the first three PN offsets in the domain (4, 8 Atoll allocates the first three PN offsets in the domain (4, 8
and 12) to the Site0’s cells. Under given constraints of and 12) to the Site0’s cells. As it is allowed to use a maximum
neighbourhood and reuse distance, same PN offsets can be of PN offsets, Atoll allocates different PN offsets to each
allocated to each site’s cells. site’s cells so that there is least repetition.
5.7.2.2 Strategy: Adjacent PN-Clusters Per Site

order.
Atoll allocates a PN cluster of adjacent PN offsets to Site0 and As it is possible to use a maximum of PN offsets, Atoll
then, one PN offset of the PN cluster to each cell. Under given allocates different PN clusters of adjacent PN offsets to sites
constraints of neighbourhood and reuse distance, the same so that there is least repetition of PN offsets.
PN cluster can be allocated to each site and same PN offsets
to each site’s cells.
5.7.2.3 Strategy: ‘Distributed PN-Clusters Per Site

order.
437
Atoll allocates a PN cluster of distributed PN offsets to Site0 As it is possible to use a maximum of PN offsets, Atoll
and then, one PN offset of the PN cluster to each cell. Under allocates different PN clusters of distributed PN offsets to
given constraints of neighbourhood and reuse distance, the sites so that there is least repetition of PN offsets.
same PN cluster can be allocated to each site and same PN
offsets to each site’s cells.
5.8 Automatic GSM-CDMA Neighbour Allocation

5.8.1 Overview
You can automatically calculate and allocate neighbours between GSM/TDMA and CDMA2000 networks. In Atoll, it is called
inter-technology neighbour allocation.
• When the CDMA coverage is not continuous. In this case, the CDMA coverage is extended by CDMA-GSM handover
into the GSM network,
• And in order to balance traffic and service distribution between both networks.
Note that the automatic inter-technology neighbour allocation algorithm takes into account both cases.
• An .atl document containing the GSM/TDMA network, GSM.atl, and another one containing the CDMA2000 network,
CDMA.atl,
• An existing link on the Transmitters folder of GSM.atl into CDMA.atl.
transmitters of GSM.atl are potential neighbours. The cells to be allocated will be called TBA cells which, being cells of
CDMA.atl, fulfill following conditions:
• They satisfy the filter criteria applied to Transmitters folder,
a group of transmitters subfolder.
Only CDMA TBA cells may be assigned neighbours.

• The inter-transmitter distance,
• The maximum number of neighbours fixed,
• Allocation options,
• The selected allocation strategy,
We assume we have a CDMA reference cell, A, and a GSM candidate neighbour, transmitter B.
438

When the automatic allocation starts, Atoll checks the following conditions:
• The distance between the CDMA reference cell and the GSM neighbour must be less than the user-definable
maximum inter-site distance. If the distance between the CDMA reference cell and the GSM neighbour is greater than
this value, then the candidate neighbour is discarded.
Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. You may choose one or
more carriers. Atoll will allocate neighbours to cells using the selected carriers.
reference CDMA cell in the candidate neighbour list. This option is automatically selected.
force/forbid a GSM transmitter to be candidate neighbour of the reference CDMA cell.
allocated to each cell is exceeded. If we consider the case for which there are 15 candidate neighbours and the maximum
number of neighbours to be allocated to the reference cell is 8. Among these 15 candidate neighbours, only 8 (having the
highest importance values) will be allocated to the reference cell. Note that the maximum number of inter-technology
neighbours can be defined at the cell level (property dialogue or cell table). If defined there, this value is taken into account
Where d is the effective distance between the CDMA reference cell and the GSM neighbour and d max is the maximum inter-
site distance.
as existing.

• The distance between the CDMA reference cell and the GSM neighbour must be less than the user-definable
maximum inter-site distance. If the distance between the CDMA reference cell and the GSM neighbour is greater than
this value, then the candidate neighbour is discarded.
Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. You may choose one or
more carriers. Atoll will allocate neighbours to cells using the selected carriers.
439
reference CDMA cell in the candidate neighbour list. This option is automatically selected.
force/forbid a GSM transmitter to be candidate neighbour of the reference CDMA cell.
• There must be an overlapping zone ( S A  S B ) with a given cell edge coverage probability where:
Four different cases may be considered for SA:
• 1st case: SA is the area where the cell A is the best serving cell of the CDMA network.
• The pilot quality from A exceeds a user-definable minimum value (minimum Ec/I0) and is the highest one.
In this case, the Ec/I0 margin must be equal to 0dB and the max Ec/I0 option disabled.
• 2nd case: SA represents the area where the pilot quality from the cell A strats decreasing but the cell A is still the
best serving cell of the CDMA network.
The Ec/I0 margin must be equal to 0dB, the max Ec/I0 option selected and a maximum Ec/I0 user-defined.
• The pilot quality from A exceeds the minimum Ec/I0 but is lower than the maximum Ec/I0.
• The pilot quality from A is the highest one.
• 3rd case: SA represents the area where the cell A is not the best serving cell but can enter the active set.
Here, the Ec/I0 margin has to be different from 0dB and the max Ec/I0 option disabled.
• The pilot quality from A is within a margin from the best Ec/I0, where the best Ec/I0 exceeds the minimum Ec/
I0.
• 4th case: SA represents the area where:
• The pilot quality from A is within a margin from the best Ec/I0 (where the best Ec/I0 exceeds the minimum Ec/
I0) and lower than the maximum Ec/I0.
In this case, the margin must be different from 0dB, the max Ec/I0 option selected and a maximum Ec/I0 user-
defined.
• 1st case: SB is the area where the cell B is the best serving cell of the GSM network.
In this case, the margin must be set to 0dB.
highest one.
• 2nd case: The margin is different from 0dB and SB is the area where:
SA  SB
Atoll calculates the percentage of covered area ( ------------------  100 ) and compares this value to the % minimum covered area. If
SA
Candidate neighbours fulfilling coverage conditions are sorted in descending order with respect to percentage of covered
area.
When the automatic allocation is based on coverage overlapping, we recommend you to

perform two successive automatic allocations:
• A first allocation in order to find handovers due to non-continuous CDMA coverage.
In this case, you have to select the max Ec/I0 option and define a high enough value.
• A second allocation in order to complete the previous list with handovers motivated
for reasons of traffic and service distribution. Here, the max Ec/I0 option must be
disabled.
440
Importance values are used by the allocation algorithm to rank the neighbours according to the distance and the allocation
reason. Atoll lists all neighbours and sorts them by importance value so as to eliminate some of them from the neighbour list
if the maximum number of neighbours to be allocated to each cell is exceeded. If we consider the case for which there are 15
candidate neighbours, only 8 (having the highest importance values) will be allocated to the reference cell. Note that the
maximum number of inter-technology neighbours can be defined at the cell level (property dialogue or cell table). If defined
there, this value is taken into account instead of the default one available in the Neighbour Allocation dialogue.
As indicated in the table below, the neighbour importance depends on the neighbourhood cause; this value varies between
0 to 100%.
Co-site transmitter If the Force co-site cells as neighbours option is selected IF
If the % minimum covered area is exceeded IF
transmitter.
d-
 Di  = 1 – ----------
d max
page 430.


IF
Yes 60%+40%{1/7%(Di)+6/7%(O)}
Where
causes.
441

5.8.2.3 Delete Existing Neighbours Option

checked. We assume that we have an existing allocation of inter-technology neighbours.
A new TBA cell i is created in CDMA.atl. Therefore, if you start a new allocation without selecting the Delete existing
neighbours option, Atoll determines the neighbour list of the cell i,
If you change some allocation criteria (e.g. increase the maximum number of neighbours or create a new GSM TBC
transmitter) and start a new allocation without selecting the Delete existing neighbours option, it examines the neighbour list
of TBA cells and checks allocation criteria if there is space in their neighbour lists. A new GSM TBC transmitter can enter the
TBA cell neighbour list if allocation criteria are satisfied. It will be the first one in the neighbour list.
442
Chapter 6
LTE Networks

• "Calculation Quick Reference" on page 450
• "Available Calculations" on page 470
• "Calculation Details" on page 485
• "Automatic Planning Algorithms" on page 563
© Forsk 2015
444
AT330_TRR_E1 Chapter 6: LTE Networks
6 LTE Networks
This chapter describes all the calculations performed in Atoll LTE documents. The first part of this chapter lists all the input
parameters in the LTE documents, their significance, location in the Atoll GUI, and their usage. It also contains the lists of the
formulas used for the calculations.
The second part describes all the calculation processes, i.e., signal level coverage predictions, point analysis calculations,
signal quality coverage predictions, calculations on subscriber lists, and Monte Carlo simulations. The calculation algorithms
used by these calculation processes are available in the next part.
The third part describes all the calculation algorithms used in all the calculations. These algorithms include the calculation of
signal levels, noise, and interference for downlink and uplink considering power control, MIMO, smart antennas, and the radio
resource management algorithms used by the different available schedulers.
If you are new to LTE, you can also see the Glossary of LTE Terms in the User Manual for information on LTE terms and
concepts, especially in the context of their user in Atoll.
• A cell refers to a transmitter-carrier (TX-c) pair. The cell being studied during a
calculation is referred to as TXi(ic) in this chapter.
• All the calculation algorithms in this section are described for two types of cells:
• A studied cell (represented by the subscript "i") comprising the studied transmitter
TXi and its carrier ic. It is the cell which is currently the focus of the calculation. For
example, a victim cell when calculating the interference it is receiving from other
cells.
• Other cells (represented by the subscript "j") comprising the other transmitter TXj
and its carrier jc. The other cells in the network can be interfering cells (downlink)
or the serving cells of interfering mobiles (uplink).
• All the calculation algorithms in this section are described for two types of receivers:
• Mi: A pixel (coverage predictions), subscriber (calculations on subscriber lists), or
mobile (Monte Carlo simulations) covered/served by the studied cell TXi(ic).
• Mj: A mobile (Monte Carlo simulations) covered/served by any other cell TXj(jc).
6.1 Definitions
This table lists the input to calculations, coverage predictions, and simulations.
D Frame 3GPP parameter (Fixed to 10 ms in Atoll) ms Frame duration
W FB 3GPP parameter (Fixed to 180 kHz in Atoll) kHz Width of a resource/frequency block
F 3GPP parameter (Fixed to 15 kHz in Atoll) kHz Subcarrier width

Number of frequency blocks for SS
N FB – SS PBCH 3GPP parameter (Fixed to 6 in Atoll) None
and PBCH transmission
N SF  Frame 3GPP parameter (Fixed to 10 in Atoll) None Number of subframes per frame
N Slots  SF 3GPP parameter (Fixed to 2 in Atoll) None Number of slots per subframe
K 1.38 x 10-23 J/K Boltzmann’s constant

Power spectral density of thermal
n0 Calculation result ( 10  Log  K  T  1000  = – 174 dBm/Hz ) dBm/Hz
noise
D CP Frame configuration or, otherwise, global parameter None Cyclic prefix duration
Number of PDCCH symbol durations
N SD – PDCCH Frame configuration or, otherwise, global parameter SD
per subframe
Average number of PUCCH
N FB – PUCCH Frame configuration or, otherwise, global parameter RB
frequency blocks per frame
445
Chapter 6: LTE Networks ©Forsk 2015

Uplink power control adjustment
M PC Global parameter dB
margin
Minimum signal to thermal noise
CNR Min Global parametera dB
threshold (interferer cutoff)
W Channel Frequency band parameter MHz Channel bandwidth
First First channel number of the

N Channel Frequency band parameter None
frequency band
Last Last channel number of the
frequency band
Start frequency of the TDD frequency
F Start – TDD Frequency band parameter MHz
band
DL start frequency of the FDD
F Start – FDD – DL Frequency band parameter MHz
frequency band
UL start frequency of the FDD
F Start – FDD – UL Frequency band parameter MHz
frequency band
F Sampling Frequency band parameter MHz Sampling frequency
f ACS Frequency band parameter dB Adjacent channel suppression factor

Number of frequency blocks per
N FB Frequency band parameter None
channel bandwidth
ICS Band Frequency band parameter MHz Inter-channel spacing
CN Band Frequency band parameter None Channel number step
F Sampling
N SCa – Total Calculation result ( N SCa – Total = --------------------
-) None Total number of subcarriers
F
N FB  W FB
N SCa – Used Calculation result ( N SCa – Used = ------------------------
-) None Number of used subcarriers
F
N SCa – DC Hard-coded parameter ( N SCa – DC = 1 ) None Number of DC subcarriers

Calculation result
N SCa – Guard None Number of guard subcarriers
( N SCa – Guard = N SCa – Total – N SCa – Used – N SCa – DC )
i Layer Layer parameter None Layer index
p Layer Layer parameter None Layer priority
Max Maximum mobile speed supported

V Layer Layer parameter km/h
by a layer
 CE Frame configuration parameter dB Cell-edge power boost
Number of cell-edge frequency
N FB – CE0 Frame configuration parameter None
blocks for PSS ID 0
blocks for PSS ID 1
blocks for PSS ID 2
B Bearer parameter None Bearer index
Mod B Bearer parameter None Modulation used by the bearer
CR B Bearer parameter None Coding rate of the bearer

bits/
B Bearer parameter Bearer efficiency
symbol
TB Bearer parameter dB Bearer selection threshold
Site Maximum S1 interface site downlink

TP S1 – DL Site parameter kbps
throughput
446
Site Maximum S1 interface site uplink

TP S1 – UL Site parameter kbps
throughput
TX Transmitter parameter (user-defined or calculated from transmitter
nf dB Transmitter noise figure
Number of antenna ports used for
N Ant – TX Transmitter parameter None
transmission
Number of antenna ports used for
N Ant – RX Transmitter parameter None
reception
TX Transmitter antenna parameter dB Antenna gain
G

L dB Transmitter loss
TX
E SA Smart antenna parameter None Number of smart antenna elements
Array
G SA Smart antenna parameter dB Array gain offset
Combining
G SA Smart antenna parameter dB Power combining gain offset
Div
G SA Smart antenna parameter dB Diversity gain (cross-polarisation)
N Channel Cell parameter None Cell’s channel number
ID  Cell parameter None Cell’s physical ID
ID  Cell’s SSS ID (one of 168 pseudo-

ID SSS Cell parameter: Floor  -------- None
 3  random sequences)
Cell’s PSS ID (one of 3 cyclic shifts of

ID PSS Cell parameter: ID  Mod 3 None
the sequence given by the SSS ID)
Cell’s v shift (also known as the
 Shift Cell parameter: ID  Mod 6 None
reference signal hopping index)
P Max Cell parameter dBm Maximum cell transmission power
Energy per resource element for the
EPRE DLRS Cell parameter dBm downlink reference signals
(User-defined or calculated)
Energy per resource element offset
EPRE SS Cell parameter dB for the SS with respect to the
downlink reference signal EPRE
EPRE PBCH Cell parameter dB for the PBCH with respect to the
EPRE PDCCH Cell parameter dB for the PDCCH with respect to the
EPRE PDSCH Cell parameter dB for the PDSCH with respect to the
T RSRP Cell parameter dB Minimum Required RSRP
TX i  ic 
T Selection Cell parameter dB Cell selection threshold
TX  ic 
i Cell parameter dB Cell individual offset
O Individual
TX i  ic 
M HO Cell parameter dB Handover margin
TL DL Cell parameter % Downlink traffic load
r DL – CE Cell parameter % Downlink cell-edge traffic ratio
447
TL UL Cell parameter % Uplink traffic load
TL DL – Max Cell parameter % Maximum downlink traffic load
TL UL – Max Cell parameter % Maximum uplink traffic load
NR UL Cell parameter dB Uplink noise rise
NRUL – ICIC Cell parameter dB ICIC uplink noise rise
NR UL – Max Cell parameter dB Maximum uplink noise rise
N Users – Max Cell parameter None Maximum number of users per cell
Number of users connected to the
N Users – DL Cell parameter None
cell in downlink
N Users – UL Cell parameter None
cell in uplink
L Path Cell parameter dB Delta path loss threshold
Number of downlink subframes per
N SF – DL Cell parameter None
frame
Number of uplink subframes per
N SF – UL Cell parameter None
frame
TX i  ic  Number of TDD special subframes
N TDD – SSF Cell parameter None
per frame
Channel and physical cell ID reuse
D Reuse Cell parameter m
distance
Average number of co-scheduled
G MU – MIMO – DL Cell parameter None
MU-MIMO users in downlink
Average number of co-scheduled
G MU – MIMO – UL Cell parameter None
MU-MIMO users in uplink
 FPC Cell parameter None Fractional power control factor
CINR PUSCH – Max Cell parameter dB Maximum PUSCH C/(I+N)

Inter – Tech
NR DL Cell parameter dB Inter-technology downlink noise rise
Inter – Tech
NR UL Cell parameter dB Inter-technology uplink noise rise
AU DL Cell parameter % Downlink AAS usage ratio

TX i  ic  Downlink multi-user diversity gain
G MUG – DL Proportional Fair scheduler parameter None
(MUG)
TX i  ic  Uplink multi-user diversity gain
G MUG – UL Proportional Fair scheduler parameter None
(MUG)
Max Maximum C/(I+N) above which no
CINR MUG Proportional Fair scheduler parameter dB
MUG gain is applied
T SU – MIMO – UL Cell reception equipment parameter dB Uplink SU-MIMO threshold
T MU – MIMO – UL Cell reception equipment parameter dB Uplink MU-MIMO threshold

Max
G SU – MIMO – UL Cell reception equipment parameter None Maximum uplink SU-MIMO gain
Receive, SU-MIMO, or MU-MIMO

G Div – UL Cell reception equipment parameter dB
diversity gain
UL Uplink secondary cell activation
T SCell Cell reception equipment parameter dB
threshold
QoS class identifier (QCI) of the
QCI Service parameter None
service
p QCI Service parameter (automatically determined from the QCI) None Service’s QCI priority
448
p Service Service parameter None User-defined service priority

Highest bearer used by a service in
B DL – Highest Service parameter None
the downlink
B UL – Highest Service parameter None
the uplink
Lowest bearer used by a service in
B DL – Lowest Service parameter None
the downlink
B UL – Lowest Service parameter None
the uplink
UL
f Act Service parameter % Uplink activity factor
DL
f Act Service parameter % Downlink activity factor
Minimum throughput demand in the

TPD Min – UL Service parameter kbps
uplink (Guaranteed Bit Rate, GBR)
TPD Min – DL Service parameter kbps
downlink (Guaranteed Bit Rate, GBR)
Maximum throughput demand in the
TPD Max – UL Service parameter kbps
uplink (Maximum Bit Rate, MBR)
TPD Max – DL Service parameter kbps
downlink (Maximum Bit Rate, MBR)
UL Average requested throughput in the
TP Average Service parameter kbps
uplink
DL Average requested throughput in the
downlink
TP Offset Service parameter kbps Throughput offset
L Body Service parameter dB Body loss
Min Minimum number of frequency

N FB – UL Service parameter None
blocks
P Min Terminal parameter dBm Minimum terminal power
P Max Terminal parameter dBm Maximum terminal power
nf Terminal parameter dB Terminal noise figure
G Terminal parameter dB Terminal antenna gain
L Terminal parameter dB Terminal loss

Number of antenna ports for
N Ant – TX Terminal parameter None
transmission
Number of antenna ports for
N Ant – RX Terminal parameter None
reception
Max – DL Maximum number of downlink
N SCell Terminal parameter None
secondary cells
Max – UL Maximum number of uplink
N SCell Terminal parameter None
secondary cells
Max – DL Maximum number of transport block
N TBB  TTI UE category parameter Bits
bits per TTI (subframe) in downlink
Max – UL Maximum number of transport block
N TBB  TTI UE category parameter Bits
bits per TTI (subframe) in uplink
Max – UL Highest modulation supported in
Mod UE UE category parameter None
uplink
Max – DL Maximum number of reception
N Ant – UE UE category parameter None
antenna ports supported in downlink
449
T SU – MIMO – DL Terminal reception equipment parameter dB SU-MIMO threshold

Max
G SU – MIMO – DL Terminal reception equipment parameter None Maximum downlink SU-MIMO gain
T MU – MIMO – DL Terminal reception equipment parameter dB Downlink MU-MIMO threshold
G Div – PBCH Terminal reception equipment parameter dB PBCH diversity gain
G Div – PDCCH Terminal reception equipment parameter dB PDCCH diversity gain

Transmit, SU-MIMO, or MU-MIMO
G Div – DL Terminal reception equipment parameter dB
diversity gain
DL Downlink secondary cell activation
T SCell Terminal reception equipment parameter dB
threshold
DL
T AAS Terminal reception equipment parameter dB Downlink AAS threshold
UL
G Div Clutter parameter dB Additional uplink diversity gain
DL
G Div Clutter parameter dB Additional downlink diversity gain
f SU – MIMO Clutter parameter None SU-MIMO gain factor
L Indoor Clutter parameter dB Indoor loss
L Path Propagation model result dB Path loss

Inter-technology downlink channel
F protection ratio for a frequency
ICPDL Network parameter None
offset F between the interfered
and interfering frequency channels
Monte Carlo simulations: Random result calculated from model
standard deviation
M Shadowing – Model dB Model shadowing margin
Coverage Predictions: Result calculated from cell edge coverage
probability and model standard deviation
M Shadowing – C  I dB C/I shadowing margin
probability and C/I standard deviation
a. Any interfering cell whose signal to thermal noise ratio is less than CNR Min will be discarded.
6.2 Calculation Quick Reference

The following tables list the formulas used in calculations.
6.2.1 Downlink Transmission Powers Calculation

Number of symbols per scheduler
N Sym  SRB N SCa – FB  N SD  Slot  N Slot  SF None
resource block
Number of DwPTS modulation
DwPTS DwPTS
N Sym  SSF N SCa – FB  N SD  SSF None symbols per scheduler resource
block in the TDD special subframes
W FB Number of subcarriers per
N SCa – FB ---------- None
F frequency block
TX  ic  TX  ic  TX  ic  TX  ic  TX  ic  Total number of symbols in
i i i i i DwPTS None
N Sym – DL N FB  N Sym  SRB  N SF – DL + N FB  N TDD – SSF  N Sym  SSF downlink
450

 TX  ic 
if  N Ant – TX = 1
i
 8  
 Number of symbols reserved for
TX  ic   TX  ic 
 16 if  N Ant – TX = 2
i i None downlink reference signals in one
N Res  SRB
 scheduler resource block

 TX  ic 
if  N Ant – TX = 4 or 8
i
 24  

Number of symbols reserved for

TX i  ic  downlink reference signals in
N Res  DwPTS See "Downlink Transmission Power Calculation" on page 485 None
DwPTS of one TDD special
subframe
TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic 
N Sym – Res N SF – DL  N FB  N Res  SRB + N TDD – SSF  N FB  N Res  DwPTS None downlink reference signals in one
frame

 TX  ic 
i
 8  
 Number of symbols for downlink
TX i  ic   TX  ic 
 8 if  N Ant – TX = 2 None reference signals in one scheduler
i
N DLRS  SRB
 resource block

 TX  ic 
i
 6  

Number of symbols for downlink

TX i  ic 
N DLRS  DwPTS See "Downlink Transmission Power Calculation" on page 485 None reference signals in DwPTS of one
TDD special subframe
TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  Number of symbols for downlink
N Sym – DLRS N SF – DL  N FB  N DLRS  SRB + N TDD – SSF  N FB  N DLRS  DwPTS None
reference signals in one frame
N Sym – PSS + N Sym – SSS = 288
Number of symbols for the PSS and
N Sym – SS Where N Sym – PSS = 2  N FB – SS PBCH  N SCa – FB = 144 None
the SSS
N Sym – SSS = 2  N FB – SS PBCH  N SCa – FB = 144
TX i  ic  Extended CP: 216

N Sym – PBCH None Number of symbols for the PBCH
Normal CP: 240
TX  ic 
if  N SD – PDCCH = 0 : 0
i
 
TX  ic  TX  ic 
if  N SD – PDCCH = 1 AND  N Ant – TX  2 :
i i
TX  ic  TX  ic  TX i  ic 
N i  i
 SD – PDCCH  N SCa – FB – 4  N FB  N SF – DL
TX  ic 
i None Number of symbols for the PDCCH
N Sym – PDCCH TX  ic  TX  ic  TX i  ic 
+  NSD – PDCCH  N SCa – FB – 4  N FB
i i
 N TDD – SSF
 
Otherwise:
TX  ic  TX  ic  TX  ic  TX i  ic 
N i  N SCa – FB – 2  Min  4 N Ant – TX   N FB
i i
 N SF – DL
 SD – PDCCH  
+  Min  2 N SD – PDCCH  N SCa – FB – 2  Min  4 N Ant – TX   N FB
i i i
 N TDD – SSF
    
TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic 
N Sym – PDSCH N Sym – DL – N Sym – Res – N Sym – SS – N Sym – PBCH – N Sym – PDCCH None Number of symbols for the PDSCH
451
TX  ic 
 P i 
 --------------------
Max
- TX  ic  TX  ic  
10  Log  10   N SD  Slot  N Slot  SF  N SF – DL + N TDD – SSF  N SD  SSF  –
10 i i DwPTS
 
  Energy per resource element for 1
modulation symbol (dBm/Sym) of
 EPRE SS
i
EPRE PBCH
i the downlink reference signals
TX  ic   TX  ic  ----------------------------------- -----------------------------------
10  L og  N Sym – DLRS + N Sym – SS  10
i i 10 10
EPRE DLRS + N Sym – PBCH  10 dBm/Sym
 With reference signal EPRE
 calculation method is set to

Calculated (equal distribution of
TX i  ic  TX i  ic  unused EPRE)
EPRE EPRE 
------------------------------------ 
PDCCH PDSCH
------------------------------------
+ N Sym – PDCCH  10
10
+ N Sym – PDSCH  10
10 



TX  ic 
 P i 
 --------------------
Max
- TX i  ic  TX i  ic  
10  Log  10   N SD  Slot  N Slot  SF  N SF – DL + N TDD – SSF  N SD  SSF  –
10 DwPTS
  
 
  Energy per resource element for 1
modulation symbol (dBm/Sym) of
 EPRE
i
EPRE
i the downlink reference signals
TX  ic   TX  ic  SS
----------------------------------
- PBCH -
----------------------------------
10  L og  N Sym – Res + N Sym – SS  10
i i 10 10
EPRE DLRS + N Sym – PBCH  10 dBm/Sym
 With reference signal EPRE
 calculation method is set to

Calculated (with boost) or
TX i  ic  TX i  ic  Calculated (without boost)
EPRE EPRE 
------------------------------------ 
PDCCH PDSCH
------------------------------------
10
10 



Energy per resource element for 1
TX i  ic  TX i  ic  TX i  ic 
EPRE SS EPRE DLRS + EPRE SS dBm/Sym modulation symbol (dBm/Sym) of
the SS
TX  ic  TX  ic  TX  ic 
i i i dBm/Sym modulation symbol (dBm/Sym) of
EPRE PBCH EPRE DLRS + EPRE PBCH
the PBCH
EPRE PDCCH EPRE DLRS + EPRE PDCCH dBm/Sym modulation symbol (dBm/Sym) of
the PDCCH
EPRE PDSCH EPRE DLRS + EPRE PDSCH dBm/Sym modulation symbol (dBm/Sym) of
the PDSCH
"Boosted" energy per resource
TX i  ic  element for 1 modulation symbol
TX i  ic  N 
EPRE DLRS + 10  Log  -------------------------
TX  ic  Sym – Res  (dBm/Sym) of downlink reference
i - dbm/Sym
EPRE DLRS  TXi  ic   signals when the reference signal
 N Sym – DLRS EPRE calculation method is set to
Calculated (with boost)
TX i  ic  TX  ic  TX  ic  Instantaneous transmission power
EPRE DLRS + 10  Log  2  N FB 
i i dBm
P DLRS  of the downlink reference signals
TX i  ic  TX i  ic  Instantaneous transmission power
P SS EPRE SS + 10  Log  N SCa – FB  N FB – SS PBCH  dBm
of the SS
TX  ic  TX  ic  Instantaneous transmission power
i i dBm
P PBCH EPRE PBCH + 10  Log  N SCa – FB  N FB – SS PBCH  of the PBCH
452
 TX  ic 
i 
TX  ic 
TX  ic 
i  N Sym–PDCCH  Average transmission power of the
i EPRE PDCCH + 10  Log  --------------------------------------------------------------------------------------------------------------------------------------
-dBm

P PDCCH  TXi  ic  TX  ic  TX  ic  TX  ic   PDCCH
 N SD – PDCCH  N SF – DL + Min  2 N SD – PDCCH  N TDD – SSF
i i i
 
 
 
 TX i  ic 

TX  ic   N 
EPRE PDSCH + 10  Log  -------------------------------------------------------------------------------------------------------------------
-
TX i  ic  i Sym–PDSCH Average transmission power of the
P PDSCH  TX i  ic  TX  ic  
dBm
PDSCH
 N  N – N  N i 
  SD  Slot Slot  SF SD – PDCCH  SF – DL 
 
  DwPTS 
TX i  ic 
 
TX i  ic  
  SD  SSF
+ N – Min 2  N  N TDD – SSF
 SD – PDCCH 
6.2.2 Co- and Adjacent Channel Overlaps Calculation

TX i  ic  TX i  ic  TX i  ic   N TXi  ic  – N First – TXi  ic 

 
F Start – Band + W Channel + ICS Band   -------------------------------------------------------
-
TX i  ic  Channel Channel Start frequency for the channel
F Start    TX i  ic   MHz
number assigned to a cell
 CN Band 
TX i  ic  TX i  ic  TX i  ic  End frequency for the channel

F End F Start + W Channel MHz
TX  ic  – TX  jc  TX  jc  TX  ic  TX  jc  TX  ic 
Min  F End  F End  – Max  F Start  F Start 
i j j i j i MHz Co-channel overlap bandwidth
W CCO    
TX i  ic  – TX j  jc 
TX  ic  – TX  jc  W CCO
i
r CCO
j ----------------------------------- None Co-channel overlap ratio
TX j  jc 
W Channel
TX i  ic  – TX j  jc  TX  jc  TX  ic  TX  jc  TX  ic  TX  ic  Bandwidth of the lower-frequency

Min  F End  F Start  – Max  F Start  F Start – W Channel
j i j i i MHz
W ACO
L     adjacent channel overlap
TX  ic  – TX  jc  W ACO Lower-frequency adjacent channel
i j L
r ACO ----------------------------------
- None
L TX j  jc  overlap ratio
W Channel
TX  ic  – TX  jc  TX j  jc  TX i  ic  TX  ic  TX  jc  TX  ic  Bandwidth of the higher-frequency

Min  F End  F End + W Channel – Max  F Start  F End 
i j i j i
W ACO MHz
H     adjacent channel overlap
TX i  ic  – TX j  jc  W ACO Higher-frequency adjacent channel
H
r ACO ----------------------------------
- None
H TX j  jc  overlap ratio
W Channel
TX i  ic  – TX j  jc  TX  ic  – TX  jc  TX  ic  – TX  jc 
i j i j
r ACO r ACO + r ACO None Adjacent channel overlap ratio
L H
TX i  ic 
TX i  ic  – TX j  jc  – f ACS
rO TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  ----------------------- None Total overlap ratio
10
r CCO + r ACO  10
6.2.3 Signal Level Calculation (DL)

The received signal levels (dBm) from any cell TXi(ic) are calculated for a pixel, subscriber, or mobile Mi as follows:
453

TX  ic  M
i i
TX  ic 
i EIRP Max – L Path – M Shadowing – Model – L Indoor + G
C Max M M M
dBm Received max cell power
i i i
–L – L Ant – L Body + f CP
TX i  ic  TX i TX i
Without smart antennas: P Max + G Ant – L
TX i  ic  With smart antennas:
EIRP Max dbm Downlink max EIRP
TX i  ic  TX i TX i TX i
+ 10  Log  E SA  +
Combining Div
P Max + G Ant – L G SA + G SA
 
TX i  ic  Mi
TX i  ic  EIRP1 DLRS – L Path – M Shadowing – Model – L Indoor + G Received downlink reference signal
C DLRS Mi Mi Mi
dBm
level
Without smart antennas: P DLRS + G Ant – L
EIRP1 DLRS dBm RS EIRP
TX  ic  TX TX TX
+ 10  Log  E SA  +
i i i i Combining
P DLRS + G Ant –L G SA
TX i  ic  Mi
TX i  ic  EIRP1 SS – L Path – M Shadowing – Model – L Indoor + G
C SS Mi Mi Mi
dBm Received SS signal level
Without smart antennas: P SS + G Ant – L
EIRP1 SS dBm SS EIRP
+ 10  Log  E SA  +
Combining Div
P SS + G Ant – L G SA + G SA
 
TX i  ic  Mi
TX i  ic  EIRP1 PBCH – L Path – M Shadowing – Model – L Indoor + G
C PBCH Mi Mi Mi
dBm Received PBCH signal level
Without smart antennas: P PBCH + G Ant – L
EIRP1 PBCH dBm PBCH EIRP
+ 10  Log  E SA  +
Combining Div
P PBCH + G Ant – L G SA + G SA
TX  ic  M
i i
TX i  ic  EIRP1 PDCCH – L Path – M Shadowing – Model – L Indoor + G
C PDCCH Mi Mi Mi
dBm Received PDCCH signal level
Without smart antennas: P PDCCH + G Ant – L
TX  ic  With smart antennas:
i dBm PDCCH EIRP
EIRP1 PDCCH
+ 10  Log  E SA  +
Combining Div
P PDCCH + G Ant – L G SA + G SA
 
TX i  ic  Mi
TX i  ic  EIRP1 PDSCH – L Path – M Shadowing – Model – L Indoor + G
C PDSCH Mi Mi Mi
dBm Received PDSCH signal level
Without smart antennas: P PDSCH + G Ant – L
TX i  ic 
EIRP1 PDSCH With smart antennas: dBm PDSCH EIRP
Array Combining Div
P PDSCH + G SA    + G SA + G SA + G SA – L
TX i  ic  Received downlink reference signal

TX i  ic  EIRP2 DLRS – L Path – M Shadowing – Model
E DLRS Mi Mi Mi Mi
dBm/Sym energy per resource element
– L Indoor + G –L – L Ant – L Body + f CP (RSRP)
454

TX  ic  TX TX
i i i
Without smart antennas: EPRE DLRS + G Ant – L
i dBm/Sym RS EIRP
EIRP2 DLRS
TX  ic  TX TX TX
+ 10  Log  E SA  + G SA
i i i i Combining
EPRE DLRS + G Ant – L
 
TX i  ic 
TX i  ic  EIRP2 SS – L Path – M Shadowing – Model Received SS energy per resource
E SS Mi Mi Mi Mi
dBm/Sym
element
– L Indoor + G –L – L Ant – L Body + f CP
Without smart antennas: EPRE SS + G Ant – L
EIRP2 SS dBm/Sym SS EIRP
+ 10  Log  E SA  + G SA
Combining Div
EPRE SS + G Ant – L + G SA
TX  ic 
i
TX i  ic  EIRP2 PBCH – L Path – M Shadowing – Model Received PBCH energy per resource
E PBCH Mi Mi Mi Mi
dBm/Sym
element
Without smart antennas: EPRE PBCH + G Ant – L
i dBm/Sym PBCH EIRP
EIRP2 PBCH
+ 10  Log  E SA  + G SA
Combining Div
EPRE PBCH + G Ant – L + G SA
 
TX i  ic 
TX i  ic  EIRP2 PDCCH – L Path – M Shadowing – Model Received PDCCH energy per resource
E PDCCH Mi Mi Mi Mi
dBm/Sym
element
Without smart antennas: EPRE PDCCH + G Ant – L
EIRP2 PDCCH dBm/Sym PDCCH EIRP
+ 10  Log  E SA  + G SA
Combining Div
EPRE PDCCH + G Ant – L + G SA
TX  ic 
i
TX i  ic  EIRP2 PDSCH – L Path – M Shadowing – Model Received PDSCH energy per resource
E PDSCH Mi Mi Mi Mi
dBm/Sym
element
TX  ic  TX TX
i i i
Without smart antennas: EPRE PDSCH + G Ant – L
TX i  ic 
EIRP2 PDSCH With smart antennas: dBm/Sym PDSCH EIRP
Array Combining Div
EPRE PDSCH + G SA    + G SA + G SA + G SA – L
TX i
L Path L Model + L Ant dB Path loss
TX i TX i
L Path + L + L Indoor + M Shadowing – Model – G
L Total Mi Mi Mi Mi
dB Total losses
+L –G + L Ant + L Body
TX i  ic 
10  Log  7  7.5  If D CP = Normal Cyclic prefix factor, i.e., the ratio of
f CP TX i  ic  dB the useful symbol energy to the total
10  Log  6  7.5  If D CP = Extended symbol energy
0 If TX i  ic  is an interferer
6.2.4 Noise Calculation (DL)

TX i  ic  Thermal noise for one resource
n 0 – Sym n 0 + 10  Log  F  dBm
element
455

TX  ic  TX  ic  M Downlink noise for one resource
i i i dBm
n Sym n 0 – Sym + nf element
6.2.5 Interference Calculation (DL)

Interfering energy per resource
element (dBm/Sym) received
TX  jc 
 TX  ic  E
j  over downlink reference signals
TX  jc  N i ---------------------
DLRS
TX  ic  – TX j  jc  (Method 1: synchronised
10  Log  -------------------  +f i
j Ant – TX 10
 DLRS  10 dBm/Sym
 TXj  jc   O transmission and reception)
 N Ant – TX  Case: RS of the interfered cell
 
collide only with RS of the
interfering cell
TX  jc 
 TX  jc  E DLRS 
j
N j --------------------- TX i  ic  – TX j  jc   N TXi  ic  – N TX j  jc  Interfering energy per resource
 Ant – TX
10  Log -------------------  10
10  + fO + 10  Log  ---------------------------------------------
Ant – TX Ant – TX
 element (dBm/Sym) received
 TXi  ic    TX i  ic 
over downlink reference signals
 N Ant – TX   N Ant – TX
  (Method 1: synchronised
TX j  jc 
 DLRS TX j  jc  TX i  ic  – TX j  jc  TX j  jc  TX i  ic  – TX j  jc  dBm/Sym transmission and reception)
E PDCCH + f PDCCH E PDSCH + f PDSCH  Case: RS of the interfered cell
----------------------------------------------------------------------- ----------------------------------------------------------------------- 
10
10
+ 3  10
10  collide with RS, PDCCH, and
----------------------------------------------------------------------------------------------------------------------------------- - PDSCH of the interfering cell
4 
 With 1 or 2 antenna ports

TX  jc  Interfering energy per resource

j
N j --------------------- TX i  ic  – TX j  jc   N TXi  ic  – N TX j  jc  element (dBm/Sym) received
 Ant – TX
10  Log  -------------------  10
10 
+ 10  Log  ---------------------------------------------
Ant – TX Ant – TX
 over downlink reference signals
TX i  ic   + fO TX i  ic 
(Method 1: synchronised
  transmission and reception)
TX  jc 
j dBm/Sym Case: RS of the interfered cell
 DLRS TX j  jc  TX i  ic  – TX j  jc  TX j  jc  TX  ic  – TX  jc 
E +f E +f
i j  collide with RS, PDCCH, and
----------------------------------------------------------------------- 
PDCCH PDCCH PDSCH PDSCH
-----------------------------------------------------------------------
10
10
+ 5  10
10  PDSCH of the interfering cell
----------------------------------------------------------------------------------------------------------------------------------- - With 4 or 8 antenna ports and
6  TX  ic 
 i
 N SD – PDCCH = 1
TX  jc  Interfering energy per resource

j
N j --------------------- TX i  ic  – TX j  jc   N TXi  ic  – N TX j  jc  element (dBm/Sym) received
10  Log  -------------------  +f + 10  Log  ---------------------------------------------
Ant – TX 10 Ant – TX Ant – TX
 10  over downlink reference signals
 TXi  ic   O  TX i  ic 
(Method 1: synchronised
  transmission and reception)
TX j  jc 
 DLRS TX j  jc  TX i  ic  – TX j  jc  TX j  jc  TX i  ic  – TX j  jc  dBm/Sym Case: RS of the interfered cell
E PDCCH + f PDCCH E PDSCH + f PDSCH  collide with RS, PDCCH, and
----------------------------------------------------------------------- ----------------------------------------------------------------------- 
10
10
+ 2  10
10  PDSCH of the interfering cell
------------------------------------------------------------------------------------------------------------------------------------  With 4 or 8 antenna ports and
3  TX i  ic 
 N SD – PDCCH  1

TX  jc  TX  ic  – TX  jc  TX  jc  TX  ic  – TX  jc  Interfering energy per resource

 EPDCCH j
+ f PDCCH
i j j
E PDSCH + f PDSCH
i j 
 ----------------------------------------------------------------------- ----------------------------------------------------------------------- element (dBm/Sym) received
 10 10
+ 3  10
10  over downlink reference signals
10  Log  ------------------------------------------------------------------------------------------------------------------------------------
 4  (Method 1: synchronised
TX  jc 
j
 DLRS   dBm/Sym transmission and reception)
  Case: RS of the interfered cell
collide with PDCCH and PDSCH of
the interfering cell
+ fO With 1 or 2 antenna ports
456

TX  jc  TX  ic  – TX  jc  TX  jc  TX  ic  – TX  jc  element (dBm/Sym) received
 E
j
+f
i j
E
j
+f
i j 
 PDCCH PDCCH
----------------------------------------------------------------------- -----------------------------------------------------------------------
PDSCH PDSCH over downlink reference signals
 10 10
+ 5  10
10  (Method 1: synchronised
10  Log  ------------------------------------------------------------------------------------------------------------------------------------
 6  transmission and reception)
TX j  jc 
 DLRS   dBm/Sym Case: RS of the interfered cell
  collide with PDCCH and PDSCH of
With 4 or 8 antenna ports and
+ fO TX i  ic 
N SD – PDCCH = 1

TX j  jc  TX i  ic  – TX j  jc  TX j  jc  TX i  ic  – TX j  jc  element (dBm/Sym) received
 EPDCCH + fPDCCH E PDSCH + f PDSCH 
 ----------------------------------------------------------------------- ----------------------------------------------------------------------- over downlink reference signals
 10 10
+ 2  10
10  (Method 1: synchronised
10  Log  ------------------------------------------------------------------------------------------------------------------------------------
 3  transmission and reception)
TX j  jc 
 DLRS   dBm/Sym Case: RS of the interfered cell
  collide with PDCCH and PDSCH of
With 4 or 8 antenna ports and
+ fO TX i  ic 
N SD – PDCCH  1
TX j  jc  TX j  jc 
 ESS E PBCH 
 --------------------
10
- ---------------------
10
TX j  jc   Interfering energy per resource
 10  N Sym – SS + 10  N Sym – PBCH
TX j  jc  10  Log  -------------------------------------------------------------------------------------------------------------
TX j  jc 
- element (dBm/Sym) received
 SS PBCH  N + N  dBm/Sym over the SS and the PBCH
 Sym – SS Sym – PBCH  (Method 1: synchronised
 
transmission and reception)
TX i  ic  – TX j  jc  TX j  jc 
+ fO + f MIMO
TX  jc 
j
 j
N Sym – DLRS in PDCCH -
-------------------- TX  ic  – TX j  jc  Interfering energy per resource
10  Log  ------------------- -  10 10  + f O i
1
 ----------------------------------------- element (dBm/Sym) received
 TX
j
 jc  TX
i
 ic  
 N Ant – TX N Sym – PDCCH  over the PDCCH
TX j  jc    (Method 1: synchronised
 PDCCH dBm/Sym
TX  jc  TX  ic  – TX  jc  transmission and reception)
 TX  ic  TX  jc  E
j
+f
i j  Case: PDCCH of the interfered cell
N i – N
j -----------------------------------------------------------------------
PDCCH PDCCH
+ 10  L og  -----------------------------------------------------------------------------  collides with PDCCH and all the RS
Sym – PDCCH Sym – DLRS in PDCCH 10
-  10
 TX  ic 
i  of the interfering cell
 N Sym – PDCCH 
 
TX  jc 
 TX  ic  TX  jc  TX  ic  E DLRS 
j
N i N
j
– N
i --------------------- Interfering energy per resource
10  Log  ------------------- -  10 10 
Ant – TX Sym – DLRS in PDCCH Sym – DLRS in PDCCH
 ------------------------------------------------------------------------------------------ element (dBm/Sym) received
 TXj  jc  TX  ic 
i 
 N Ant – TX N Sym – PDCCH  over the PDCCH
 PDCCH dBm/Sym
 TX  jc  E
j
+f
i j  Case: PDCCH of the interfered cell
N j PDCCH PDCCH
----------------------------------------------------------------------- TX  ic  – TX j  jc 
+ 10  L og  -----------------------------  +f i collides with PDCCH and some RS
Sym – PDCCH 10
 10
 TX
i
 ic   O
of the interfering cell
 
over the PDCCH
TX j  jc  TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  (Method 1: synchronised
 PDCCH E PDCCH + f PDCCH + fO dBm/Sym
Case: PDCCH of the interfered cell
collides only with PDCCH of the
interfering cell
457
TX  jc 
 TX  jc  E
j 
 N
j DLRS 
--------------------
- TX  ic  – TX  jc  Interfering energy per resource
10  Log  -------------------  +f i
1 Sym – DLRS in PDSCH 10 j
 -----------------------------------------
-  10 element (dBm/Sym) received
TX  jc 
j
TX  ic 
i  O
 N Ant – TX N Sym – PDSCH  over the PDSCH
 PDSCH dBm/Sym
 TX  ic  TX  jc 
j
E PDSCH + f PDSCH
i j  Case: PDSCH of the interfered cell
N i – N
j -----------------------------------------------------------------------
+ 10  L og  -----------------------------------------------------------------------------  collides with PDSCH and all the RS
Sym – PDSCH Sym – DLRS in PDSCH 10
 10
 TX  ic 
i  of the interfering cell
 N Sym – PDSCH 
 
TX  jc 
 TX  ic  TX  jc  TX  ic  E DLRS 
j
N i N
j
– N
i --------------------- Interfering energy per resource
10  Log  ------------------- -  10 10 
Ant – TX Sym – DLRS in PDSCH Sym – DLRS in PDSCH
 ----------------------------------------------------------------------------------------- element (dBm/Sym) received
 TXj  jc  TX  ic 
i 
 N Ant – TX N Sym – PDSCH  over the PDSCH
 PDSCH dBm/Sym
 TX  jc  E
j
+f
i j  Case: PDSCH of the interfered cell
N j -----------------------------------------------------------------------
PDSCH PDSCH
TX  ic  – TX j  jc 
+ 10  L og  ----------------------------  +f i collides with PDSCH and some RS
Sym – PDSCH 10
-  10
 TX
i
 ic   O
of the interfering cell
 
over the PDSCH
TX j  jc  TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  (Method 1: synchronised
 PDSCH E PDSCH + f PDSCH + fO dBm/Sym
Case: PDSCH of the interfered cell
collides only with PDSCH of the
interfering cell
TX  jc  TX  jc  TX  ic  – TX  jc 
 E j TX j  jc 
j
E PDCCH + f PDCCH
i j
TX j  jc 
 --------------------
DLRS -
N ----------------------------------------------------------------------- N Sym – PDCCH

10  Log  10
10 Sym – DLRS
 -------------------------
- + 10 10
 -----------------------------
TX j  jc  TX j  jc  Interfering energy per resource
 N Sym – DL N Sym – DL element (dBm/Sym) received
TX  jc  
j dBm/Sym over downlink reference signals
 DLRS
TX j  jc  TX i  ic  – TX j  jc 
E PDSCH + f PDSCH TX j  jc 
 (Method 2: non-synchronised
----------------------------------------------------------------------- N  TX  ic  – TX j  jc  transmission and reception)
10 Sym – PDSCH 
+ 10 - + fO i
 ----------------------------
TX j  jc  
N Sym – DL 

TX  jc  TX  jc  TX  jc  TX  jc 
j j j j
 ESS +f
MIMO- E PBCH + f MIMO  Interfering energy per resource
 ----------------------------------------------
10
-----------------------------------------------
10
TX  jc 
j  element (dBm/Sym) received
TX  jc 
j  10  N Sym – SS + 10  N Sym – PBCH TX i  ic  – TX j  jc 
 SS PBCH 10  Log  ---------------------------------------------------------------------------------------------------------------------------------------------------
TX j  jc 
- + fO over the SS and the PBCH
 N Sym – SS + N Sym – PBCH  (Method 2: non-synchronised
  dBm/Sym
 
j
E PDCCH + f PDCCH
i j
TX j  jc 
 --------------------
DLRS -
N Sym – DLRS ----------------------------------------------------------------------- N Sym – PDCCH
10  Log  10
10 10
 -------------------------- + 10  -----------------------------
 TX j  jc  TX j  jc  Interfering energy per resource
TX  jc  
j dBm/Sym over the PDCCH
 PDCCH
----------------------------------------------------------------------- N  TX  ic  – TX j  jc  transmission and reception)
+ 10 - + fO i
 ----------------------------
TX j  jc  
N Sym – DL 

458
 E j TX  jc  E
j
+f
i j
TX  jc 
 --------------------
DLRS
- j
N Sym – DLRS
PDCCH PDCCH
----------------------------------------------------------------------- j
N Sym – PDCCH
10  Log  10
10 10
 -------------------------
TX  jc 
- + 10  -----------------------------
TX  jc 
j j Interfering energy per resource
TX j  jc  
 PDSCH dBm/Sym over the PDSCH
E PDSCH + f PDSCH TX  jc 
----------------------------------------------------------------------- N j  TX  ic  – TX j  jc  transmission and reception)
+ 10 - + fO i
 ----------------------------
TX  jc 
j 
N Sym – DL 

 EPDSCH + fPDSCH
 ----------------------------------------------------------------------
10
- TX j  jc 
 10  N Sym – PDSCH
10  Log  ---------------------------------------------------------------------------------------------
 10 Interfering energy per frequency
 N + N block (dBm/RB) received over 1
 Sym – PDSCH Sym – PDCCH
frequency block during an OFDM
TX j  jc  
 RSSI dBm/RB symbol carrying reference signals
E PDCCH + f PDCCH TX j  jc  
-----------------------------------------------------------------------
10
TX j  jc  E DLRS  For number of antenna ports > 1,
10  N Sym – PDCCH --------------------
10
-

TX j  jc  

-  10 + 10
+ ---------------------------------------------------------------------------------------------  2  Min 2 N Ant – TX  + f O 8 is used instead of encircled 10
TX j  jc  TX j  jc   
N Sym – PDSCH + N Sym – PDCCH 

TX  jc  TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc 
 j
f MIMO + f TL
j
+ f ICIC – DL
i j
+ f ABS – DL
i j 
 TX  jc  -
-------------------------------------------------------------------------------------------------------------------------------------------------
  1 – AU j   10 10 
TX i  ic  – TX j  jc   DL   PDCCH interference weighting
f PDCCH   dB
factor
 TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  
 f
TL
+ f
ICIC – DL
+ f
ABS – DL
------------------------------------------------------------------------------------------------------------------------

 TX
j
 jc 
10 
 + AU DL  10 
 j
f MIMO + f TL
j
+ f ICIC – DL
i j
+ f ABS – DL
i j

 TX j  jc  -
-------------------------------------------------------------------------------------------------------------------------------------------------

 1 – AU DL  10

TX i  ic  – TX j  jc     10  PDSCH interference weighting
f PDSCH   dB
factor
  TX TX TX  ic  – TX  jc  TX  ic  – TX  jc  
j j  i j i j
 G SA    – G SA    + f ICIC – DL + f ABS – DL
   
 TX  jc  --------------------------------------------------------------------------------------------------------------------------------------------------------
- 
 + AU DLj  10 10

TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  Interference reduction factor due

fO 10  Log  r O  dB
  to channel overlap
Interference reduction factor due

TX  ic  – TX  jc  TX i  ic  – TX j  jc 
10  Log  p Collision 
i j dB to static downlink ICIC using
f ICIC – DL   fractional frequency reuse
TX j  jc  TX  jc  Interference reduction factor due
10  Log  TLDL 
j dB
f TL  to the downlink traffic load
Interference increment due to

TX j  jc  TX  jc 
10  Log  N Ant – TX
j dB more than one transmission
f MIMO
antenna port
TX k
 P DL – Rec 
 --------------------------------------
-
Downlink inter-technology

Inter – Tech
I DL  F  TX i  ic  TX k 
W
interference
TX k  ICP DL 
6.2.6 C/N Calculation (DL)

CNR DLRS E DLRS – n Sym dB Downlink reference signals C/N
459

i i i dB SS C/N
CNR SS E SS – n Sym
i i
TX  ic  E PBCH – n Sym
i dB PBCH C/N
CNR PBCH TX i  ic  TX i  ic  Mi
DL
With MIMO: CNR PBCH = CNR PBCH + G Div – PBCH + G Div
TX i  ic  TX i  ic 
TX i  ic  E PDCCH – n Sym
CNR PDCCH dB PDCCH C/N
TX i  ic  TX i  ic  Mi
DL
With MIMO: CNR PDCCH = CNR PDCCH + G Div – PDCCH + G Div
TX  ic  E PDSCH – n Sym
i dB PDSCH C/N
CNR PDSCH TX i  ic  TX i  ic  Mi
DL
With MIMO: CNR PDSCH = CNR PDSCH + G Div – DL + G Div
6.2.7 C/(I+N) Calculation (DL)

TX  jc  TX  ic 
    j  n Sym 
i 
TX i  ic     -------------------
DLRS 
TX i  ic  - --------------------- 
CINR DLRS E DLRS –  10  Log   10 10  + I Inter – Tech + 10 10  + NR Inter – Tech
 dB Downlink reference signals C/(I+N)
    DL  DL 
  All TXj  jc     
     
    j  n Sym 
i 
TX i  ic     -------------------------
SS PBCH
TX  ic  - --------------------- 
–  10  Log    10  +I  + NR dB 
i 10 Inter – Tech 10 Inter – Tech
CINR SS E SS + 10 SS C/(I+N)
    DL  DL 
  All TXj  jc     
     
    j  n Sym 
i 
TX i  ic     -------------------------
SS PBCH
- --------------------- 
E PBCH –  10  Log   10
  +I  + NR 
10 Inter – Tech 10 Inter – Tech
+ 10
TX i  ic      DL  DL 
CINR PBCH   All TXj  jc     dB  PBCH C/(I+N)
     
TX  ic  TX  ic  M
i i i DL
With MIMO: CINR PBCH = CINR PBCH + G Div – PBCH + G Div
    j  n Sym 
i 
TX i  ic     --------------------
PDCCH-
--------------------- 
E PDCCH –  10  Log    10  +I  + NR 
  + 10
TX i  ic  DL  DL 
CINR PDCCH   All TXj  jc     dB  PDCCH C/(I+N)
     
DL
With MIMO: CINR PDCCH = CINR PDCCH + G Div – PDCCH + G Div
    j  n Sym 
i 
TX i  ic     --------------------
PDSCH-
--------------------- 
E PDSCH –  10  Log    10 10  + I Inter – Tech + 10 10  + NR Inter – Tech
TX i  ic    DL  DL 
CINR PDSCH   All TXj  jc     dB PDSCH C/(I+N)
     
DL
With MIMO: CINR PDSCH = CINR PDSCH + G Div – DL + G Div
TX i  ic  TX i  ic  TX  ic  TX i  ic  Reference signal received quality

RSRQ 10  Log  N FB  + E i – RSSI dB
  DLRS (RSRQ)
460
   j  n
i 
 TX  ic   -------------------
RSSI 
-
Sym
--------------------- 
10  Log   RSSI +   10  +I  12 +
i 10 Inter – Tech 10
+ 10
  DL
 All TX j  jc    
TX i  ic      dBm
Received signal strength indicator
RSSI (RSSI)
TX  ic 
+ 10  Log  N FB 
Inter – Tech i
NR DL 
   j  n Sym 
i
  -------------------
DLRS -
---------------------
10  Log    10 10  + I Inter – Tech + 10 10 
  DL 
 All TXj  jc    
TX i  ic      dBm
Downlink reference signals total
 I + N  DLRS noise (I+N)
TX  ic 
+ 10  Log  2  N FB 
Inter – Tech i
+ NR DL 
   j  n
i 
  -------------------------
SS PBCH
- ---------------------
Sym
10  Log    10  +I 
10 Inter – Tech 10
+ 10
   DL 
 All TXj  jc    
TX i  ic      dBm SS and PBCH total noise (I+N)
 I + N  SS PBCH
Inter – Tech
+ NR DL + 10  Log  N SCa – FB  N FB – SS PBCH 
   j  n Sym 
i
  --------------------
PDCCH-
---------------------
10  Log    10 10  + I Inter – Tech + 10 10 
  DL  PDCCH total noise (I+N)
TX i  ic   All TXj  jc    
 I + N  PDCCH     dBm (Method 1: synchronised
TX i  ic  transmission and reception)
 N Sym – PDCCH 

+ 10  Log  TX  ic  - + NR Inter
---------------------------------------------- – Tech
TX i  ic   DL
i
 N SF – DL + N TDD – SSF
   j  n Sym 
i
  --------------------
PDCCH
- ---------------------
10  Log    10 10  + I Inter – Tech + 10 10 
  DL  PDCCH total noise (I+N)
TX  ic   All TXj  jc    
I +
i
N  PDCCH     dBm (Method 2: non-synchronised
 N TXi  ic  TX  ic 
i 
 Sym – PDSCH + N Sym – PDCCH  – Tech
- + NR Inter
+ 10  Log -------------------------------------------------------------------
 TX  ic  
i
DL
 N SD  Slot  N Slot  SF  N SF – DL
   j  n
i 
  --------------------
PDSCH
- ---------------------
Sym
10  Log    10  +I 
+ 10
   DL  PDSCH total noise (I+N)
TX i  ic   All TXj  jc    
 I + N  PDSCH     dBm (Method 1: synchronised
TX i  ic  transmission and reception)
 N Sym – PDSCH 

+ 10  Log TX  ic  - + NR Inter
---------------------------------------------- – Tech
 i
TX i  ic   DL
 N SF – DL + N TDD – SSF
461
   j  n
i 
  --------------------
PDSCH
- ---------------------
Sym
10  Log   10  +I 
+ 10
  DL  PDSCH total noise (I+N)
TX  ic   All TXj  jc    
I +
i
N  PDSCH     dBm (Method 2: non-synchronised
 N TXi  ic  TX  ic 
i 
 Sym – PDSCH + N Sym – PDCCH  – Tech
+ 10  Log - + NR Inter
-------------------------------------------------------------------
 TX  ic  
i
DL
 N SD  Slot  N Slot  SF  N SF – DL
6.2.8 Signal Level Calculation (UL)

P O_PUSCH CINR PUSCH – Max + NR UL + n PUSCH PUCCH – 10  Log  N FB  dBm Nominal PUSCH power
 
Mi  Mi TX i  ic  TX i  ic  TX i  ic   Maximum allowed transmission

P Allowed Min  P Max 10  Log  N FB  + P O_PUSCH +  FPC  L Total  dBm
power of a user equipment
 
Mi TX i
Mi EIRP PUSCH PUCCH – L Path – M Shadowing – Model – L Indoor + G Ant Received PUSCH and PUCCH signal
C PUSCH PUCCH TX i Mi Mi
dBm
level
Mi Mi Mi
P +G –L
Mi Mi Mi PUSCH and PUCCH EIRP of a user
EIRP PUSCH PUCCH With P = P Allowed without power control adjustment and dBm
equipment
Mi Mi
P = P Eff after power control adjustment
TX i  ic 
10  Log  7  7.5  If D CP = Normal Cyclic prefix factor, i.e., the ratio of
f CP TX i  ic  dB the useful symbol energy to the total
10  Log  6  7.5  If D CP = Extended symbol energy
0 If M i is an interferer
6.2.9 Noise Calculation (UL)

n 0 – PUSCH PUCCH n 0 + 10  Log  N FB  W FB  1000 dBm PUSCH and PUCCH thermal noise
n PUSCH PUCCH n 0 – PUSCH PUCCH + nf dBm PUSCH and PUCCH noise
6.2.10 Interference Calculation (UL)

Mj Mj TX i  ic  – TX j  jc  Mj TX i  ic  – TX j  jc  Received PUSCH and PUCCH
I PUSCH PUCCH C PUSCH PUCCH + f O + f TL – UL + f ICIC – UL dBm
interference
TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  Interference reduction factor due to
fO 10  Log  r O  dB
  the co- and adjacent channel overlap
Interference reduction factor due to

M M
10  Log  TL UL
j j dB the interfering mobile’s uplink traffic
f TL – UL   load
462

TX  ic  – TX  jc  TX  ic  – TX  jc 
10  Log  p Collision 
i j i j dB static uplink ICIC using fractional
f ICIC – UL  frequency reuse
6.2.11 Noise Rise Calculation (UL)

M
 j
 IPUSCH PUCCH  TX i  ic  
   non-ICIC M i n PUSCH PUCCH
  -----------------------------------------------------------------------------
- -------------------------------------------
10  Log   
10 10
  10  + 10
 Uplink noise rise for any mobile Mi in
TX  ic 
i  All M j    
NR UL   
dB cell centre of the interfered cell
 All TX  jc     TXi(ic)
j
TX i  ic 
Inter – Tech
+ NR UL – n PUSCH PUCCH
M
 j
   ICIC M  n 
 
i
------------------------------------------------------------------- -------------------------------------------
PUSCH  PUCCH
10  Log   
10 10
  10  + 10 
 All Mj    
TX i  ic   Uplink noise rise for any mobile Mi in
  dB
 All TXj  jc   
NR UL – ICIC
 cell-edge of the interfered cell TXi(ic)
TX i  ic 
Inter – Tech
+ NR UL – n PUSCH PUCCH
For any mobile Mi in cell centre of the interfered cell TXi(ic):

TX  ic  NRUL + n PUSCH PUCCH
i dBm PUSCH and PUCCH total noise (I+N)
I + N  PUSCH PUCCH For any mobile Mi in cell-edge of the interfered cell TXi(ic):
NRUL – ICIC + n PUSCH PUCCH
6.2.12 C/N Calculation (UL)

Mi TX i  ic 
C PUSCH PUCCH – n PUSCH PUCCH
Mi
CNR PUSCH PUCCH With MIMO: dB PUSCH and PUCCH C/N
Mi Mi TX i  ic 
UL
CNR PUSCH PUCCH = CNR PUSCH PUCCH + G Div – UL + G Div
6.2.13 C/(I+N) Calculation (UL)

For any mobile Mi in cell centre of the interfered cell TXi(ic):
Mi TX i  ic 
CNR PUSCH PUCCH – NR UL
For any mobile Mi in cell-edge of the interfered cell TXi(ic):
Mi
CINR PUSCH PUCCH Mi TX i  ic  dB PUSCH and PUCCH C/(I+N)
CNR PUSCH PUCCH – NR ICIC – UL
With MIMO:
UL
CINR PUSCH PUCCH = CINR PUSCH PUCCH + G Div – UL + G Div
463
 Mi  M
i  TXi  ic    Mi  Effective transmission power of a
M
i Max  PAllowed –  CINR PUSCH PUCCH –  T M + M PC   P Min dBm user equipment after power control
P Eff    B i  
UL adjustment
6.2.14 Calculation of Downlink Cell Resources

Number of modulation symbols per
scheduler resource block
Number of DwPTS modulation
DwPTS DwPTS
N Sym  SSF N SCa – FB  N SD  SSF None symbols per scheduler resource
block in the TDD special subframes
W FB Number of subcarriers per frequency
N SCa – FB ---------
- None
F block
TX  ic  TX  ic  TX  ic  TX  ic  Total number of modulation symbols
i i i i None
N Sym – DL N FB  N Sym  SRB  N SF – DL + N Sym – DwPTS in downlink
TX i  ic  TX i  ic  TX i  ic  Number of modulation symbols in
DwPTS None
N Sym – DwPTS N FB  N TDD – SSF  N Sym  SSF DwPTS
TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  Number of PDSCH modulation
R DL N Sym – DL – O DLRS – O PSS – O SSS – O PBCH – O PDCCH – O UERS None
symbols
TX i  ic  TX i  ic  TX i  ic  TX i  ic  Number of PDSCH modulation
R DwPTS N Sym – DwPTS – O DLRS  DwPTS – O PDCCH  DwPTS None
symbols in the DwPTS
TX  ic  TX  ic  TX  ic  TX  ic  TX  ic 
i i i i i None Downlink reference signals overhead
O DLRS N FB  N DLRS  SRB  N SF – DL + O DLRS  DwPTS
TX i  ic  TX i  ic  TX i  ic  TX i  ic  Downlink reference signals overhead

O DLRS  DwPTS N FB  N DLRS  DwPTS  N TDD – SSF None
in the DwPTS

 TX  ic 
if  N Ant – TX = 1
i
 8
 Number of symbols reserved for
TX i  ic   TX  ic 
 16 if  N Ant – TX = 2 None downlink reference signals in one
i
N DLRS  SRB
   scheduler resource block

 TX  ic 
i
 24  


TX i  ic 
N DLRS  DwPTS See "Calculation of Downlink Cell Resources" on page 538 None downlink reference signals in DwPTS
of one TDD special subframe
O PSS 2  NFB – SS PBCH  N SCa – FB = 144 None PSS overhead
O SSS 2  NFB – SS PBCH  N SCa – FB = 144 None SSS overhead

TX i  ic  Extended CP: 216
O PBCH None PBCH overhead
Normal CP: 240
TX  ic 
if  N SD – PDCCH = 0 : 0
i
 
if  N SD – PDCCH = 1 AND  N Ant – TX  2 :
i i
   
TX i  ic 
O PDCCH TX  ic  TX  ic  TX  ic  TX  ic  None PDCCH overhead
N i  N SCa – FB – 4  N FB
i i i
 N SF – DL + O PDCCH  DwPTS
 SD – PDCCH 
Otherwise:
TX  ic  TX  ic  TX  ic  TX i  ic  TX i  ic 
N i  i  i
 SD – PDCCH  N SCa – FB – 2  Min  4 N Ant – TX   N FB  N SF – DL + O PDCCH  DwPTS
464
TX  ic 
if  N SD – PDCCH = 0 : 0
i
 
if  N SD – PDCCH = 1 AND  N Ant – TX  2 :
i i
   
TX i  ic 
O PDCCH  DwPTS TX  ic  TX  ic  TX i  ic  PDCCH overhead in the DwPTS
N i  N SCa – FB – 4  N FB
i
 N TDD – SSF
 SD – PDCCH 
Otherwise:
TX  ic  TX  ic  TX  ic  TX  ic 
 Min  2 N i   i  i i
  SD – PDCCH  N SCa – FB – 2  Min  4 N Ant – TX   N FB  N TDD – SSF
Without smart antennas and MIMO: 0

With smart antennas and without MIMO: 12  N FB  N SF – DL
UE-specific reference signals or
i i i None demodulation reference signal
O DMRS With smart antennas and with MIMO: 24  N FB  N SF – DL
overhead
Without smart antennas and with SU-MIMO or MU-MIMO and
N Ant – TX  4 : 24  N FB  N SF – DL
6.2.15 Calculation of Uplink Cell Resources

Number of modulation symbols per
scheduler resource block
W FB Number of subcarriers per frequency
N SCa – FB ---------
- None
F block
TX i  ic  TX  ic  TX  ic  TX  ic  Total number of modulation symbols

N Sym – UL N i – N FB – PUCCH  N Sym  SRB  N SF – UL
i i None
 FB  in uplink
TX i  ic  TX i  ic  TX i  ic  TX i  ic  Nnumber of PUSCH modulation
R UL N Sym – UL – O ULSRS – O ULDRS None
symbols
TX i  ic  N SCa – FB TX i  ic  Uplink sounding reference signal

O ULSRS -  N Sym
--------------------- – UL None
N Sym  SRB overhead
TX  ic  N SCa – FB TX i  ic  Uplink demodulation reference

i
O ULDRS 2  ---------------------
-  N Sym – UL None
N Sym  SRB signal overhead
6.2.16 Calculation of Downlink UE Capacity

Max
N i + N TDD – SSF
i
Maximum downlink throughput
TP UE – DL Max – DL  SF – DL  bps
N TBB  TTI  ---------------------------------------------------- capacity of a UE category
D Frame
6.2.17 Calculation of Uplink UE Capacity

TX i  ic 
Max Max – UL N SF – UL
Maximum uplink throughput
TP UE – UL N TBB  TTI  ----------------
- bps
D Frame capacity of a UE category
465
6.2.18 Channel Throughput, Cell Capacity, Allocated Bandwidth

Throughput, and Per-user Throughput Calculation
TX i  ic 
R DL  M
i
B DL
---------------------------------
D Frame
TX  ic 
i
R DL  M
i
B DL TX  ic 
i
Mi For proportional fair schedulers: ---------------------------------  G MUG – DL Downlink peak RLC channel
D Frame kbps
CTP P – DL throughput
With SU-MIMO:
Max – M
  1 + f SU – MIMO  G SU – MIMO – DL – 1 
i
 Mi =  Mi
B DL B DL
With MU-MIMO in throughput coverage predictions:

Mi TX i  ic 
CTP P – DL  G MU – MIMO – DL
M M M Downlink effective RLC channel

CTP P – DL   1 – BLER  B DL 
i i i kbps
CTP E – DL    throughput
Mi
Mi Mi f TP – Scaling Mi Downlink application channel
CTP A – DL CTP E – DL  ------------------------
- – TP Offset kbps
100 throughput
Cap P – DL CTP P – DL  TL DL – Max kbps Downlink peak RLC cell capacity
M M M
Cap P – DL   1 – BLER  BDL 
i i i kbps Downlink effective RLC cell capacity
Cap E – DL   
Mi
Mi Mi f TP – Scaling Mi
Cap A – DL Cap E – DL  ------------------------
- – TP Offset kbps Downlink application cell capacity
100
Mi
M Cap P – DL Downlink peak RLC throughput per
PUTP P – DL
i -----------------------
- kbps
TX i  ic  user
N Users – DL
Mi
Mi Cap E – DL Downlink effective RLC throughput
PUTP E – DL -----------------------
- kbps
TX i  ic  per user
N Users – DL
M
Mi i Downlink application throughput per
Mif TP – Scaling Mi
kbps
PUTP A – DL PUTP E – DL  ------------------------
- – TPOffset user
100
TX  ic 
i
R UL  M
i
B UL
---------------------------------
D Frame
TX  ic 
i
R UL  Mi
B TX  ic 
UL i
Mi For proportional fair schedulers: ---------------------------------  G MUG – UL
D Frame kbps Uplink peak RLC channel throughput
CTP P – UL
With SU-MIMO:
Max – TX  ic 
  1 + f SU – MIMO  G SU – MIMO – UL – 1 
i
 Mi =  Mi
B UL B UL
With MU-MIMO in throughput coverage predictions:

Mi TX i  ic 
CTP P – UL  G MU – MIMO – UL
466
M M M Uplink effective RLC channel

CTP P – UL   1 – BLER  B UL 
i i i kbps
CTP E – UL    throughput
M
M i Uplink application channel
i M
i f TP – Scaling M
i kbps
CTP A – UL CTP E – UL  ------------------------
- – TP Offset throughput
100
Cap P – UL CTP P – UL  TL UL – Max kbps Uplink peak RLC cell capacity
M M M
Cap P – UL   1 – BLER  B UL 
i i i kbps Uplink effective RLC cell capacity
Cap E – UL   
Mi
Cap A – UL Cap E – UL  ------------------------
- – TP Offset kbps Uplink application cell capacity
100
Mi
M Mi N FB – UL Uplink peak RLC allocated bandwidth
i
ABTP P – UL CTP P – UL  -----------------
TX i  ic 
- kbps
throughput
N FB
M M M Uplink effective RLC allocated

ABTP P – UL   1 – BLER  B UL 
i i i kbps
ABTP E – UL    bandwidth throughput
Mi
Mi Mi f TP – Scaling Mi Uplink application allocated
ABTP A – UL ABTP E – UL  ------------------------
100 bandwidth throughput
 Cap Mi M 
Min  -----------------------
- ABTP P –i UL
Mi P – UL
PUTP P – UL  TX i  ic   kbps Uplink peak RLC throughput per user
 N Users – UL 
 Cap Mi M 
Min  -----------------------
- ABTP E –i UL
Mi E – UL Uplink effective RLC throughput per
PUTP E – UL  TXi  ic   kbps
user
Mi
Mi M f TP – Scaling M Uplink application throughput per
PUTP A – UL i
PUTP E – UL  ------------------------
i
100 user
6.2.19 Scheduling and Radio Resource Management

Sel
Mi
Sel TPD Min – DL Resources allocated to a mobile to
Mi --------------------------- None satisfy its minimum throughput
R Min – DL Sel
Mi demand in downlink
CTP P – DL
Sel
Mi
Sel TPD Min – UL Resources allocated to a mobile to
R Min – UL Sel
Mi demand in uplink
CTP P – UL
Sel
TX i  ic  Mi Remaining downlink cell resources
TX i  ic 
R Rem – DL
TL DL – Max –  R Min – DL None after allocation for minimum
Sel
Mi throughput demands
Sel
TX  ic  M Remaining uplink cell resources after

TX i  ic  i i
TL UL – Max – R Min – UL None allocation for minimum throughput
R Rem – UL
Sel
Mi demands
Sel Sel
Sel
M  Mi Mi
Max  Remaining throughput demand for a
i
TPD Rem – DL Min  TPD Max – DL – TPD Min – DL TP UE – DL kbps
mobile in downlink
 
467
Sel Sel
Sel
M  M
i
M
i Max  Remaining throughput demand for a
i
TPD Rem – UL Min  TPD Max – UL – TPD Min – UL TP UE – UL kbps
mobile in uplink
 
Sel Sel Downlink peak channel throughput

CTP P – DL  G MUG – DL kbps with multi-user diversity gain
CTP P – DL Without MUG (Proportional Fair)
Sel Sel Uplink peak channel throughput with

M M TX  ic 
i i kbps multi-user diversity gain
i
CTP P – UL CTP P – UL  G MUG – UL
Without MUG (Proportional Fair)
Sel
Mi
Sel TPD Rem – DL Remaining resource demand for a
Mi ---------------------------- None
RD Rem – DL Sel
Mi mobile in downlink
CTP P – DL
Sel
Mi
Sel TPD Rem – UL Remaining resource demand for a
Mi ---------------------------- None
RD Rem – UL Mi
Sel mobile in uplink
CTP P – UL
Sel TX i  ic 
 Mi R Rem – DL
Proportional Fair: Min  RD Rem – DL ---------------------
 N 
Sel TX i  ic 
Round Robin: Min  RD Rem – DL --------------------
-
 N 
Sel
Sel
Mi Resources allocated to a mobile to
Mi TX i  ic 
RD Rem – DL
Proportional Demand: R Eff – Rem – DL  ----------------------------------
- None satisfy its maximum throughput
R Max – DL Sel
 RDRem – DL
Sel
Mi
Sel
M
i
TPD Rem – DL
Max C/I: ----------------------------
Sel
M
i
CTP P – DL
Sel TX i  ic 
 Mi R Rem – UL
Proportional Fair: Min  RD Rem – UL --------------------
-
 N 
Sel TX  ic 
i
-
 N 
Sel
Sel
Mi Resources allocated to a mobile to
Mi TX i  ic 
RD Rem – UL
Proportional Demand: R Eff – Rem – UL  ----------------------------------
- None satisfy its maximum throughput
R Max – UL Sel
Mi demand in uplink
 RDRem – UL
Sel
Mi
Sel
Mi
TPD Rem – UL
Max C/I: ---------------------------
Sel
-
Mi
CTP P – UL
 Sel 
 TXi  ic  Mi  Effective remaining downlink
TX i  ic 
R Eff – Rem – DL
Min  R Rem – DL
 
RD Rem – DL

None resources in a cell
 M
Sel  (Proportional Demand)
i
468
 Sel 
 TXi  ic  M  Effective remaining uplink resources
TX  ic 

i
i Min  R Rem – UL RD Rem – UL None in a cell
R Eff – Rem – UL  
 M
i
  Mi
Sel
Mi 
Sel 
 
   R Max – DL  CTP E – DL
  
 Sel
 Site

Max  1 -----------------------------------------------------------------------------------------------------
-
Site M
i Site backhaul overflow ratio in
BHOF DL None
Sel Sel  downlink
  Mi Mi  

Site
 TP S1 – DL –  R Min – DL  CTP E – DL 
  
 M
Sel
i
 Site 
  Mi
Sel
Mi 
Sel 
 
   R Max – UL  CTP E – UL
  
 Sel
M i  Site

BHOF UL
Site
Max  1 ------------------------------------------------------------------------------------------------------ None Site backhaul overflow ratio in uplink
 Sel Sel 
  Mi Mi  

Site
 TPS1 – UL –  R Min – UL  CTP E – UL 
  
 Sel
M i  Site 
Sel Sel
M M
 Mi
Sel
Mi 
Sel R i i 
Max – DL  CTP P – DL
Sel Sel  R Min – DL  CTP P – DL +  ------------------------------------------------ Site 
Total resources assigned to a mobile
Mi Mi    BHOF DL  None in downlink
TL DL = R DL ------------------------------------------------------------------------------------------------------------------
Sel (Downlink traffic load of the mobile)
Mi
CTP P – DL
Sel Sel
Mi Mi
 Mi
Sel
Mi 
Sel R 
Max – UL  CTP P – UL
Sel Sel  R Min – UL  CTP P – UL +  ------------------------------------------------ Site
-

M
i
M
i    BHOF UL  None in uplink
TL UL = R UL ------------------------------------------------------------------------------------------------------------------
Sel (Uplink traffic load of the mobile)
Mi
CTP P – UL
6.2.20 User Throughput Calculation

Sel Sel Sel
Mi Mi Mi
kbps Downlink peak RLC user throughput
UTP P – DL R DL  CTP P – DL
Sel Sel
Sel
Mi Mi   Mi   Downlink effective RLC user
UTP E – DL UTP P – DL   1 – BLER  B DL   kbps
throughput
  
Sel
Sel Mi
Mi Sel Sel Downlink application user
Mif TP – Scaling Mi kbps
UTP A – DL UTP E – DL  ------------------------
- – TP Offset throughput
100
Sel Sel Sel
Mi Mi Mi
kbps Uplink peak RLC user throughput
UTP P – UL R UL  CTP P – UL
Sel Sel
Sel
Mi M
i   Mi  
UTP E – UL UTP P – UL   1 – BLER  B UL   kbps Uplink effective RLC user throughput
  
Sel
Sel Sel Mi Sel
M Mif TP – Scaling Mi
i kbps Uplink application user throughput
UTP A – UL UTP E – UL  ------------------------- – TP Offset
100
469
6.3 Available Calculations

6.3.1.1 Profile View
The point analysis profile view displays the following calculation results for the selected transmitter based on the calculation
algorithm described in "Signal Level Calculation (DL)" on page 501.
TX  ic 
i
• Downlink reference signal level C DLRS
• Path loss L Path
• Total losses L Total
M M M M
i i i i
L ,G , L Ant , L Body , and f CP are not used in the calculations performed for the profile view.
6.3.1.2 Reception View

Analysis provided in the reception view is based on path loss matrices. So, you can display received downlink reference signal
levels from the cells for which calculated path loss matrices are available. For each cell, Atoll displays the received RSRP or
reference signal, SS, or PDSCH signal levels.
Reception level bar graphs show the RSRP or signal levels in decreasing order. The maximum number of bars in the graph
depends on the studied signal level of the best server. The bar graph displays cells whose received RSRP are higher than their
minimum RSRP thresholds and are within a 30 dB margin from the studied signal level of the best server.
You can use a value other than 30 dB for the margin from the studied signal level of the best server, for example a smaller
value for improving the calculation speed. For more information on defining a different value for this margin, see the
Administrator Manual.
6.3.1.3 Interference View

Analysis provided in the interference view is based on path loss matrices. So, you can display the received signal level from
the best server and interfering signal levels from other cells for which calculated path loss matrices are available. For each cell,
Atoll displays the best server RS, SS, or PDSCH signal level, and interference from other cells.
Interference level bar graphs show the interference levels on different channels in decreasing order. The maximum number
of bars in the graph depends on the highest interference level on the studied channel. The bar graph displays cells whose C/
N are higher than the minimum interferer C/N threshold and whose interference levels are within a 30 dB margin from the
highest interference level on the studied channel.
You can use a value other than 30 dB for the margin from the highest interference level on the studied channel, for example
a smaller value for improving the calculation speed. For more information on defining a different value for this margin, see
6.3.1.4 Details View

Analysis provided in the details view is based on path loss matrices. So, you can display the received signal level from the best
server and interfering signal levels from other cells for which calculated path loss matrices are available. For each cell, Atoll
displays the RSRP and RS, SS, PBCH, PDCCH, and PDSCH signal levels, as well as interference levels on these channels from
interfering cells.
The results for the best server (first row) are displayed using bold italic characters. Other cells are listed in the decreasing order
of RSRP. All the cells from which the received RSRP is higher than their minimum RSRP thresholds are listed in the table. As
well, interference values are listed for all the cells whose C/N are higher than the minimum interferer C/N threshold and
whose interference levels are within a 30 dB margin from the highest interference level on RS.
You can use a value other than 30 dB for the margin from the highest interference level on RS, for example a smaller value for
improving the calculation speed. For more information on defining a different value for this margin, see the Administrator
Manual.
470
6.3.2 Coverage Predictions

6.3.2.1 Downlink Signal Level Coverage Predictions
The following coverage predictions are based on the received downlink reference signal levels:
• Coverage by Transmitter
• Coverage by Signal Level
• Overlapping Zones
For these calculations, Atoll calculates the received downlink reference signal level. Then, Atoll determines the selected
display parameter on each pixel inside the cell’s calculation area. Each pixel within the calculation area of TXi(ic) is considered
a non-interfering receiver.
These coverage predictions do not depend on the traffic input. Therefore, these calculations are of special interest before and
during the deployment stage of the network to study the coverage footprint of the system.
Mi Mi Mi Mi
L , G , L Ant , and L Body are not considered in the calculations performed for the downlink signal level based coverage
predictions.
For more information on downlink reference signal level calculations, see "Signal Level Calculation (DL)" on page 501.
For more information on coverage area determination and available display options, see:
• "Coverage Area Determination" on page 471.
• "Coverage Display Types" on page 472.
Coverage Area Determination
Atoll uses parameters entered in the Condition tab of the coverage prediction properties dialogue to determine coverage
areas to display. There are three possibilities.
• All Servers
The coverage area of each cell TXi(ic) corresponds to the pixels where.
MinimumThreshold  C DLRS  or L Total or L Path   MaximumThreshold
i i i
 
• Best Signal Level and a Margin

MinimumThreshold  C DLRS  or L Total or L Path   MaximumThreshold
i i i
 
AND
TX i  ic  TX j  jc 
C DLRS  Best  C DLRS  – M
ji  
Where M is the specified margin (dB). The Best function considers the highest value from a list of values.
• If M = 0 dB, Atoll considers pixels where the received downlink reference signal level from TXi(ic) is the highest.
• If M = 2 dB, Atoll considers pixels where the received downlink reference signal level from TXi(ic) is either the
highest or within a 2 dB margin from the highest
• If M = -2 dB, Atoll considers pixels where the received downlink reference signal level from TXi(ic) is 2 dB higher
than the received downlink reference signal levels from the cells which are 2nd best servers
• Second Best Signal Level and a Margin
MinimumThreshold  C DLRS  or L Total or L Path   MaximumThreshold
i i i
AND
471
TX  ic 
nd TX  jc 
C DLRS  2 Best  C DLRS  – M
i j
 
ji
Where M is the specified margin (dB). The 2nd Best function considers the second highest value from a list of values.
• If M = 0 dB, Atoll considers pixels where the received downlink reference signal level from TXi(ic) is the second
highest.
• If M = 2 dB, Atoll considers pixels where the received downlink reference signal level from TXi(ic) is either the
second highest or within a 2 dB margin from the second highest.
• If M = -2 dB, Atoll considers pixels where the received downlink reference signal level from TXi(ic) is 2 dB higher
than the received downlink reference signal levels from the cells which are 3rd best servers.
Coverage Display Types
A pixel of a coverage area is coloured if the calculated value of the selected display parameter is greater than or equal to the
defined thresholds values. Coverage consists of several independent layers that can be displayed and hidden on the map.
It is possible to display the coverage predictions with colours depending on any transmitter or cell attribute, and other criteria
such as:
• Signal Level (dBm, dBµV, dBµV/m)
• Best Signal Level (dBm, dBµV, dBµV/m): Where cell coverage areas overlap, Atoll keeps the highest value of the signal
level.
• Path Loss (dB)
• Total Losses (dB)
• Best Server Path Loss (dB): Where cell coverage areas overlap, Atoll determines the best cell (i.e., the cell with the
highest downlink reference signal level) and evaluates the path loss from this cell.
• Best Server Total Losses (dB): Where cell coverage areas overlap, Atoll determines the best cell (i.e., the cell with the
highest downlink reference signal level) and evaluates the total losses from this cell.
• Number of Servers: Atoll evaluates the number of cells that cover a pixel (i.e., the pixel falls within the coverage areas
of these cells).
6.3.2.2 Effective Signal Analysis Coverage Predictions

The following coverage predictions are based on the received downlink reference signal, SS, PDSCH, and PUSCH and PUCCH
Mi Mi Mi Mi
signal levels and noise, and take into account the receiver characteristics ( L ,G , L Ant , and L Body ) when calculating the
required parameter:
• Effective Signal Analysis (DL)
• Effective Signal Analysis (UL)
For these calculations, Atoll calculates the received signal level or C/N level at each pixel for the channel type being studied,
i.e., RS, SS, PBCH, PDCCH, PDSCH, PUSCH and PUCCH. Each pixel within the calculation area of TXi(ic) is considered a non-
interfering receiver. The properties of the non-interfering probe receiver are set by selecting a terminal, a mobility type, and
a service.
For more information on signal level calculations, see:
• "Signal Level Calculation (DL)" on page 501.
• "Signal Level Calculation (UL)" on page 523.
For more information on C/N level calculations, see:
• "C/N Calculation (DL)" on page 516.
• "C/N Calculation (UL)" on page 529.
472
These coverage predictions are best server coverage predictions, i.e., the coverage area of each cell comprises the pixels
where the cell is the best server. Best server for each pixel is calculated as explained in "Best Server Determination" on
page 535.
A pixel of a coverage area is coloured if the calculated value of the selected display type parameter is greater than or equal to
the defined thresholds values. Coverage consists of several independent layers that can be displayed and hidden on the map.
It is possible to display the Effective Signal Analysis (DL) coverage prediction with colours depending on the following display
options:
• RSRP (RS EPRE) Level (DL) (dBm)
• RS Signal Level (DL) (dBm)
• SS Signal Level (DL) (dBm)
• PBCH Signal Level (DL) (dBm)
• PDCCH Signal Level (DL) (dBm)
• PDSCH Signal Level (DL) (dBm)
• RS C/N Level (DL) (dB)
• SS C/N Level (DL) (dB)
• PBCH C/N Level (DL) (dB)
• PDCCH C/N Level (DL) (dB)
• PDSCH C/N Level (DL) (dB)
TX  jc 
j
• Delta Path Loss (dB): Atoll calculates the difference of the total losses from the second best serving cells ( L Total ) and
TX i  ic  TX j  jc  TX i  ic 
the total losses from the best serving cells ( L Total ) on each pixel of their coverage areas ( L Total – L Total ). Pixels are
coloured according to the thresholds defined in the coverage prediction. Total losses are calculated as explained in
"Signal Level Calculation (DL)" on page 453.
• ICIC Cell-edge Areas: Based on the delta path loss calculation as above. Pixels are coloured according to the colours of
the transmitter symbols on the map. The prediction is based on the delta path loss thresholds defined per cell.
It is possible to display the Effective Signal Analysis (UL) coverage prediction with colours depending on the following display
options:
• PUSCH & PUCCH Signal Level (UL) (dBm)
• PUSCH & PUCCH C/N Level (UL) (dB)
6.3.2.3 C/(I+N)-based Coverage Predictions

The following coverage predictions are based on the received signal levels, total noise, and interference.
• Coverage by C/(I+N) Level (DL)
• Service Area Analysis (DL)
• Coverage by Throughput (DL)
• Coverage by Quality Indicator (DL)
• Coverage by C/(I+N) Level (UL)
• Service Area Analysis (UL)
• Coverage by Throughput (UL)
• Coverage by Quality Indicator (UL)
Mi Mi Mi Mi
These coverage predictions take into account the receiver characteristics ( L ,G , L Ant , and L Body ) when calculating the
required parameter. For these calculations, Atoll calculates the received signal level, noise, and interference at each pixel.
Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. The properties of the non-interfering
probe receiver are set by selecting a terminal, a mobility type, and a service.
The downlink coverage predictions are based on the downlink traffic loads of the cells, and the uplink coverage predictions
are based on the uplink noise rise values. These parameters can either be calculated by Atoll during the Monte Carlo
simulations, or set manually by the user for all the cells.
For more information on RSRQ, RSSI, C/(I+N), (I+N), and bearer calculations, see:
• "C/(I+N) and Bearer Calculation (DL)" on page 518.
• "C/(I+N) and Bearer Calculation (UL)" on page 532.
473
For more information on thoughput calculations, see:

• "Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-user Throughput Calculation" on
page 547.
These coverage predictions are all best server coverage predictions, i.e., the coverage area of each cell comprises the pixels
page 535.
It is possible to display the Coverage by C/(I+N) Level (DL) coverage prediction with colours depending on the following display
options:
• RSRQ Level (DL) (dB)
• RSSI Level (DL) (dBm)
• RS C/(I+N) Level (DL) (dB)
• SS C/(I+N) Level (DL) (dB)
• PBCH C/(I+N) Level (DL) (dB)
• PDCCH C/(I+N) Level (DL) (dB)
• SS & PBCH Total Noise (I+N) (DL) (dBm)
• PDSCH C/(I+N) Level (DL) (dB)
• PDSCH & PDCCH Total Noise (I+N) (DL) (dBm)
It is possible to display the Service Area Analysis (DL) coverage prediction with colours depending on the following display
options:
• Bearer (DL)
• Modulation (DL): Modulation used by the bearer
• Service
It is possible to display the Coverage by Throughput (DL) coverage prediction with colours depending on the following display
options:
• Peak RLC Channel Throughput (DL) (kbps)
• Effective RLC Channel Throughput (DL) (kbps)
• Application Channel Throughput (DL) (kbps)
• Peak RLC Cell Capacity (DL) (kbps)
• Effective RLC Cell Capacity (DL) (kbps)
• Application Cell Capacity (DL) (kbps)
• Peak RLC Throughput per User (DL) (kbps)
• Effective RLC Throughput per User (DL) (kbps)
• Application Throughput per User (DL) (kbps)
It is possible to display the Coverage by Quality Indicator (DL) coverage prediction with colours depending on the following
display options:
• Quality indicators available in the document (Quality Indicators table): Atoll calculates the PDSCH C/(I+N) levels
received from the best serving cells at each pixel of their coverage areas. From the C/(I+N), Atoll determines the best
bearer available on each pixel. Then, for the calculated C/(I+N) and bearer, it determines the value of the selected
quality indicator from the quality graphs defined in the reception equipment of the selected terminal.
It is possible to display the Coverage by C/(I+N) Level (UL) coverage prediction with colours depending on the following display
options:
• PUSCH & PUCCH C/(I+N) Level (UL) (dB)
• PUSCH & PUCCH Total Noise (I+N) (UL) (dBm)
• Allocated Bandwidth (UL) (No. of Frequency Blocks)
M
i
• PUSCH & PUCCH C/(I+N) Level for 1 Frequency Block (UL) (dB): PUSCH & PUCCH C/(I+N) level with N FB – UL = 1
• Transmission Power (UL) (dBm)
474
It is possible to display the Service Area Analysis (UL) coverage prediction with colours depending on the following display
options:
• Bearer (UL)
• Modulation (UL): Modulation used by the bearer
• Service
It is possible to display the Coverage by Throughput (UL) coverage prediction with colours depending on the following display
options:
• Peak RLC Channel Throughput (UL) (kbps)
• Effective RLC Channel Throughput (UL) (kbps)
• Application Channel Throughput (UL) (kbps)
• Peak RLC Cell Capacity (UL) (kbps)
• Effective RLC Cell Capacity (UL) (kbps)
• Application Cell Capacity (UL) (kbps)
• Peak RLC Allocated Bandwidth Throughput (UL) (kbps)
• Effective RLC Allocated Bandwidth Throughput (UL) (kbps)
• Application Allocated Bandwidth Throughput (UL) (kbps)
• Peak RLC Throughput per User (UL) (kbps)
• Effective RLC Throughput per User (UL) (kbps)
• Application Throughput per User (UL) (kbps)
It is possible to display the Coverage by Quality Indicator (UL) coverage prediction with colours depending on the following
display options:
• Quality indicators available in the document (Quality Indicators table): Atoll calculates the PUSCH and PUCCH C/(I+N)
levels received at the best serving cells from each pixel of their coverage areas. From the C/(I+N), Atoll determines the
best bearer available on each pixel. Then, for the calculated C/(I+N) and bearer, it determines the value of the selected
quality indicator from the quality graphs defined in the reception equipment of the best serving cell.
6.3.2.4 Cell Identifier Collision Zones Coverage Prediction

The Cell Identifier Collision Zones coverage prediction is based on the received downlink reference signal levels. Atoll
calculates the received downlink reference signal level then Atoll determines the selected display parameter on each pixel
Mi
inside the cell’s calculation area. Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. L ,
Mi Mi Mi
G , L Ant , and L Body are not considered in the calculations.
For more information on downlink reference signal level calculations, see "Signal Level Calculation (DL)" on page 501.
areas to display. It is possible to determine the coverage area based on the best signal level. The coverage area of each cell
TXi(ic) corresponds to the pixels where:
TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX j  jc 
MinimumThreshold  C DLRS  or L Total or L Path   MaximumThreshold AND C DLRS  Best  C DLRS  – M
ji
• If M = 0 dB, Atoll considers pixels where the received downlink reference signal level from TXi(ic) is the highest.
• If M = 2 dB, Atoll considers pixels where the received downlink reference signal level from TXi(ic) is either the highest
or within a 2 dB margin from the highest
• If M = -2 dB, Atoll considers pixels where the received downlink reference signal level from TXi(ic) is 2 dB higher than
the received downlink reference signal levels from the cells which are 2nd best servers
475
It is possible to display the coverage predictions with colours per cell or:
• Number of interferers
• Number of interferers per cell
6.3.3 Calculations on Subscriber Lists

When calculations are performed on a list of subscribers by running the Automatic Server Allocation, Atoll calculates the path
loss again for the subscriber locations and heights because the subscriber heights can be different from the default receiver
height used for calculating the path loss matrices.
Atoll calculates the following parameters for each subscriber in the list whose Lock Status is set to None.
• Serving Base Station and Reference Cell as described in "Best Server Determination" on page 535.
Atoll calculates the following parameters for each subscriber in the list that has a serving base station assigned and whose
Lock Status is set to None or Server.
• Azimuth (  ): Angle with respect to the north for pointing the subscriber terminal antenna towards its serving base
station.
• Mechanical Downtilt (  ): Angle with respect to the horizontal for pointing the subscriber terminal antenna towards
its serving base station.
Atoll calculates the remaining parameters for each subscriber in the list that has a serving base station assigned, using the
properties of the default terminal and service. For more information, see:
• "Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-user Throughput Calculation" on
page 547.
6.3.4 Monte Carlo Simulations

The simulation process is divided into two steps.
• Generating a realistic user distribution as explained in "User Distribution" on page 476.
Atoll generates user distributions as part of the Monte Carlo algorithm based on traffic data. The resulting user
distribution complies with the traffic database and maps selected when creating simulations.
• Scheduling and Radio Resource Management as explained under "Simulation Process" on page 479.
6.3.4.1 User Distribution

During each simulation, Atoll performs two random trials. The first random trial generates the number of users and their
activity status as explained in the following sections depending on the type of traffic input.
• "Simulations Based on User Profile Traffic Maps and Subscriber Lists" on page 476.
• "Simulations Based on Sector Traffic Maps" on page 478.
Once all the user characteristics have been determined, a second random trial is performed to obtain their geographical
locations weighted according to the clutter classes, and whether they are indoor or outdoor according to the percentage of
indoor users per clutter class.
[Simulation]
RandomTotalUsers=0
6.3.4.1.1 Simulations Based on User Profile Traffic Maps and Subscriber Lists
user profiles with an associated mobility type and a given density, i.e., number of users of a user profile per km².
476
User profile traffic maps: Each polygon or line of the map is assigned a density of users with a given user profile and mobility
type. If the map is composed of points, each point is assigned a number of users with given user profile and mobility type.
Fixed subscribers listed in subscriber lists have a user profile assigned to each of them.
User profiles model the behaviour of the different user categories. Each user profile contains a list of services and parameters
describing how these services are accessed by the user.
The number of users of each user profile is calculated from the surface area (SEnv) of each environment class map (or each
polygon) and the user profile density (DUP).
N Users = S Env  D UP
• In case of user profile traffic maps composed of lines, the number of users of each
user profile is calculated from the line length (L) and the user profile density (DUP)
(users per km): N Users = L  D UP
• The number of users is a direct input when a user profile traffic map is composed of
points.
Atoll calculates the probability for a user being active at a given instant in the uplink and in the downlink according to the
service usage characteristics described in the user profiles, i.e., the number of voice calls or data sessions, the average
duration of each voice call, or the volume of the data transfer in the uplink and the downlink in each data session.
Voice Service (v)
User profile parameters for voice type services are:

• The user terminal equipment used for the service (from the Terminals table).
• The average number of calls per hour N Call .
• The average duration of a call (seconds) D Call .
N Call  D Call
Calculation of the service usage duration per hour ( p 0 : probability of an active call): p 0 = ----------------------------
-
3600
Calculation of the number of users trying to access the service v ( n v ): n v = N Users  p 0
The activity status of each user depends on the activity periods during the call, i.e., the uplink and downlink activity
UL DL
factors defined for the voice type service v, f Act and f Act .
Calculation of activity probabilities:

UL DL
Probability of being inactive: p Inactive =  1 – f Act    1 – f Act 
UL UL DL
Probability of being active in the uplink: p Active = f Act   1 – f Act 
DL DL UL
Probability of being active in the downlink: p Active = f Act   1 – f Act 
UL + DL UL DL
Probability of being active in the uplink and downlink both: p Active = f Act  f Act
Calculation of number of users per activity status:
Number of inactive users: n v – Inactive = n v  p Inactive
UL UL
Number of users active in the uplink: n v – Active = n v  p Active
DL DL
Number of users active in the downlink: n v – Active = n v  p Active
UL + DL UL + DL
Number of users active in the uplink and downlink both: n v – Active = n v  p Active
Therefore, a connected user can be either active on both links, inactive on both links, active on UL only, or active on
DL only.
Data Service (d)
User profile parameters for data type services are:

477
• The average number of data sessions per hour N Session .

DL UL
• The average data volume (in kBytes) transferred in the downlink V and the uplink V during a session.
DL UL
• The average throughputs in the downlink TP Average and the uplink TP Average for the service d.
UL DL
UL N Session  V  8 DL N Session  V  8
Calculation of activity probabilities: f = ------------------------------------------
- and f = ------------------------------------------
-
UL DL
TP Average  3600 TP Average  3600
UL DL
Probability of being inactive: p Inactive =  1 – f    1 – f 
UL UL DL
Probability of being active in the uplink: p Active = f  1 – f 
DL DL UL
Probability of being active in the downlink: p Active = f  1 – f 
UL + DL UL DL
Probability of being active in the uplink and downlink both: p Active = f f
Calculation of number of users:
Number of inactive users: n d – Inactive = N Users  p Inactive
UL UL
Number of users active in the uplink: n d – Active = N Users  p Active
DL DL
Number of users active in the downlink: n d – Active = N Users  p Active
UL + DL UL + DL
Number of users active in the uplink and downlink both: n d – Active = N Users  p Active
Calculation of the number of active users trying to access the service d (nd):
UL DL UL + DL
n d = n d – Active + n d – Active + n d – Active
are average distributions. The service and the activity status of each user are randomly
drawn in each simulation. Therefore, if you calculate several simulations at once, the
average number of users per service and average numbers of inactive, active on UL,
active on DL and active on UL and DL users, respectively, will correspond to calculated
distributions. But if you check each simulation, the user distribution between services as
well as the activity status distribution between users can be different in each of them.

Sector traffic maps are also referred to as live traffic maps. Live traffic data from the O&M is spread over the best server
coverage areas of the transmitters included in the traffic map. Either throughput demands per service or the number of active
users per service are assigned to the coverage areas of each transmitter.
For each transmitter TXi and each service s,
• Sector Traffic Maps (Throughputs)
Atoll calculates the number of active users of each service s on UL and DL in the coverage area of TXi as follows:
UL DL
UL TP Cell DL TP Cell
N = ----------------------
- and N = ----------------------
-
UL DL
TP Average TP Average
UL
Where TP Cell is the total uplink throughput demand defined in the map for any service s for the coverage area of the
DL
transmitter, TP Cell is the total downlink throughput demand defined in the map for any service s for the coverage
UL DL
area of the transmitter, TP Average is the average uplink requested throughput of the service s, and TP Average is the
average downlink requested throughput of the service s.
• Sector Traffic Maps (# Active Users)
UL DL
Atoll directly uses the defined N and N values, i.e., the number of active users on UL and DL in the transmitter
coverage area using the service s.
478
At any given instant, Atoll calculates the probability for a user being active in the uplink and in the downlink as follows:
UL DL
Users active in the uplink and downlink both are included in the N and N values. Therefore, it is necessary to
UL DL UL + DL
accurately determine the number of active users in the uplink ( n Active ), in the downlink ( n Active ), and both ( n Active ).
As for the other types of traffic maps, Atoll considers both active and inactive users.
UL DL
factors defined for the service, f Act and f Act .

UL DL
UL UL DL
DL DL UL
UL + DL UL DL
Calculation of the number of active users trying to access the service:

UL UL UL + DL DL DL UL + DL
We have: N =  p Active + p Active   n and N =  p Active + p Active   n
Where, n is the total number of active users in the transmitter coverage area using the service.
UL UL + DL DL UL + DL
UL + DL  N  p Active N  p Active 
Number of users active in the uplink and downlink both: n Active = Min  ---------------------------------------
UL + DL
 ---------------------------------------
+ DL
 or
 p Active + p Active p Active + p UL
UL DL
Active 
UL + DL UL DL DL UL
simply, n Active = Min  N  f Act N  f Act 
UL UL UL + DL
Number of users active in the uplink: n Active = N – n Active
DL DL UL + DL
Number of users active in the downlink: n Active = N – n Active
UL DL UL + DL
And, n = n Active + n Active + n Active
Calculation of the number of inactive users attempting to access the service:
nv
-  p Inactive
Number of inactive users: n Inactive = ---------------------------
1 – p Inactive
simulation, the activity status of each user is randomly drawn. Therefore, if you calculate
simulation, the activity status distribution between users can be different in each of
them.
6.3.4.2 Simulation Process

LTE cells include intelligent schedulers and radio resource management features for regulating network traffic loads,
optimising spectral efficiency, and satisfying the QoS demands of the users. Each Monte Carlo simulation in the Atoll LTE
module is a snap-shot of the network with resource allocation carried out over a duration of 1 second (100 frames). The steps
of this algorithm are listed below.
The simulation process can be summed up into the following iterative steps.
1. Generates mobiles according to the input traffic data as explained in "User Distribution" on page 476.
2. Sets initial values for the following parameters:
• Cell transmission powers and EPREs are calculated from the maximum power and EPRE offset values defined by
the user as explained in "Downlink Transmission Power Calculation" on page 485.
479
M
i
• Mobile transmission power is set to the maximum mobile power ( P Max ).
TX  ic  TX  ic  TX  ic  TX  ic  TX  ic  TX  ic 
i i i i i i
• Cell loads ( TL DL , TL UL , NR UL , NRUL – ICIC , r DL – CE , and AU DL ) are set to their current values in the
Cells table.
3. Determines the best servers for all the mobiles generated for the simulation, and determines whether they are in the
cell centre or cell-edge, as explained in "Best Server Determination" on page 535.
4. Sets the maximum PUSCH C/(I+N) of each cell to a value high enough to ensure that it will not cause any power
constraints for cell-edge mobiles.
TX i  ic 
For all the mobiles Mi served by any cell TXi(ic) in the uplink, Atoll calculates CINR PUSCH – Max as follows to ensure
access to the highest bearer using all the frequency blocks.
From fractional power control (see "Signal Level Calculation (UL)" on page 523), we know that:
Mi
P Allowed = CINR PUSCH – Max + NRUL + n PUSCH PUCCH +  FPC  L Total (1)
Where CINR PUSCH – Max is the maximum PUSCH C/(I+N), NRUL is the noise rise, n PUSCH PUCCH is the uplink thermal
noise,  FPC is the fractional power control factor, and L Total are the total losses.
Mi
Transmitting P Allowed , a mobile Mi can access the highest bearer if:
Mi Mi
P Allowed – NR UL – n PUSCH PUCCH – L Total = T B (2)
Mi
Where T B is the bearer selection thresholds of the highest bearer defined in the reception equipment used by the
cell TXi(ic).
Mi
Combining equations (1) and (2), we get the CINR PUSCH – Max for each mobile Mi that ensures access to the highest
bearer:
M M TX  ic 
CINR PUSCH – Max = T B +  1 –  FPC   L Total
i i i
 
For each cell TXi(ic), the highest value is kept:
TX  ic  M
CINR PUSCH – Max = Max  CINR PUSCH – Max
i i
All M
 
i
For each iteration k, the simulation process,

5. Determines the downlink and uplink C/(I+N) and bearers for each of these mobiles as explained in "C/(I+N) and Bearer
Calculation (DL)" on page 518 and "C/(I+N) and Bearer Calculation (UL)" on page 532 respectively.
6. Determines the channel throughputs at the mobile as explained in "Throughput Calculation" on page 538.
7. Performs radio resource management and scheduling to determine the amount of resources to allocate to each
mobile according to the service priorities and throughput demands of each mobile using the selected scheduler as
explained in "Scheduling and Radio Resource Management" on page 552.
8. Calculates the user throughputs after allocating resources to each mobile as explained in "User Throughput
Calculation" on page 561.
480
Figure 6.1: LTE Simulation Algorithm
9. Updates the traffic loads, and noise rise values of all the cells according to the resources in use and the total resources
as follows:
Calculation of Traffic Loads:
Atoll calculates the traffic loads for all the cells TXi(ic).
TX  ic  M TX  ic  M
 RDL and TLUL  RUL
i i i i
TL DL = =
Mi Mi
MU – MIMO – DL MU – MIMO – UL
TX i  ic  Mi TX i  ic  Mi
For MU-MIMO, TL DL =  RC DL and TL UL =  RC UL
MU – MIMO – DL MU – MIMO – UL
M M
i i
Calculation of Uplink Noise Rise:

For each victim cell TXi(ic), the uplink noise rise is calculated and updated by considering each interfering mobile Mj
as explained in "Interference Calculation (UL)" on page 525.
Calculation of Downlink Cell-edge Traffic Ratio:
Atoll calculates the downlink cell-edge traffic ratio for all the cells as follows:
CE
Mi
 R DL
TX i  ic  CE
Mi
r DL – CE = ---------------------
TX i  ic 
TL DL
481
CE
M

i
Where R DL is the sum of the percentages of the downlink cell resources allocated to mobiles in the cell-edge.
CE
M
i
Calculation of Downlink AAS Usage:

Atoll calculates the downlink AAS usages for all the cells as follows:
M

i
R DL
AAS
TX i  ic  Mi
AU DL = -------------------------------
AAS -
TX i  ic 
TL DL
M

i
Where R DL is the sum of the percentages of the downlink cell resources allocated to mobiles served by the
AAS
Mi
AAS
smart antennas.
10. Performs uplink noise rise control as follows:
For each cell TXi(ic), Atoll calculates the difference between the current and the maximum noise rise values (in terms
of IoT, i.e., the ratio of interference over thermal noise I/N which can be calculated from the noise rise:
IoT = I/N = (I+N)/N - 1):
  NR i   NR i 
TX i  ic    ------------------------
UL
-   -------------------------------
UL – Max
- 
=  10  Log  10 – 1  – 10  Log  10 – 1
10 10
NR UL
    
    
    
TX  ic 
i
Here NRUL is the uplink noise rise of the cell TXi(ic) calculated in step 9.
The default method of uplink noise rise control is the best effort method. This means that uplink noise rise control is
not part of the simulation convergence criteria. In other words, a simulation will converge once the downlink and
uplink traffic loads and the uplink noise rise values are stable, irrespective of whether or not the noise rise control has
been successful. The resulting noise rise values may be higher than the maximum allowed values defined per cell. If
the resulting noise rise values are higher than the maximum allowed, this means that the noise rise control requires
more iterations for stabilising the overall network’s noise rise than those needed by the simulation to converge. If you
wish to achieve optimum noise rise control, you should decrease the uplink noise rise convergence threshold defined
for the simulation so that the simulation takes more iterations to converge and allows noise rise control to reach its
goal. The best effort noise rise control works as follows:
TX i  ic 
• If NR UL  0 , the cell TXi(ic) requests its neighbouring cells to decrease the uplink transmission powers of the
mobiles they serve (mobiles interfering TXi(ic)).
TX i  ic 
• If 0  NR UL  M NRC , the cell TXi(ic) does not request any change.
TX i  ic 
• If NR UL  M NRC , the cell TXi(ic) requests its neighbouring cells to increase the uplink transmission powers of
the mobiles they serve (mobiles interfering TXi(ic)).
Here M NRC is a noise rise control margin set to -1 dB by default. This value can be changed through Atoll.ini file by
adding the following lines and setting it to a value other than "1" (positive values are considered as negative margins):
[LTE]
NR_CONTROL_MARGIN_MIN = 1
The uplink transmission powers of the mobiles in neighbouring cells of the cell TXi(ic) are adjusted according to the
request in the next iteration by updating the maximum PUSCH C/(I+N) for the neighbouring cells TXj(jc):
TX j  jc  TX j  jc  TX i  ic  TX  jc 
= Min  CINR PUSCH – Max  CINR PUSCH – Max CINR PUSCH – Limit
j
CINR PUSCH – Max – NRUL
k  k–1 
482
TX  jc 
j
Here CINR PUSCH – Max is the maximum PUSCH C/(I+N) for the neighbouring cell TXj(jc) in the current iteration k,
k
TX  jc 
j
CINR PUSCH – Max is the maximum PUSCH C/(I+N) for the neighbouring cell TXj(jc) in the previous iteration k-1,
k–1
TX  jc 
j
CINR PUSCH – Limit is an upper limit fixed at 50 dB, and CINR PUSCH – Max is the maximum PUSCH C/(I+N) for the
neighbouring cell TXj(jc) as calculated in step 4.
If you wish to include the uplink noise rise control in the simulation convergence criteria, you can change the uplink
noise rise control method from best effort to strict by setting the following option in the Atoll.ini file:
[LTE]
ULNRControlMethod = 1
The strict uplink noise rise control method makes the uplink noise rise control a part of the simulation convergence
criteria. In other words, a simulation will converge once the downlink and uplink traffic loads and the uplink noise rise
values are stable, and the uplink noise rise values of all the cells are less than or equal to the defined maximum uplink
noise rise. The strict noise rise control works as follows:
TX i  ic 
• If NR UL  0 , the cell TXi(ic) requests its neighbouring cells to decrease the uplink transmission powers of the
mobiles they serve (mobiles interfering TXi(ic)).
TX i  ic 
• If NR UL  m NRC , the cell TXi(ic) requests its neighbouring cells to increase the uplink transmission powers of
the mobiles they serve (mobiles interfering TXi(ic)).
Here m NRC is a noise rise control precision level set to 0.5 dB by default. This value can be changed through Atoll.ini
file by adding the following lines:
[LTE]
ULNRControlPrecision = 5
Setting this option to X means that the precision will be taken as 0.X dB. The default value is 5 (= 0.5 dB).
The uplink transmission powers of the mobiles in neighbouring cells of the cell TXi(ic) are adjusted according to the
request in the next iteration by updating the maximum PUSCH C/(I+N) for the neighbouring cells TXj(jc):
TX j  jc  TX  jc  TX i  ic  TX  jc 
= Min  CINR PUSCH – Max  CINR PUSCH – Max CINR PUSCH – Limit
j j
CINR PUSCH – Max – NR UL
k k–1
TX  jc 
j
Here CINR PUSCH – Max is the maximum PUSCH C/(I+N) for the neighbouring cell TXj(jc) in the current iteration k,
k
TX j  jc 
CINR PUSCH – Max is the maximum PUSCH C/(I+N) for the neighbouring cell TXj(jc) in the previous iteration k-1, and
k–1
TX j  jc 
CINR PUSCH – Max is the maximum PUSCH C/(I+N) for the neighbouring cell TXj(jc) as calculated in step 4.
At most six neighbouring cells are considered in uplink noise rise control. These six
neighbouring cells are those whose served mobiles generate the highest interference for
the studied cell.
11. Performs the convergence test to see whether the differences between the previous and current values are within the
TX i  ic  TX  ic  TX i  ic 
Max  TL DL 
i
TL DL = – TL DL
k All TX  ic 
 k k – 1
i
Max  TL UL 
i
TL UL = – TL UL
k All TX  ic  k k – 1
i
Max  NR UL 
i i i
NR UL = – NR UL
i
483
i i i
If TL DL , TL UL , and NR UL are the simulation convergence thresholds defined when creating
Req Req Req
the simulation, Atoll stops the simulation in the following cases.
Convergence: Simulation has converged between iteration k - 1 and k, with the best effort uplink noise rise control, if:
i i i i i i
TL DL  TL DL AND TL UL  TL UL AND NR UL  NR UL
k Req k Req k Req
Simulation has converged between iteration k - 1 and k, with the strict uplink noise rise control, if:
TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic 
TL DL  TL DL AND TL UL  TL UL AND NR UL  NR UL AND
k Req k Req k Req
i i
NRUL  NR UL – Max
k
No convergence: Simulation has not converged even after the defined maximum number of iterations, with the best
effort uplink noise rise control, if:
TL DL  TL DL OR TL UL  TL UL OR NR UL  NR UL
k Req k Req k Req
Simulation has not converged even after the defined maximum number of iterations, with the strict uplink noise rise
control, if:
TL DL  TL DL OR TL UL  TL UL OR NR UL  NR UL OR
k Req k Req k Req
NRUL  NR UL – Max
k
12. Repeats the above steps (from step 3.) for the iteration k+1 using the new calculated loads as the current loads.
Simulation Results
At the end of the simulation process, the main results obtained are:
• Downlink traffic loads
• Uplink traffic loads
• Uplink noise rise
• Downlink ICIC ratio
• Uplink ICIC noise rise
• Downlink AAS usage
• Number of co-scheduled MU-MIMO users (DL)
• Number of co-scheduled MU-MIMO users (UL)
• Maximum PUSCH C/(I+N)
• Number of connected users in downlink
• Number of connected users in uplink
These results can be used as input for C/(I+N)-based coverage predictions.
In addition to the above parameters, the simulations also list the connection status of each mobile. Mobiles can be rejected
due to:
• No Coverage: If an LTE mobile does not have any best serving cell with cell type "LTE" and if an LTE-A mobile does not
have any best serving primary cell with cell type "LTE-A PCell" (step 3.)
• No Service: If the mobile is not able to access a bearer in the direction of its activity (step 5.), i.e., UL, DL, or DL+UL, or
if the mobile’s minimum throughput demand is higher than the UE throughput capacity.
• Scheduler Saturation: If the mobile is not in the list of mobiles selected for scheduling (step 7.). For LTE-A mobiles,
this applies to the mobiles selected for scheduling by their primary cells.
• Resource Saturation: If all the cell resources are used up before allocation to the mobile or if, for a user active in
uplink, the minimum uplink throughput demand is higher than the uplink allocated bandwidth throughput (step 7.).
For LTE-A mobiles, this applies to their primary cells.
• Backhaul Saturation: If allocating resources to a mobile makes the effective RLC aggregate site throughputs exceed
the maximum S1 interface throughputs defined for the site. This condition is only verified if the simulation was created
with the Backhaul capacity check box selected (step 7.)
Connected mobiles (step 7.) can be:
• Connected UL: If a mobile active in UL is allocated resources in UL.
• Connected DL: If a mobile active in DL is allocated resources in DL.
• Connected DL+UL: If a mobile active in DL+UL is allocated resources in DL+UL.
484
6.4 Calculation Details

The following sections describe all the calculation algorithms used in point analysis, calculation of coverage predictions,
calculations on subscriber lists, and Monte Carlo simulations.
6.4.1 Downlink Transmission Power Calculation

LTE eNode-Bs have a maximum transmission power which is shared by downlink channels. These channels include the
downlink reference signals, SSS, PSS, PBCH, PDCCH (which is considered to include the PHICH and PCFICH), and PDSCH. The
transmission powers of various channels are determined from the distribution of the total energy over a frame among the
resource elements corresponding to these channels. The energy per resource element (EPRE) of the downlink reference
signals is considered to be the reference with respect to which the EPRE of other channels is determined. You can either define
the reference signal EPRE for each cell, or let Atoll calculate it from the cell’s maximum power and the EPRE offsets of other
channels. The EPRE offsets of channels other than the downlink reference signals can be positive values meaning a relative
boost with respect to the downlink reference signals EPRE, or negative values meaning a reduction with respect to the
downlink reference signals EPRE.
Atoll first determines the EPRE for each channel in the downlink and then the transmission power corresponding to each
channel from the EPRE values.
Input
• F : Subcarrier width (15 kHz).

• W FB : Width of a frequency block (180 kHz).
• N FB – SS PBCH : Number of frequency blocks that carry the SS and the PBCH (6).
• N Slot  SF : Number of slots per subframe (2).
TX i  ic 
• D CP : Cyclic prefix duration defined in the TXi(ic) frame configuration or, otherwise, global network settings.
• N SD  Slot : Number of symbol durations per slot (7 is D CP is Normal, 6 if D CP is Extended).
TX  ic 
i
• N SD – PDCCH: Number of PDCCH symbol durations per subframe defined in the TXi(ic) frame configuration or,
otherwise, global network settings.
TX  ic  TX  jc 
i j
• N FB and N FB : Total number of frequency blocks defined in the frequency bands table for the channel
bandwidth used by the cell.
• N FB – CE0 and N FB – CE0 : Number of cell-edge frequency blocks defined in the frame configurations table for the
channel bandwidth used by the cell and PSS ID 0.
TX i  ic 
• N SF – DL : Number of downlink subframes in the frame for the cell TXi(ic). It is equal to 10 for FDD frequency bands, and
is determined from the cell’s TDD frame configuration for TDD frequency bands.
TX i  ic 
• N TDD – SSF : Number of TDD special subframes (containing DwPTS, GP, and UpPTS) in the frame for the cell TXi(ic). It is
equal to 0 for FDD frequency bands, and is determined from the cell’s TDD frame configuration for TDD frequency
bands.
N SF – DL and N TDD – SSF are determined as follows:
Configuration N SF – DL N TDD – SSF
FDD 10 0
DSUUU-DSUUU 2 2
DSUUD-DSUUD 4 2
DSUDD-DSUDD 6 2
DSUUU-DSUUD 3 2
485
Configuration i i
N SF – DL N TDD – SSF
DSUUU-DDDDD 6 1
DSUUD-DDDDD 7 1
DSUDD-DDDDD 8 1
TX i  ic 
• N Ant – TX : Number of transmission (downlink) antenna ports defined for the cell TXi(ic).
TX i  ic 
• P Max : Maximum transmission power of the cell TXi(ic).
TX  ic 
i
• EPRE DLRS : Downlink reference signal EPRE of the cell TXi(ic).
You can either set the P Max or EPRE DLRS for a cell.
TX  ic 
i
• EPRE SS : Energy per resource element offset for the SS with respect to the downlink reference signals EPRE.
TX i  ic 
• EPRE PBCH : Energy per resource element offset for the PBCH with respect to the downlink reference signals EPRE.
TX i  ic 
• EPRE PDCCH : Energy per resource element offset for the PDCCH with respect to the downlink reference signals EPRE.
TX i  ic 
• EPRE PDSCH : Energy per resource element offset for the PDSCH with respect to the downlink reference signals EPRE.
Calculations
If you have directly entered the downlink reference signal EPRE for the cell, you can skip the section "Calculation of Downlink
Reference Signal EPRE" on page 486 and go directly to the section "Calculation of Other EPREs and Per-channel Powers" on
page 491.
Calculation of Downlink Reference Signal EPRE
In LTE, a resource block (RB) is defined as 1 frequency block by 1 slot. However, schedulers are able to perform resource
allocation every subframe (2 slots). 1 frequency block by 1 subframe (2 slots) is called a scheduler resource block (SRB) in the
calculations below.
The number of modulation symbols (resource elements) per scheduler resource block is calculated as follows:
N Sym  SRB = N SCa – FB  N SD  Slot  N Slot  SF
Where N SCa – FB is the number of subcarriers per frequency block calculated as follows:
W FB
N SCa – FB = ---------
-
F
The number of modulation symbols (resource elements) corresponding to the DwPTS per scheduler resource block in the TDD
special subframes is calculated as follows:
DwPTS DwPTS
N Sym  SSF = N SCa – FB  N SD  SSF
DwPTS
Where N SD  SSF is the number of DwPTS symbol durations (OFDM symbols) per special subframe, determined from the TDD
special subframe configuration according to the 3GPP specifications as follows:
486
Cyclic Prefix = Normal Cyclic Prefix = Extended

Special
Subframe DwPTS GP UpPTS DwPTS GP UpPTS
Configuration DwPTS
N SD  SSF
GP
N SD  SSF
UpPTS
N SD  SSF
DwPTS
N SD  SSF
GP
N SD  SSF
UpPTS
N SD  SSF
0 3 10 3 8
1 9 4 8 3
1
2 10 3 1 9 2
3 11 2 10 1
4 12 1 3 7
5 3 9 8 2 2
6 9 3 9 1
2
7 10 2
8 11 1
The total number of modulation symbols (resource elements) in downlink is calculated as follows:
DwPTS
N Sym – DL = N FB  N Sym  SRB  N SF – DL + N FB  N TDD – SSF  N Sym  SSF
Out of the total number of modulation symbols, Atoll then determines the numbers of modulation symbols corresponding to
each control channel as follows:
The number of modulation symbols for the downlink reference signals
The number of modulation symbols reserved for downlink reference signal transmission in one scheduler resource
block depends on the number of transmission antenna ports:

 TX  ic 
i
 8  

TX i  ic   TX  ic 
For all subframes except the TDD special subframes: N Res  SRB =  16 if  N Ant – TX = 2
i


 TX  ic 
i
 24  

For TDD special subframes:
Special Cyclic Prefix = Normal Cyclic Prefix = Extended

Subframe DwPTS TX  ic  TX  ic  DwPTS TX  ic  TX  ic 
i i i i
Configuration N SD  SSF N Ant – TX N Res  DwPTS N SD  SSF N Ant – TX N Res  DwPTS
1 2 1 2
2 4 2 4
0 3 3
4 8 4 8
8 8 8 8
1 6 1 6
2 12 2 12
1 9 8
4 20 4 20
8 20 8 20
1 6 1 6
2 12 2 12
2 10 9
4 20 4 20
8 20 8 20
1 6 1 8
2 12 2 16
3 11 10
4 20 4 24
8 20 8 24
487

i i i i
Configuration N SD  SSF N Ant – TX N Res  DwPTS N SD  SSF N Ant – TX N Res  DwPTS
1 8 1 2
2 16 2 4
4 12 3
4 24 4 8
8 24 8 8
1 2 1 6
2 4 2 12
5 3 8
4 8 4 20
8 8 8 20
1 6 1 6
2 12 2 12
6 9 9
4 20 4 20
8 20 8 20
1 6
2 12
7 10
4 20
8 20
1 6
2 12
8 11
4 20
8 20
This gives a number of reserved modulation symbols per frame:

N Sym – Res = N SF – DL  N FB  N Res  SRB + N TDD – SSF  N FB  N Res  DwPTS
The number of modulation symbols used for downlink reference signal transmission in one scheduler resource block
is:

 TX  ic 
if  N Ant – TX = 1
i
 8

TX i  ic   TX  ic 
For all subframes except the TDD special subframes: N DLRS  SRB =  8 if  N Ant – TX = 2
i


 TX  ic 
i
 6  

For TDD special subframes:

i i i i
Configuration N SD  SSF N Ant – TX N DLRS  DwPTS N SD  SSF N Ant – TX N DLRS  DwPTS
1 2 1 2
2 2 2 2
0 3 3
4 2 4 2
8 2 8 2
1 6 1 6
2 6 2 6
1 9 8
4 5 4 5
8 5 8 5
488

i i i i
1 6 1 6
2 6 2 6
2 10 9
4 5 4 5
8 5 8 5
1 6 1 8
2 6 2 8
3 11 10
4 5 4 6
8 5 8 6
1 8 1 2
2 8 2 2
4 12 3
4 6 4 2
8 6 8 2
1 2 1 6
2 2 2 6
5 3 8
4 2 4 5
8 2 8 5
1 6 1 6
2 6 2 6
6 9 9
4 5 4 5
8 5 8 5
1 6
2 6
7 10
4 5
8 5
1 6
2 6
8 11
4 5
8 5
This gives a number of downlink reference signal modulation symbols per frame:
N Sym – DLRS = N SF – DL  N FB  N DLRS  SRB + N TDD – SSF  N FB  N DLRS  DwPTS
The number of modulation symbols for the SS

The primary and secondary synchonisation signals are transmitted on 1 symbol duration each in the 1st and the 6th
downlink subframes, over the center 6 frequency blocks. Therefore,
N Sym – PSS = 2  N FB – SS PBCH  N SCa – FB = 144
N Sym – SSS = 2  N FB – SS PBCH  N SCa – FB = 144
And, N Sym – SS = N Sym – PSS + N Sym – SSS = 288
The number of modulation symbols for the PBCH

The physical broadcast channel is transmitted on four symbol durations in the 1st downlink subframe over the center
6 frequency blocks. The physical broadcast channel overlaps with the downlink reference signals, therefore, some
modulation symbols reserved for downlink reference signals are subtracted:
216 for extended cyclic prefix
240 for normal cyclic prefix
The number of modulation symbols for the PDCCH
489
The physical downlink control channel can be transmitted over up to 4 symbol durations in each subframe. The
number of symbol durations for the PDCCH is defined in the global network settings. The physical downlink control
channel overlaps with the downlink reference signals, therefore, some modulation symbols reserved for downlink
reference signals are subtracted:
TX  ic 
if  N SD – PDCCH = 0 :
i
TX i  ic 
N Sym – PDCCH = 0
i i
N i  i i
 SD – PDCCH  N SCa – FB – 4  N FB
TX i  ic   N SF – DL
N Sym – PDCCH =
+  N SD – PDCCH  N SCa – FB – 4  N FB  N TDD – SSF
 
Otherwise:
N i  i  i
 SD – PDCCH  N SCa – FB – 2  Min  4 N Ant – TX   N FB
TX i  ic   N SF – DL
N Sym – PDCCH =
+  Min  2 N SD – PDCCH  N SCa – FB – 2  Min  4 N Ant – TX   N FB
i i i
 N TDD – SSF
The number of modulation symbols for the PDSCH

The total number of modulation symbols in the frame excluding all the control channel modulation symbols gives the
number of modulation symbols available for user data, i.e., for the PDSCH:
N Sym – PDSCH = N Sym – DL – N Sym – Res – N Sym – SS – N Sym – PBCH – N Sym – PDCCH
The energy per resource element for 1 modulation symbol (dBm/Sym) of the downlink reference signals is calculated as
follows:
• If the reference signal EPRE calculation method is set to Calculated (equal distribution of unused EPRE):
TX  ic 
 P i 
TX i  ic   --------------------
Max
- TX i  ic  TX i  ic  
EPRE DLRS 
= 10  Log 10
10  DwPTS  
 N SD  Slot  N Slot  SF  N SF – DL + N TDD – SSF  N SD  SSF –
  
 
 
 EPRE SS
i
EPRE PBCH
i
 TX  ic  ----------------------------------- -----------------------------------
10  L og  N Sym – DLRS + N Sym – SS  10
i 10 10
+ N Sym – PBCH  10



EPRE PDCCH EPRE PDSCH 
------------------------------------ ------------------------------------ 
10
10 



• If the reference signal EPRE calculation method is set to Calculated (with boost) or Calculated (without boost):
490
TX  ic 
 P i 
TX  ic   --------------------
Max
- TX  ic  TX  ic  
= 10  Log  10   N SD  Slot  N Slot  SF  N SF – DL + N TDD – SSF  N SD  SSF  –
i 10 i i DwPTS
EPRE DLRS
 
 
 EPRE SS
i
EPRE PBCH
i
 TX  ic  ----------------------------------- -----------------------------------
10  L og  N Sym – Res + N Sym – SS  10
i 10 10
+ N Sym – PBCH  10


EPRE PDSCH 
i i
EPRE PDCCH
------------------------------------ ------------------------------------ 
10
10 



Calculation of Other EPREs and Per-channel Powers

The energy per resource element for 1 modulation symbol (dBm/Sym) of the SS is calculated as follows:
EPRE SS = EPRE DLRS + EPRE SS
The energy per resource element for 1 modulation symbol (dBm/Sym) of the PBCH is calculated as follows:
EPRE PBCH = EPRE DLRS + EPRE PBCH
The energy per resource element for 1 modulation symbol (dBm/Sym) of the PDCCH is calculated as follows:
EPRE PDCCH = EPRE DLRS + EPRE PDCCH
The energy per resource element for 1 modulation symbol (dBm/Sym) of the PDSCH is calculated as follows:
EPRE PDSCH = EPRE DLRS + EPRE PDSCH
If the reference signal EPRE calculation method is set to Calculated (with boost), the "boosted" RS energy per resource
element is calculated as follows:
TX i  ic  TX i  ic   N TXi  ic  
+ 10  Log  -------------------------
Sym – Res 
EPRE DLRS = EPRE DLRS -
 TXi  ic  
 N Sym – DLRS
The instantaneous downlink reference signal transmission power is calculated as follows:
P DLRS = EPRE DLRS + 10  Log  2  N FB 
 
TX  ic 
i
Where 2  NFB implies that at the instant when downlink reference signals are transmitted, they are transmitted using 2
subcarriers in each frequency block.
The instantaneous SS transmission power is calculated as follows:
P SS = EPRE SS + 10  Log  N SCa – FB  N FB – SS PBCH 
The instantaneous PBCH transmission power is calculated as follows:

P PBCH = EPRE PBCH + 10  Log  N SCa – FB  N FB – SS PBCH 
Where N SCa – FB  N FB – SS PBCH implies that at the instant when the SS and the PBCH are transmitted, they are transmitted
using all the subcarriers in the centre 6 consecutive frequency blocks.
The average PDCCH transmission power is calculated as follows:
 TX i  ic  
TX i  ic  
TX i  ic  N Sym–PDCCH 
P PDCCH = EPRE PDCCH + 10  Log  ---------------------------------------------------------------------------------------------------------------------------------------
 TX  ic  TX  ic  TX  ic  TX  ic  
 N SD – PDCCH  N SF – DL + Min  2 N SD – PDCCH  N TDD – SSF
i i i i
The average PDSCH transmission power is calculated as follows:
491
 
 
 
 TX  ic 

TX  ic  TX  ic   i
N Sym–PDSCH 
= EPRE PDSCH + 10  Log --------------------------------------------------------------------------------------------------------------------

i i
P PDSCH
 TX  ic  TX  ic  
 N  N – N
i  N i 
  SD  Slot Slot  SF SD – PDCCH SF – DL 
 
  DwPTS 
TX i  ic 
 
TX i  ic  
 +  N SD  SSF – Min  2 N SD – PDCCH   N TDD – SSF
As the number of subcarriers used for the PDCCH and PDSCH transmission varies over time, i.e., from one symbol duration to
the next, the instantaneous powers of the PDCCH and the PDSCH also vary over time. This is why average transmission powers
are calculated and used in Atoll.
EPRE and Transmission Power adjustment for ICIC
The following applies to RS, PDCCH, and PDSCH EPREs for cells using downlink static ICIC.
1. No ICIC, time-switched FFR, and hard FFR
Cell-edge and cell-centre frequency blocks are not transmitted at the same time. Therefore,
TX i  ic  TX i  ic  N FB TX i  ic  TX i  ic  N FB
EPRE DLRS CC = EPRE DLRS  -----------------
TX i  ic 
and EPRE DLRS CE = EPRE DLRS  -----------------
TX i  ic 
N FB – CC N FB – CE
N FB TX i  ic  TX i  ic  N FB
EPRE PDCCH CC = EPRE PDCCH  -----------------
TX i  ic 
and EPRE PDCCH CE = EPRE PDCCH  -----------------
TX i  ic 
i i
N FB TX i  ic  TX i  ic  N FB
EPRE PDSCH CC = EPRE PDSCH  -----------------
TX  ic 
and EPRE PDSCH CE = EPRE PDSCH  -----------------
TX  ic 
i i
P DLRS CC = P DLRS CE = P DLRS
P PDCCH CC = P PDCCH CE = P PDCCH
P PDSCH CC = P PDSCH CE = P PDSCH
2. Soft and partial soft FFR

Cell-edge and cell-centre frequency blocks are transmitted at the same time; therefore, power is divided among cell-
centre and cell-edge frequency blocks. Therefore, we have,
TX i  ic 
TX i  ic  TX i  ic  N FB TX i  ic  TX i  ic  TX i  ic 
EPRE DLRS CC = EPRE DLRS  ---------------------------------------------------------------------- and EPRE DLRS CE = EPRE DLRS CC   CE
 i  N i 
 CE FB – CE + N FB – CC
TX i  ic 
EPRE PDCCH CC = EPRE PDCCH  ---------------------------------------------------------------------
- and EPRE PDCCH  CE = EPRE PDCCH CC   CE
 i  N i 
 CE FB – CE + N FB – CC
TX i  ic 
EPRE PDSCH CC = EPRE PDSCH  ---------------------------------------------------------------------- and EPRE PDSCH CE = EPRE PDSCH CC   CE
  N FB – CE + N FB – CC
 CE 
TX i  ic  TX i  ic  EPRE DLRS CC TX i  ic  TX i  ic  EPRE
DLRS CE
P DLRS CC = P DLRS  -----------------------------
TX i  ic 
- and P DLRS  CE = P DLRS  -----------------------------
TX i  ic 
-
EPRE DLRS EPRE DLRS
EPRE PDCCH CC TX i  ic  TX i  ic  EPRE PDCCH CE
P PDCCH CC = P PDCCH  ---------------------------------
TX i  ic 
- and P PDCCH  CE = P PDCCH  ---------------------------------
TX i  ic 
EPRE PDCCH EPRE PDCCH
492
i i
TX  ic  TX  ic  EPRE PDSCH CC TX  ic  TX  ic  EPRE PDSCH CE
i i i i
P PDSCH CC = P PDSCH  --------------------------------
TX  ic 
- and P PDSCH  CE = P PDSCH  --------------------------------
TX  ic 
-
i i
EPRE PDSCH EPRE PDSCH
TX  ic  EPRE CE
i
In the above,  CE is the cell-edge power boost for cell TXi(ic)’s frame configuration. By definition:  CE = -----------------
-
EPRE CC
TX i  ic 
TX i  ic  TX i  ic  N FB – CC
If  CE is left empty, it is automatically calculated as follows:  CE = -----------------
TX i  ic 
N FB – CE
N FB – CC and N FB – CE are respectively the numbers of frequency blocks in cell centre and cell-edge of TXi(ic).
Number of frequency blocks in

ICIC mode
Cell centre Cell edge
No FFR N FB N FB
Time-switched FFR N FB N FB – CEx
Hard FFR N FB – CEx N FB – CEx
Soft FFR N FB – N FB – CEx N FB – CEx
TX i  ic  TX  ic  TX  ic  TX  ic  TX i  ic 
–  N FB – CE0 + N FB – CE1 + N FB – CE2
Partial soft FFR i i i
N FB N FB – CEx
 
TX i  ic  TX i  ic  TX i  ic  TX i  ic 
Where N FB – CEx can be N FB – CE0 , N FB – CE1 , or N FB – CE2 depending on the PSS ID of TXi(ic).
Output
TX  ic 
i
• EPRE DLRS : Energy per resource element of the downlink reference signals for cell TXi(ic).
TX i  ic 
• EPRE SS : Energy per resource element of the SS for cell TXi(ic).
TX i  ic 
• EPRE PBCH : Energy per resource element of the PBCH for cell TXi(ic).
TX i  ic 
• EPRE PDCCH : Energy per resource element of the PDCCH for cell TXi(ic).
TX i  ic 
• EPRE PDSCH : Energy per resource element of the PDSCH for cell TXi(ic).
TX  ic 
i
• P DLRS : Instantaneous transmission power of the downlink reference signals for cell TXi(ic).
TX i  ic 
• P SS : Instantaneous transmission power of the SS for cell TXi(ic).
TX i  ic 
• P PBCH : Instantaneous transmission power of the PBCH for cell TXi(ic).
TX i  ic 
• P PDCCH : Average transmission power of the PDCCH for cell TXi(ic).
TX i  ic 
• P PDSCH : Average transmission power of the PDSCH for cell TXi(ic).

An LTE network can consist of cells that use different channel bandwidths. Therefore, the start and end frequencies of all the
channels may not exactly coincide. Channel bandwidths of cells can overlap each other with different ratios.
493
Figure 6.2: Co-Channel and Adjacent Channel Overlaps
The following sections describe how the co- and adjacent channel overlaps are calculated between the channels used by any
studied cell TXi(ic) and any other cell TXj(jc) of the network. In terms of interference calculation, the studied cell can be
considered a victim of interference received from the other cells that might be interfering the studied cell.
TX  ic 
i
If the studied cell is assigned a channel number N Channel , it receives co-channel interference on the channel bandwidth of
N Channel , and adjacent channel interference on the adjacent channel bandwidths, i.e., corresponding to N Channel – 1 and
TX i  ic 
N Channel + 1 .
In order to calculate the co- and adjacent channel overlaps between two channels, it is necessary to calculate the start and
end frequencies of both channels (explained in "Conversion From Channel Numbers to Start and End Frequencies" on
page 494). Once the start and end frequencies are known for the studied and other cells, the co- and adjacent overlaps and
the total overlap ratio are calculated as respectively explained in:
• "Co-Channel Overlap Calculation" on page 495.
• "Adjacent Channel Overlap Calculation" on page 496.
• "Total Overlap Ratio Calculation" on page 496.
6.4.2.1 Conversion From Channel Numbers to Start and End Frequencies

Input
• F Start – Band and F Start – Band : Start frequencies of the frequency bands assigned to the cells TXi(ic) and TXj(jc).
F Start – Band can be the start frequency of a TDD frequency band ( F Start – TDD ), or the uplink or the downlink start
frequency of an FDD frequency band ( F Start – FDD – UL or F Start – FDD – DL ).
First – TX i  ic  First – TX j  jc 
• N Channel and N Channel : First channel numbers the frequency band assigned to the cells TXi(ic) and TXj(jc).
• N Channel and N Channel : Channel numbers assigned to cells TXi(ic) and TXj(jc).
For FDD networks, Atoll considers that the same channel number is assigned to a cell in the downlink and uplink, i.e.,
the channel number you assign to a cell is considered for uplink and downlink both.
• W Channel and W Channel : Bandwidths of the channels assigned to cells TXi(ic) and TXj(jc).
• ICS Band and ICS Band : Inter-channel spacing of the frequency bands assigned to cells TXi(ic) and TXj(jc).
• CN Band and CN Band : Channel number step of the frequency bands assigned to cells TXi(ic) and TXj(jc).
494
Calculations
Channel numbers are converted into start and end frequencies as follows:
For cell TXi(ic):
TX  ic  TX  ic  TX  ic  TX  ic   N TXi  ic  – N First – TX i  ic 
+ W Channel + ICS Band   -------------------------------------------------------
  -
i i i i Channel Channel
F Start = F Start – Band   TX i  ic  
 CN Band 
F End = F Start + W Channel
For cell TXj(jc):
TX j  jc  TX j  jc  TX j  jc  TX j  jc   N TXj  jc  – N First – TX j  jc 
= F Start – Band + W Channel + ICS Band   -------------------------------------------------------
  -
Channel Channel
F Start   TX  jc 
j 
 CN Band 
TX j  jc  TX j  jc  TX j  jc 
Output
i j
• F Start and F Start : Start frequencies for the cells TXi(ic) and TXj(jc).
• F End and F End : End frequencies for the cells TXi(ic) and TXj(jc).
6.4.2.2 Co-Channel Overlap Calculation

Input
• F Start and F Start : Start frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel
Numbers to Start and End Frequencies" on page 494.
• F End and F End : End frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel
TX  ic 
i
• W Channel : Bandwidth of the channel assigned to the studied cell TXi(ic).
Calculations
Atoll first verifies that co-channel overlap exists between the cells TXi(ic) and TXj(jc).
Co-channel overlap exists if:
TX i  ic  TX j  jc  TX i  ic  TX j  jc 
F Start  F End AND F End  F Start
Otherwise there is no co-channel overlap.

Atoll calculates the bandwidth of the co-channel overlap as follows:
TX i  ic  – TX j  jc  TX  jc  TX  ic  TX  jc  TX  ic 
= Min  FEnd  F End  – Max  F Start  F Start 
j i j i
W CCO
   
The co-channel overlap ratio is given by:
TX  ic  – TX  jc 
TX i  ic  – TX j  jc  i j
W CCO
r CCO = ----------------------------------
-
TX j  jc 
W Channel
Output
• r CCO : Co-channel overlap ratio between the cells TXi(ic) and TXj(jc).
495
6.4.2.3 Adjacent Channel Overlap Calculation

Input
i j
TX i  ic 
Calculations
Atoll first verifies that adjacent channel overlaps exist between (the lower-frequency and the higher-frequency adjacent
channels of) the cells TXi(ic) and TXj(jc).
Adjacent channel overlap exists on the lower-frequency adjacent channel if:
TX i  ic  TX i  ic  TX j  jc  TX i  ic  TX j  jc 
F Start – W Channel  F End AND F Start  F Start
Adjacent channel overlap exists on the higher-frequency adjacent channel if:

TX i  ic  TX j  jc  TX i  ic  TX i  ic  TX j  jc 
F End  F End AND F End + W Channel  F Start
Otherwise there is no adjacent channel overlap.

Atoll determines the adjacent channel overlap ratio as follows:
Bandwidth of the lower-frequency adjacent channel overlap:
TX i  ic  – TX j  jc  TX  jc  TX  ic  TX  jc  TX  ic  TX  ic 
= Min  F End  F Start  – Max  F Start  F Start – W Channel
j i j i i
W ACO
L
The lower-frequency adjacent channel overlap ratio is given by:

TX i  ic  – TX j  jc  W ACO
L
r ACO = ----------------------------------
-
L TX j  jc 
W Channel
Bandwidth of the higher-frequency adjacent channel overlap:
TX i  ic  – TX j  jc  TX j  jc  TX i  ic  TX  ic  TX  jc  TX  ic 
= Min  F End  F End + W Channel – Max  F Start  F End 
i j i
W ACO
H    
The higher-frequency adjacent channel overlap ratio is given by:

H
r ACO = ----------------------------------
-
H TX  jc 
j
W Channel
The adjacent channel overlap ratio is given by:

TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc 
r ACO = r ACO + r ACO
L H
Output
TX  ic  – TX  jc 
i j
• r ACO : Adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc).
6.4.2.4 Total Overlap Ratio Calculation

Input
• r CCO : Co-channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co-Channel Overlap
496
TX  ic  – TX  jc 
i j
• r ACO : Adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Adjacent Channel
Overlap Calculation" on page 496.
TX  ic 
i
• f ACS : Adjacent channel suppression factor defined for the frequency band of the cell TXi(ic).
Calculations
The total overlap ratio is:

TX i  ic 
– f ACS
TX  ic  – TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc  -----------------------
i j i j i j 10
rO = r CCO + r ACO  10
Output
• rO : Total co- and adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc).
6.4.3 Subframe Pattern Collision Calculation

Subframe transmission and reception patterns can be defined for each cell using the Almost Blank Subframe (ABS) Pattern
field. The ABS pattern is a bit map, i.e., a series of 0’s and 1’s where each bit corresponds to one subframe. In an ABS pattern,
each 0 signifies a normal subframe and 1 implies an almost blank subframe. Almost blank subframes do not carry any traffic.
Only reference signals are transmitted over an ABS. For example, the ABS pattern "0100001000" means that subframes 1 and
6 are almost blank subframes whereas all the other subframes are normal subframes carrying traffic.
ABS patterns are used in conjunction with cell range expansion for eICIC (enhanced inter-cell interference coordination, also
known as time-domain ICIC) in an effort to minimise cell-edge interference between macro and small cells in heterogeneous
LTE networks (HetNets).
In order to calculate the collision between normal and almost blank subframes, the cells’ ABS patterns must be normalised,
and the used downlink, uplink, and special subframe patterns determined from the ABS patterns. These calculations are
respectively explained in:
• "Subframe Pattern Normalisation" on page 497.
• "Determination of Effective Subframe Patterns" on page 498.
• "Calculation of Subframe Collision Probabilities" on page 499.
6.4.3.1 Subframe Pattern Normalisation

Prior to the calculation of subframe collision probabilities, Atoll normalises the different ABS pattern lengths in order to
perform logical (bit by bit) AND and OR operations afterwards.
ABS patterns are normalised as follows:
1. The standard length of the ABS pattern of a cell is determined from its frequency band’s duplexing method and, in
case the cell’s frequency band is a TDD band, from the cell’s TDD frame configuration. The standard lengths of the ABS
pattern bit maps as defined by the 3GPP are as follows:
• FDD cells: 40 bits
• TDD cells using the frame configuration 0: 70 bits
• TDD cells using the frame configuration 1 through 5: 20 bits
• TDD cells using the frame configuration 6: 60 bits
The normalised ABS pattern length used in Atoll is 80 bits, which covers all the standard lengths. ABS patterns of
different lengths are normalised to 80 bits by Atoll.
2. The ABS pattern is corrected to match the standard ABS pattern lengths determined in step 1.:
• If the ABS pattern contains an asterisk, the pattern of 0’s and 1’s leading the asterisk is cyclically repeated until it
matches the standard ABS pattern length. Any 0’s and 1’s entered after an asterisk will be ignored.
FDD example: “0100010000*” = “0100010000010001000001000100000100010000”
• If the ABS pattern is shorter than the standard ABS pattern length, it is filled with 0’s to match the standard ABS
pattern length.
FDD example: “01000100000100010000” = “0100010000010001000000000000000000000000”
• If the ABS pattern is longer than the standard ABS pattern length, it is truncated to match the standard ABS pattern
length.
FDD example: “01000100000100010000010001000001000100000111110000” =
“0100010000010001000001000100000100010000”
497
• If the ABS pattern is empty, it means that there are no almost blank subframes defined and all the subframes can
carry traffic.
FDD example: NULL = “0” = “0*” = “0000000000000000000000000000000000000000”
3. The ABS pattern determined in step 2. is resized to 80 bits. More precisely, the ABS pattern is concatenated with itself.
Examples:
• FDD:
“0100010000010001000001000100000100010000” =
01000100000100010000010001000001000100000100010000010001000001000100000100010000
• TDD frame configuration 0:
“0100010000010001000001000100000100010000010001000001000100000100010000” =
01000100000100010000010001000001000100000100010000010001000001000100000100010000
• TDD frame configurations 1 through 5:
“01000100000100010000” =
01000100000100010000010001000001000100000100010000010001000001000100000100010000
• TDD frame configuration 6:
“010001000001000100000100010000010001000001000100000100010000” =
01000100000100010000010001000001000100000100010000010001000001000100000100010000
Once the ABS pattern SFP ABS has been normalised, it is inverted to determine the used subframe pattern SFP Used that is
used in further calculations:
SFP Used = !SFP ABS
6.4.3.2 Determination of Effective Subframe Patterns

Effective downlink, uplink, and special subframe patterns ( SFP DL , SFP UL , and SFP SSF ) are determined as follows by applying
masks ( SFM DL , SFM UL , and SFM SSF ) to the normalised used subframe patterns SFP Used determined as explained in
"Subframe Pattern Normalisation" on page 497:
SFP DL = SFP Used AND SFM DL
SFP UL = SFP Used AND SFM UL
SFP SSF = SFP Used AND SFM SSF
SFM DL , SFM UL , and SFM SSF are, respectively, the downlink, uplink and special subframe masks listed below:
FDD
SFM DL
SFM UL 11111111111111111111111111111111111111111111111111111111111111111111111111111111
SFM SSF
TDD frame configuration 0-DSUUU DSUUU

SFM DL 10000100001000010000100001000010000100001000010000100001000010000100001000010000
SFM UL 00111001110011100111001110011100111001110011100111001110011100111001110011100111
SFM SSF 01000010000100001000010000100001000010000100001000010000100001000010000100001000

TDD frame configuration 1-DSUUD DSUUD
SFM DL 10001100011000110001100011000110001100011000110001100011000110001100011000110001
SFM UL 00110001100011000110001100011000110001100011000110001100011000110001100011000110
SFM SSF 01000010000100001000010000100001000010000100001000010000100001000010000100001000

TDD frame configuration 2-DSUDD DSUDD
SFM DL 10011100111001110011100111001110011100111001110011100111001110011100111001110011
SFM UL 00100001000010000100001000010000100001000010000100001000010000100001000010000100
498
SFM SSF 01000010000100001000010000100001000010000100001000010000100001000010000100001000

TDD frame configuration 3-DSUUU DDDDD
SFM DL 10000111111000011111100001111110000111111000011111100001111110000111111000011111
SFM UL 00111000000011100000001110000000111000000011100000001110000000111000000011100000
SFM SSF 01000000000100000000010000000001000000000100000000010000000001000000000100000000

TDD frame configuration 4-DSUUD DDDDD
SFM DL 10001111111000111111100011111110001111111000111111100011111110001111111000111111
SFM UL 00110000000011000000001100000000110000000011000000001100000000110000000011000000
SFM SSF 01000000000100000000010000000001000000000100000000010000000001000000000100000000

TDD frame configuration 5-DSUDD DDDDD
SFM DL 10011111111001111111100111111110011111111001111111100111111110011111111001111111
SFM UL 00100000000010000000001000000000100000000010000000001000000000100000000010000000
SFM SSF 01000000000100000000010000000001000000000100000000010000000001000000000100000000

TDD frame configuration 6-DSUUU DSUUD
SFM DL 10000100011000010001100001000110000100011000010001100001000110000100011000010001
SFM UL 00111001100011100110001110011000111001100011100110001110011000111001100011100110
SFM SSF 01000010000100001000010000100001000010000100001000010000100001000010000100001000
6.4.3.3 Calculation of Subframe Collision Probabilities

The probabilities of collision of subframes between a studied cell TXi(ic) and any interfering cell TXj(jc) are calculated as
follows.
• In the following equations, the operator  X implies the sum of 1’s in a given
1
series of bits, X.
• In the following equations, AND and OR are logical bit-by-bit operators.
Method 1: ABS Patterns Used Only at Cell Edges

By default, ABS patterns are considered only to be used for serving users at cell edges. This means that all subframes are
considered non-ABS subframes in the cell centre. This is equivalent to setting the following Atoll.ini option:
[LTE]
UseABSonCellEdgeOnly = 1
Different collision probabilities are calculated depending on the location of the served pixel, subscriber, or mobile in cell
TXi(ic):
• Subframe collision between cell centre of TXi(ic) and cell centre of TXj(jc):
 TX  ic  TX  ic  TX  jc  TX  jc  
   SFMDL   OR SFMSSF  
i i j j
OR SFM SSF  AND  SFM DL
TX i  ic  – TX j  jc  
p ABS – DL – CC = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1 -
TX  ic  TX i  ic 
 SFM i 
  DL OR SFM SSF

1
 TX i  ic  TX i  ic  TX  jc  TX  jc  
   SFMUL  AND  SFM j OR SFMSSF  
j
OR SFM SSF
  UL 
TX i  ic  – TX j  jc  
p ABS – UL – CC = ------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1 -
 SFM 
  UL OR SFM SSF 
1
499
• Subframe collision between cell edge of TXi(ic) and cell centre of TXj(jc):
   SFPDL  AND  SFM DL OR SFM SSF  
i i j j
OR SFPSSF
  
TX  ic  – TX  jc 
i j 
p ABS – DL – CC = --------------------------------------------------------------------------------------------------------------------------------------------------------------------
1 -
 SFP i 

i
OR SFPSSF 
 DL
1
   SFPUL   OR SFM SSF  
i i j j
OR SFPSSF  AND  SFM UL
TX i  ic  – TX j  jc  
p ABS – UL – CC = --------------------------------------------------------------------------------------------------------------------------------------------------------------------
1 -
 SFP i 
  UL OR SFPSSF

1
• Subframe collision between cell centre of TXi(ic) and cell edge of TXj(jc):
   SFMDL   OR SFP SSF  
i i j j
OR SFM SSF  AND  SFP DL
TX i  ic  – TX j  jc  
p ABS – DL – CE = --------------------------------------------------------------------------------------------------------------------------------------------------------------------
1 -
 SFM 
  DL OR SFMSSF

1
   SFMUL  AND  SFP j OR SFP SSF  
j
OR SFM SSF
  UL 
TX i  ic  – TX j  jc  
p ABS – UL – CE = --------------------------------------------------------------------------------------------------------------------------------------------------------------------
1 -
 SFM i 

i
OR SFMSSF 
 UL
1
• Subframe collision between cell edge of TXi(ic) and cell edge of TXj(jc):
 TX i  ic  TX i  ic  TX j  jc  TX j  jc  
   SFP DL OR SFPSSF  AND  SFP DL OR SFP SSF  
  
TX i  ic  – TX j  jc   
p ABS – DL – CE = ----------------------------------------------------------------------------------------------------------------------------------------------------------------
1 -
 SFP 
  DL OR SFP SSF 
1
   SFPUL  AND  SFP j OR SFP SSF  
j
OR SFPSSF   UL
TX i  ic  – TX j  jc  
p ABS – UL – CE = ----------------------------------------------------------------------------------------------------------------------------------------------------------------
1 -
 SFP i 
  UL OR SFP SSF

1
This method enables you to include the cell-edge traffic ratio in the calculation of interference. The downlink interference
reduction factor due to subframe collisions for any served pixel, subscriber, or mobile in cell TXi(ic) is calculated as follows:
TX i  ic  – TX j  jc  TX  jc  TX  ic  – TX  jc  TX  jc  TX  ic  – TX  jc 
= 10  Log  r DL – CE  p ABS – DL – CE +  1 – r DL – CE  p ABS – DL – CC 
j i j j i j
f ABS – DL
   
The uplink interference reduction factor due to subframe collisions for any served pixel, subscriber, or mobile in cell TXi(ic) is
TX i  ic  – TX j  jc  TX  ic  – TX  jc  TX  ic  – TX j  jc  TX  ic  – TX  jc 
= 10  Log  p ABS – UL – CE  or f ABS – UL = 10  Log  p ABS – UL – CC 
i j i i j
f ABS – UL
   
Method 2: ABS Patterns Used Throughout the Cell
If you wish to apply the ABS patterns throughout the cell, irrespective of the cell-edge area and the cell-edge traffic ratio, you
can do so by adding the following lines in the Atoll.ini file:
[LTE]
UseABSonCellEdgeOnly = 0
The following collision probabilities are calculated between cells TXi(ic) and TXj(jc):
500
   SFPDL   OR SFPSSF  
i i j j
OR SFP SSF  AND  SFP DL
TX  ic  – TX  jc 
i j 
p ABS – DL = ----------------------------------------------------------------------------------------------------------------------------------------------------------------
1 -
 SFP i 

i
OR SFP SSF
 DL 
1
   SFPUL  AND  SFP j OR SFPSSF  
j
OR SFP SSF   UL
TX i  ic  – TX j  jc  
p ABS – UL = ----------------------------------------------------------------------------------------------------------------------------------------------------------------
1 -
 SFP 
  UL OR SFP SSF

1
The downlink interference reduction factor due to subframe collisions for any pixel, subscriber, or mobile is calculated as
follows:
TX i  ic  – TX j  jc  TX i  ic  – TX j  jc 
f ABS – DL = 10  Log  p ABS – DL 
 
The uplink interference reduction factor due to subframe collisions for any pixel, subscriber, or mobile is calculated as follows:
f ABS – UL = 10  Log  p ABS – UL 
 
6.4.4 Signal Level and Signal Quality Calculations

These calculations include the calculation of the received signal levels, and noise and interference. The following sections
describe how the received signal levels, the noise and interference, C/N, and C/(I+N) ratios are calculated in Atoll:
• "Noise Calculation (DL)" on page 505.
• "Interference Calculation (DL)" on page 505.
• "C/N Calculation (DL)" on page 516.
• "Noise Calculation (UL)" on page 525.
• "Interference Calculation (UL)" on page 525.
• "Noise Rise Calculation (UL)" on page 528.
• "C/N Calculation (UL)" on page 529.
6.4.4.1 Signal Level Calculation (DL)

Input
TX i  ic 
• P Max : Max power of the cell TXi(ic).
TX  ic 
i
• P DLRS : Transmission power of the downlink reference signals for cell TXi(ic) as calculated in "Downlink Transmission
Power Calculation" on page 485.
TX  ic 
i
• P SS : Transmission power of the SS for cell TXi(ic) as calculated in "Downlink Transmission Power Calculation" on
page 485.
TX  ic 
i
• P PBCH : Transmission power of the PBCH for cell TXi(ic) as calculated in "Downlink Transmission Power Calculation"
on page 485.
TX i  ic 
• P PDCCH : Transmission power of the PDCCH for cell TXi(ic) as calculated in "Downlink Transmission Power Calculation"
on page 485.
TX i  ic 
• P PDSCH : Transmission power of the PDSCH for cell TXi(ic) as calculated in "Downlink Transmission Power Calculation"
on page 485.
TX i  ic 
• EPRE DLRS : Energy per resource element of the downlink reference signals for cell TXi(ic) as calculated in "Downlink
Transmission Power Calculation" on page 485.
501
TX  ic 
i
• EPRE SS : Energy per resource element of the SS for cell TXi(ic) as calculated in "Downlink Transmission Power
TX  ic 
i
• EPRE PBCH : Energy per resource element of the PBCH for cell TXi(ic) as calculated in "Downlink Transmission Power
TX i  ic 
• EPRE PDCCH : Energy per resource element of the PDCCH for cell TXi(ic) as calculated in "Downlink Transmission Power
TX i  ic 
• EPRE PDSCH : Energy per resource element of the PDSCH for cell TXi(ic) as calculated in "Downlink Transmission Power
TX i
• E SA : Number of antenna elements defined for the smart antenna equipment used by the transmitter TXi.
Array
• G SA : Smart antenna array gain offset defined per clutter class.
Combining
• G SA : Smart power combining gain offset defined per clutter class.
Div
• G SA : Smart antenna diversity gain (for cross-polarised smart antennas) defined per clutter class.
TX i
• G Ant : Transmitter antenna gain for the antenna used by the transmitter TXi.
TX i
• G SA    : Smart antenna gain in the direction  of the served pixel, subscriber, or mobile Mi. For more information on
TX i
the calculation of G SA    , see "Beamforming Smart Antenna Models" on page 43.
TX i
Array
The smart antenna gain ( G SA    ) and the smart antenna array gain offset ( G SA ) are
applied only if the AAS criterion (RS C/N, RS C/(I+N), or PDSCH C/(I+N)) is less than the
DL
AAS threshold ( T AAS ) defined in the properties of the reception equipment used by the
pixel, subscriber, or mobile Mi.
TX i TX i
• L : Total transmitter losses for the transmitter TXi ( L = L Total – DL ).
• L Model : Loss on the transmitter-receiver path (path loss) calculated using a propagation model.
TX i
• L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the transmitter TXi.
• M Shadowing – Model : Shadowing margin based on the model standard deviation.
In coverage predictions, shadowing margins are taken into account when the option "Shadowing taken into account"
is selected.
• L Indoor : Indoor losses taken into account when the option "Indoor coverage" is selected.
M
i
• L : Receiver terminal losses for the pixel, subscriber, or mobile Mi.
M
i
• G : Receiver terminal’s antenna gain for the pixel, subscriber, or mobile Mi.
M
i
• L Ant : Receiver terminal’s antenna attenuation calculated for the pixel, subscriber, or mobile Mi.
Mi
For calculating the useful signal level from the best serving cell, L Ant is determined in the direction (H,V) = (0,0) from
Mi
the antenna patterns of the antenna used by Mi. For calculating the interfering signal level from any interferer, L Ant
is determined in the direction of the interfering cell from the antenna patterns of the antenna used by Mi, while the
antenna is pointed towards Mi’s best serving cell.
Mi
• L Body : Body loss defined for the service used by the pixel, subscriber, or mobile Mi.
Mi Mi Mi Mi
L , G , L Ant , and L Body are not used in the calculations performed for the point
analysis tool’s profile tab, delta path loss calculation, and the downlink reference signal
level based coverage predictions.
502
TX  ic 
i
• D CP : Cyclic prefix duration defined in the TXi(ic) frame configuration or, otherwise, in the global network settings.
Calculations
The received signal levels (dBm) from any cell TXi(ic) are calculated for a pixel, subscriber, or mobile Mi as follows:
TX i  ic  TX i  ic  Mi Mi Mi Mi
C Max = EIRP Max – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body + f CP
TX i  ic  TX i  ic  TX i TX i
Without smart antennas: EIRP Max = P Max + G Ant – L
TX i  ic  TX i  ic  TX i TX i TX
+ 10  Log  E SA  + G SA
i Combining Div
With smart antennas: EIRP Max = P Max + G Ant – L + G SA
C DLRS = EIRP1 DLRS – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body + f CP
Without smart antennas: EIRP1 DLRS = P DLRS + G Ant – L
TX i  ic  TX i  ic  TX i TX i TX i
+ 10  Log  E SA  + G SA
Combining
With smart antennas: EIRP1 DLRS = P DLRS + G Ant – L
 
TX  ic  TX  ic  M M M M
i i i i i i
C SS = EIRP1 SS – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body + f CP
Without smart antennas: EIRP1 SS = P SS + G Ant – L
+ 10  Log  E SA  + G SA
Combining Div
With smart antennas: EIRP1 SS = P SS + G Ant – L + G SA
 
C PBCH = EIRP1 PBCH – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body + f CP
Without smart antennas: EIRP1 PBCH = P PBCH + G Ant – L
TX  ic  TX  ic  TX TX TX
+ 10  Log  E SA  + G SA
i i i i i Combining Div
With smart antennas: EIRP1 PBCH = P PBCH + G Ant – L + G SA
C PDCCH = EIRP1 PDCCH – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body + f CP
Without smart antennas: EIRP1 PDCCH = P PDCCH + G Ant – L
+ 10  Log  E SA  + G SA
Combining Div
With smart antennas: EIRP1 PDCCH = P PDCCH + G Ant – L + G SA
 
C PDSCH = EIRP1 PDSCH – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body + f CP
Without smart antennas: EIRP1 PDSCH = P PDSCH + G Ant – L
Array Combining Div
With smart antennas: EIRP1 PDSCH = P PDSCH + G SA    + G SA + G SA + G SA – L
The energy per resource element (dBm/Sym) received from any cell TXi(ic) are calculated for a pixel, subscriber, or mobile Mi
as follows:
RSRP: E DLRS = EIRP2 DLRS – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body + f CP
Without smart antennas: EIRP2 DLRS = EPRE DLRS + G Ant – L
+ 10  Log  E SA  + G SA
i i i i i Combining
With smart antennas: EIRP2 DLRS = EPRE DLRS + G Ant – L
503
TX  ic  TX  ic  M M M M
i i i i i i
E SS = EIRP2 SS – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body + f CP
TX  ic  TX  ic  TX TX
i i i i
Without smart antennas: EIRP2 SS = EPRE SS + G Ant – L
+ 10  Log  E SA  + G SA
With smart antennas: EIRP2 SS = EPRE SS + G Ant – L + G SA
 
E PBCH = EIRP2 PBCH – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body + f CP
Without smart antennas: EIRP2 PBCH = EPRE PBCH + G Ant – L
+ 10  Log  E SA  + G SA
With smart antennas: EIRP2 PBCH = EPRE PBCH + G Ant – L + G SA
E PDCCH = EIRP2 PDCCH – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body + f CP
i i i i
Without smart antennas: EIRP2 PDCCH = EPRE PDCCH + G Ant – L
+ 10  Log  E SA  + G SA
Combining Div
With smart antennas: EIRP2 PDCCH = EPRE PDCCH + G Ant – L + G SA
 
E PDSCH = EIRP2 PDSCH – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body + f CP
Without smart antennas: EIRP2 PDSCH = EPRE PDSCH + G Ant – L
Array Combining Div
With smart antennas: EIRP2 PDSCH = EPRE PDSCH + G SA    + G SA + G SA + G SA – L
In the above, L Path is the path loss (dB) calculated as follows:
TX i
L Path = L Model + L Ant
Furthermore, the total losses between the cell and the pixel, subscriber, or mobile Mi can be calculated as follows:
TX TX M M M M
i i i i i i
L Total = L Path + L + L Indoor + M Shadowing – Model – G +L –G + L Ant + L Body
f CP is the cyclic prefix factor, i.e., the ratio of the useful symbol energy to the total symbol energy.
The total symbol duration of a modulation symbol comprises the useful symbol duration, carrying the actual data bits, and a
cyclic prefix, added to the useful data bits as padding against multi-path to avoid inter-symbol interference. Hence, the total
energy within a modulation symbol belongs in part to the useful data bits and in part to the cyclic prefix. Once a modulation
symbol is received, only the energy of the useful data bits can be used for extracting the data. The energy belonging to the
cyclic prefix is lost once it has served its purpose of combatting inter-symbol interference. Therefore, f CP implies that the
energy belonging to the cyclic prefix is excluded from the useful signal level.
 TX  ic 
 10  Log  7  7.5  If D CPi = Normal

f CP =  TX i  ic 
 10  Log  6  7.5  If D CP = Extended

 0 If TX i  ic  is an interferer
The cyclic prefix energy and the useful data bits energy are both taken into account when calculating interfering signal levels.
Output
TX i  ic 
• C Max : Received max signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX i  ic 
• C DLRS : Received downlink reference signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX i  ic 
• C SS : Received SS signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
504
TX  ic 
i
• C PBCH : Received PBCH signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX  ic 
i
• C PDCCH : Received PDCCH signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX  ic 
i
• C PDSCH : Received PDSCH signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX i  ic 
• E DLRS : Received downlink reference signal energy per resource element from the cell TXi(ic) at the pixel, subscriber,
or mobile Mi.
TX i  ic 
• E SS : Received SS energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX i  ic 
• E PBCH : Received PBCH energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX i  ic 
• E PDCCH : Received PDCCH energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX  ic 
i
• E PDSCH : Received PDSCH eneregy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
• L Path : Path loss between the cell TXi(ic) and the pixel, subscriber, or mobile Mi.
• L Total : Total losses between the cell TXi(ic) and the pixel, subscriber, or mobile Mi.
6.4.4.2 Noise Calculation (DL)

For determining the C/N and C/(I+N), Atoll calculates the downlink noise which comprises thermal noise and the noise figure
of the equipment. The thermal noise density depends on the temperature, i.e., it remains constant for a given temperature.
However, the value of the thermal noise varies with the used bandwidth.
Input
• K: Boltzmann’s constant.
• T: Temperature in Kelvin.
Mi
• nf : Noise figure of the terminal used for calculations by the pixel, subscriber, or mobile Mi.
Calculations
The power spectral density of thermal noise is calculated as follows:

n 0 = 10  Log  K  T  1000  = – 174 dBm/Hz
The thermal noise for one resource element, i.e., over one subcarrier, is calculated as follows:
TX i  ic 
n 0 – Sym = n 0 + 10  Log  F 
The downlink noise is the sum of the thermal noise and the noise figure of the terminal used for the calculations by the pixel,
subscriber, or mobile Mi. The downlink noise for one resource element, i.e., over one subcarrier, is calculated as follows:
n Sym = n 0 – Sym + nf
Output
TX i  ic 
• n Sym : Downlink noise for one subcarrier.
6.4.4.3 Interference Calculation (DL)

The interference received by any pixel, subscriber, or mobile, served by a cell TXi(ic) from other cells TXj(jc) can be defined as
the signal levels received from interfering cells TXj(jc) depending on the overlap that exists between the channels used by the
cells TXi(ic) and TXj(jc), on the traffic loads of the interfering cells TXj(jc), and whether the cells support ICIC or not.
Input
TX  jc 
j
• E DLRS : Received downlink reference energy per resource element received from any interfering cell TXj(jc) as
calculated in "Signal Level Calculation (DL)" on page 501 at the pixel, subscriber, or mobile Mi covered by the cell
TXi(ic).
505
TX  jc 
j
• E SS : Received SS energy per resource element received from any interfering cell TXj(jc) as calculated in "Signal
Level Calculation (DL)" on page 501 at the pixel, subscriber, or mobile Mi covered by the cell TXi(ic).
TX  jc 
j
• E PBCH : Received PBCH energy per resource element received from any interfering cell TXj(jc) as calculated in "Signal
TX j  jc 
• E PDCCH : Received PDCCH energy per resource element received from any interfering cell TXj(jc) as calculated in "Signal
TX j  jc 
• E PDSCH : Received PDSCH energy per resource element received from any interfering cell TXj(jc) as calculated in "Signal
TX j
• G SA    : Smart antenna gain in the direction  . For more information, see "Beamforming Smart Antenna Models" on
page 43.
TX j
• G SA    : Smart antenna gain in the direction  calculated from the average array correlation matrix:
H
G SA    = g n     S   R Avg  S  . For more information, see "Beamforming Smart Antenna Models" on page 43.
• M Shadowing – C  I : Shadowing margin based on the C/I standard deviation.
In Monte Carlo simulations, the received energies per resource element from interferers already include
M Shadowing – Model , as explained in "Signal Level Calculation (DL)" on page 501.
In coverage predictions, the ratio M Shadowing – Model – M Shadowing – C  I is applied to the interfering signals (for more
information, see "Shadow Fading Model" on page 90). As the received energies per resource element from interferers
already include M Shadowing – Model , M Shadowing – C  I is added to the received energies per resource element from
interferers in order to achieve the ratio M Shadowing – Model – M Shadowing – C  I :
TX  jc  TX  jc 
j j
E = E + M Shadowing – C  I
is selected.
TX j  jc 
• N Sym – DLRS : Number of downlink reference signal resource elements as calculated in "Downlink Transmission Power
• N Sym – SS : Number of SS resource elements as calculated in "Downlink Transmission Power Calculation" on page 485.
TX j  jc 
• N Sym – PBCH : Number of PBCH resource elements as calculated in "Downlink Transmission Power Calculation" on
page 485.
TX j  jc 
• N Sym – PDCCH : Number of PDCCH resource elements as calculated in "Downlink Transmission Power Calculation" on
page 485.
TX j  jc 
• N Sym – PDSCH : Number of PDSCH resource elements as calculated in "Downlink Transmission Power Calculation" on
page 485.
TX j  jc 
• N Sym – DL : Total number of downlink resource elements as calculated in "Downlink Transmission Power Calculation"
on page 485.
• rO : Total channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co- and Adjacent
Channel Overlaps Calculation" on page 493.
TX j  jc 
• TL DL : Downlink traffic load of the interfering cell TXj(jc).
Traffic loads can either be calculated using Monte Carlo simulations, or entered manually for each cell. Calculation of
traffic loads is explained in "Simulation Process" on page 479.
TX j  jc 
• AU DL : Downlink AAS usage of the interfering cell TXj(jc).
• W FB : Width of a frequency block in the frequency domain (180 kHz).
506
i j
i j
• F Start and F Start : Start frequencies of the channels assigned to the cells TXi(ic) and TXj(jc) calculated as explained
in "Conversion From Channel Numbers to Start and End Frequencies" on page 494.
• ID  and ID  : Physical cell IDs of the cells TXi(ic) and TXj(jc).
• r DL – CE and r DL – CE : Downlink cell-edge traffic ratios of the cells TXi(ic) and TXj(jc).
TX i  ic 
TX  jc 
j
• N Ant – TX : Number of transmission (downlink) antenna ports defined for the cell TXj(jc).
Calculations
Two interference calculation methods exist in Atoll.
TX j  jc  TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc 
• Calculations of f MIMO , f TL , f ICIC – DL , f PDCCH , and f PDSCH are
explained at the end of this section.
TX j  jc 
• In the calculations below, E DLRS is weighted by the downlink subframe collision
probability f ABS – DL when the relevant option is set in the Atoll.ini file:
[LTE]
eICIConRS = 1
Method 1: Synchronised Transmission and Reception

Atoll calculates the interference between two cells using this method when:
• The frequency channels assigned to the interfered and interfering cells have the same centre frequency, and
• The interfered and interfering cells both have an even number of frequency blocks or both have an odd number of
frequency blocks, and
• The Atoll.ini file does not contain the following option:
[LTE]
SameItf_PDSCH_RS_PDCCH = 1
Synchronised transmission and reception means that the OFDM symbols of the interfered and interfering frames overlap and
match each other in time.
The interfering energy per resource element (dBm/Sym) received over downlink reference signals from any cell TXj(jc) at a
pixel, subscriber, or mobile Mi is calculated as follows:
• RS of the interfered cell TXi(ic) collide only with RS of the interfering cell TXj(jc)
TX i  ic  TX j  jc  TX j  jc  TX i  ic 
This occurs when ID PSS = ID PSS and N Ant – TX  N Ant – TX
For the calculation of the probability of collision, here N Ant – TX = Min  4 NAnt – TX  .
507
TX  jc 
 TX  ic  E
j 
TX  jc  N i ---------------------
DLRS
TX  ic  – TX  jc 
= 10  Log  -------------------  +f i
j Ant – TX 10 j
 DLRS  10
TX  jc 
j  O
 N Ant – TX 
 
• RS of the interfered cell TXi(ic) collide with RS, PDCCH, and PDSCH of the interfering cell TXj(jc)
This occurs when ID PSS = ID PSS and N Ant – TX  N Ant – TX
For the calculation of the probability of collision, here N Ant – TX = Min  4 N Ant – TX  .
With 1 or 2 antenna ports:
TX  jc 
j
TX j  jc  N j --------------------- TX i  ic  – TX j  jc 
= 10  Log  -------------------
Ant – TX 10 
 DLRS  10
TX i  ic   + fO
 
TX  jc  TX  ic  – TX  jc  TX  jc  TX  ic  – TX  jc 
 j
E PDCCH + f PDCCH
i j j
E PDSCH + f PDSCH
i j 
 TXi  ic  TX j  jc  ----------------------------------------------------------------------- -----------------------------------------------------------------------
 N Ant – TX – N Ant – TX 10 10
+ 3  10
10 
+ 10  L og  ---------------------------------------------
TX i  ic 
 ------------------------------------------------------------------------------------------------------------------------------------
 4 
 
TX  ic 
i
With 4 or 8 antenna ports and N SD – PDCCH = 1 :
TX  jc 
 TX  jc  E
j 
TX j  jc  N j ---------------------
DLRS
= 10  Log  -------------------
Ant – TX 10 
 DLRS  10 + fO
 TXi  ic  
 
 j
E PDCCH + f PDCCH
i j j
E PDSCH + f PDSCH
i j 
 TXi  ic  TX j  jc  ----------------------------------------------------------------------- -----------------------------------------------------------------------
 N Ant – TX – N Ant – TX 10 10
+ 5  10
10 
+ 10  L og  ---------------------------------------------
TX i  ic 
 ------------------------------------------------------------------------------------------------------------------------------------
 6 
 
TX i  ic 
With 4 or 8 antenna ports and N SD – PDCCH  1 :
TX  jc 
j
TX j  jc  N j --------------------- TX i  ic  – TX j  jc 
= 10  Log  -------------------
Ant – TX 10 
 DLRS  10
TX i  ic   + fO
 
 j
E PDCCH + f PDCCH
i j j
E PDSCH + f PDSCH
i j 
 TXi  ic  TX j  jc  ----------------------------------------------------------------------- -----------------------------------------------------------------------
 N Ant – TX – N Ant – TX 10 10
+ 2  10
10 
+ 10  L og  ---------------------------------------------
TX i  ic 
 ------------------------------------------------------------------------------------------------------------------------------------
 3 
 
• RS of the interfered cell TXi(ic) collide only with PDCCH and PDSCH of the interfering cell TXj(jc)
TX i  ic  TX j  jc  TX i  ic  TX j  jc  TX i  ic  TX j  jc 
This occurs when ( ID PSS = ID PSS and  Shift =  Shift  3 and N Ant – TX = N Ant – TX = 1 ) OR
ID PSS  ID PSS
With 1 or 2 antenna ports:
508
 EPDCCH j
+f
i j
E
j
+f
i j 
 -----------------------------------------------------------------------
PDCCH
-----------------------------------------------------------------------
PDSCH PDSCH
TX  jc 
j  10 10
+ 3  10
10  TX  ic  – TX  jc 
i j
 DLRS = 10  Log  ------------------------------------------------------------------------------------------------------------------------------------ + f O
 4 
 
 
TX i  ic 
With 4 or 8 antenna ports and N SD – PDCCH = 1 :
 EPDCCH j
+ f PDCCH
i j j
E PDSCH + f PDSCH
i j 
 ----------------------------------------------------------------------- -----------------------------------------------------------------------
TX j  jc   10 10
+ 5  10
10  TX i  ic  – TX j  jc 
 DLRS = 10  Log  ------------------------------------------------------------------------------------------------------------------------------------ + f O
 6 
 
 
TX i  ic 
With 4 or 8 antenna ports and N SD – PDCCH  1 :
 EPDCCH j
+ f PDCCH
i j j
E PDSCH + f PDSCH
i j 
 ----------------------------------------------------------------------- -----------------------------------------------------------------------
TX j  jc   10 10
+ 2  10
10  TX i  ic  – TX j  jc 
 DLRS = 10  Log  ------------------------------------------------------------------------------------------------------------------------------------ + f O
 3 
 
 
The interfering energy per resource element (dBm/Sym) received over the SS and the PBCH from any cell TXj(jc) at a pixel,
subscriber, or mobile Mi is calculated as follows:
j j
 ESS E PBCH 
 --------------------
10
- ---------------------
10
TX j  jc  
TX j  jc   10  N Sym – SS + 10  N Sym – PBCH TX  ic  – TX j  jc  TX j  jc 
 SS PBCH - + f O i
= 10  Log  ------------------------------------------------------------------------------------------------------------- + f MIMO
TX j  jc 
 N Sym – SS + N Sym – PBCH 
 
 
The interfering energy per resource element (dBm/Sym) received over the PDCCH from any cell TXj(jc) at a pixel, subscriber,
or mobile Mi is calculated as follows:
• PDCCH of the interfered cell TXi(ic) collides with PDCCH and all the RS of the interfering cell TXj(jc)
TX  ic  TX  jc  TX  ic  TX  jc  TX  ic  TX  jc 
i j i j i j
i j
ID PSS  ID PSS
For the calculation of the probability of collision, here N Ant – TX = Min  4 NAnt – TX  .
TX  jc 
 TX j  jc  E
j 
TX j  jc   N
DLRS 
--------------------
- TX  ic  – TX j  jc 
= 10  Log  -------------------  +f i
1 Sym – DLRS in PDCCH 10
 PDCCH -  10
 -----------------------------------------
 TX j  jc  TX i  ic   O
 N Ant – TX N Sym – PDCCH 
 
TX  jc  TX  ic  – TX  jc 
 TX  ic  TX j  jc 
j
E PDCCH + f PDCCH
i j 
N i – N -----------------------------------------------------------------------
+ 10  L og  ----------------------------------------------------------------------------- 
Sym – PDCCH Sym – DLRS in PDCCH 10
-  10
 TX i  ic  
 
Here, N Sym – DLRS in PDCCH is the number of downlink reference signal resource elements that fall within the PDCCH, and
N Sym – PDCCH is the number of PDCCH resource elements per frame.
• PDCCH of the interfered cell TXi(ic) collides with PDCCH and some RS of the interfering cell TXj(jc)
509
TX  jc 
 TX  ic  TX j  jc  TX i  ic  E DLRS 
j
TX j  jc  N i N Sym – DLRS in PDCCH – N Sym – DLRS in PDCCH -
-------------------- TX  ic  – TX j  jc 
= 10  Log  ------------------- -  10 10  + f O i
Ant – TX
 PDCCH  ------------------------------------------------------------------------------------------
 TX
j
 jc  TX
i
 ic  
 N Ant – TX N Sym – PDCCH 
 
TX  jc  TX  ic  – TX  jc 
 TX  jc  j
E PDCCH + f PDCCH
i j 
N j -----------------------------------------------------------------------
+ 10  L og  ----------------------------- 
Sym – PDCCH 10
 10
 TXi  ic  
 
Here, N Sym – DLRS in PDCCH is the number of downlink reference signal resource elements that fall within the PDCCH, and
N Sym – PDCCH is the number of PDCCH resource elements per frame.
• PDCCH of the interfered cell TXi(ic) collides only with PDCCH of the interfering cell TXj(jc)
TX j  jc  TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc 
 PDCCH = E PDCCH + f PDCCH + fO
The interfering energy per resource element (dBm/Sym) received over the PDSCH from any cell TXj(jc) at a pixel, subscriber,
• PDSCH of the interfered cell TXi(ic) collides with PDSCH and all the RS of the interfering cell TXj(jc)
TX  ic  TX  jc  TX  ic  TX  jc  TX  ic  TX  jc 
i j i j i j
ID PSS  ID PSS
TX  jc 
 TX  jc  E
j 
TX j  jc   N
j DLRS -
-------------------- TX i  ic  – TX j  jc 
 PDSCH  1
= 10  Log ------------------- Sym – DLRS in PDSCH
-  10
 -----------------------------------------
10 
+ fO
 TXj  jc  TX i  ic  
 N Ant – TX N Sym – PDSCH 
 
TX  jc  TX  ic  – TX  jc 
 TX  ic  TX j  jc  E
j
+f
i j 
N i – N -----------------------------------------------------------------------
PDSCH PDSCH
+ 10  L og  ----------------------------------------------------------------------------- 
Sym – PDSCH Sym – DLRS in PDSCH 10
 10
 TX i  ic  
 
Here, N Sym – DLRS in PDSCH is the number of downlink reference signal resource elements that fall within the PDSCH, and
N Sym – PDSCH is the number of PDSCH resource elements per frame.
• PDSCH of the interfered cell TXi(ic) collides with PDSCH and some RS of the interfering cell TXj(jc)
510
TX  jc 
 TX  ic  TX  jc  TX  ic  E
j 
TX  jc  N i N
j
– N
i ---------------------
DLRS
TX  ic  – TX  jc 
= 10  Log  -------------------  +f i
j Ant – TX Sym – DLRS in PDSCH Sym – DLRS in PDSCH 10 j
 PDSCH  -----------------------------------------------------------------------------------------
-  10
TX  jc 
j
TX  ic 
i  O
 N Ant – TX N Sym – PDSCH 
 
TX  jc  TX  ic  – TX  jc 
 TX  jc  j
E PDSCH + f PDSCH
i j 
N j -----------------------------------------------------------------------
+ 10  L og  ---------------------------- 
Sym – PDSCH 10
-  10
TX i  ic  
 
Here, N Sym – DLRS in PDSCH is the number of downlink reference signal resource elements that fall within the PDSCH, and
N Sym – PDSCH is the number of PDSCH resource elements per frame.
• PDSCH of the interfered cell TXi(ic) collides only with PDSCH of the interfering cell TXj(jc)
 PDSCH = E PDSCH + f PDSCH + fO
Method 2: Non-synchronised Transmission and Reception

Atoll calculates the interference between two cells using this method when:
• The frequency channels assigned to the interfered and interfering cells do not have the same centre frequency, or
• The interfered and interfering cells do not both have an even number of frequency blocks or do not both have an odd
number of frequency blocks, or
• The Atoll.ini file contains the following option:
[LTE]
SameItf_PDSCH_RS_PDCCH = 1
This method is also used for calculating the interference received from LTE cells of an external network in co-planning mode,
i.e., inter-technology interference received from LTE cells calculated using the inter-technology IRFs.
The interfering energy per resource element (dBm/Sym) received over downlink reference signals from any cell TXj(jc) at a
j
E PDCCH + f PDCCH
i j
TX  jc 
TX j  jc   --------------------
DLRS -
N ----------------------------------------------------------------------- N j
 DLRS 
= 10  Log 10
10 Sym – DLRS
 -------------------------
- + 10 10 Sym – PDCCH
 -----------------------------
 TX j  jc  TX j  jc 
 N Sym – DL N Sym – DL


-----------------------------------------------------------------------
10 N Sym – PDSCH  TX  ic  – TX j  jc 
+ 10 - + fO i
 ----------------------------
TX j  jc  
N Sym – DL 

The interfering energy per resource element (dBm/Sym) received over the SS and the PBCH from any cell TXj(jc) at a pixel,
TX j  jc  TX j  jc  TX j  jc  TX j  jc 
 ESS + f MIMO E PBCH + f MIMO 
 ----------------------------------------------
10
- -----------------------------------------------
10
TX j  jc  
TX j  jc   10  N Sym – SS + 10  N Sym – PBCH TX  ic  – TX j  jc 
 SS PBCH - + f O i
= 10  Log  ---------------------------------------------------------------------------------------------------------------------------------------------------
TX j  jc 
 N Sym – SS + N Sym – PBCH 
 
 
The interfering energy per resource element (dBm/Sym) received over the PDSCH and the PDCCH from any cell TXj(jc) at a
511
 E j TX  jc  E
j
+f
i j
TX  jc 
TX  jc   --------------------
DLRS
- N
j PDCCH PDCCH
----------------------------------------------------------------------- j
N Sym – PDCCH
j
 PDSCH 
= 10  Log  10
10 Sym – DLRS
 -------------------------
- + 10 10
 -----------------------------
j j

TX  jc  TX  ic  – TX  jc 
j
E PDSCH + f PDSCH
i j
TX j  jc 

-----------------------------------------------------------------------
+ 10 - + fO i
 ----------------------------
TX j  jc  
N Sym – DL 

j
E PDCCH + f PDCCH
i j
TX j  jc 
TX j  jc   --------------------
DLRS -
N Sym – DLRS ----------------------------------------------------------------------- N Sym – PDCCH
 PDCCH 
= 10  Log 10
10
 TX  jc  + 10
-------------------------
- 10
 -----------------------------
 j
TX j  jc 


-----------------------------------------------------------------------
+ 10 - + fO i
 ----------------------------
TX j  jc  
N Sym – DL 

E-UTRA carrier RSSI is measured on the OFDM symbols that contain reference signals. Therefore, the interfering energy per
frequency block (dBm/RB) received from any cell TXj(jc) at a pixel, subscriber, or mobile Mi over 1 frequency block during an
OFDM symbol carrying reference signals, is given as follows:
For cells using more than 1 antenna port, the encircled 10 in the formulas below is
replaced with 8.
 EPDSCH + fPDSCH
 ----------------------------------------------------------------------
10
- TX j  jc 
TX j  jc   10  N Sym – PDSCH
 RSSI = 10  Log  ---------------------------------------------------------------------------------------------
 10
 N + N
 Sym – PDSCH Sym – PDCCH

E
PDCCH
+f
PDCCH
TX j  jc  
-----------------------------------------------------------------------
10
TX j  jc  E DLRS 
10  N Sym – PDCCH ---------------------
10 
TX j  jc  

-  10 + 10
+ ---------------------------------------------------------------------------------------------  2  Min 2 N Ant – TX  + f O
TX j  jc  TX j  jc   
N Sym – PDSCH + N Sym – PDCCH 

Calculation of PDCCH and PDSCH Interference Weighting Factors

The PDCCH and PDSCH interference weighting factors ( f PDCCH and f PDSCH ) are calculated as follows:
 f MIMO + f TL + f ICIC – DL + f ABS – DL 
 TX j  jc  -
-------------------------------------------------------------------------------------------------------------------------------------------------
  1 – AU   10 10 
TX i  ic  – TX j  jc   DL  
f PDCCH =  
 TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc 
j i j i j 
 f TL + f ICIC – DL + f ABS – DL
-----------------------------------------------------------------------------------------------------------------------
-

 TX j  jc  10 
 + AU DL  10 
 j
f MIMO + f TL
j
+ f ICIC – DL
i j
+ f ABS – DL
i j

 TX j  jc  --------------------------------------------------------------------------------------------------------------------------------------------------

 1 – AU DL  10

TX i  ic  – TX j  jc     10 
f PDSCH =  
  TX j TX j  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  
 SA    –     + +
 
G G
SA 
f
ICIC – DL
f
ABS – DL 
 TX j  jc  --------------------------------------------------------------------------------------------------------------------------------------------------------- 
10
 + AU DL  10 
Calculation of MIMO/Antenna Diversity Interference Factors
512
TX  jc  TX  jc  TX  jc 
f MIMO is the interference increment due to more than one transmission antenna port: f MIMO = 10  Log  N Ant – TX
j j j
TX  jc 
j
If you do not wish to apply f MIMO , add the following lines in the Atoll.ini file:
[LTE]
MultiAntennaInterference = 0
MultiAntennaInterference is set to 1 by default.

Calculation of Interference Reduction Factors
TX  jc  TX  ic  – TX  jc 
j i j
Calculations for the interference reduction factors due to traffic load f TL , channel overlapping ( f O ), and static
downlink ICIC using fractional frequency reuse ( f ICIC – DL ) are explained below:
Interference reduction due to the traffic loads of the interfering cells:

Interference reduction due to the traffic loads of the interfering cells TXj(jc) is calculated as follows:
f TL = 10  Log  TL DL 
 
Interference reduction due to the co- and adjacent channel overlap between the studied and the interfering cells:
Interference reduction due to the co- and adjacent channel overlap between the cells TXi(ic) and TXj(jc) is calculated
as follows:
fO = 10  Log  r O 

Interference reduction due to static downlink ICIC using fractional frequency reuse:
The interference reduction factor due to static downlink ICIC using fractional frequency reuse for any pixel, subscriber,
TX i  ic  – TX j  jc  TX  ic  – TX j  jc 
= 10  Log  p Collision 
i
f ICIC – DL 
Whether a pixel, subscriber, or mobile Mi is in cell centre or cell-edge is determined as explained in "Best Server
Depending on the ICIC mode defined for the frame configuration of the cells TXi(ic) and TXj(jc), f ICIC – DL is
• If both TXi(ic) and TXj(jc) use time-switched FFR
The cell-edge and cell-centre resources are time-divided. Therefore, an interfered user may receive interference
from the cell-edge and cell-centre parts of the frame depending on time-domain switching points between the
cell-edge and cell-centre parts of the frames.
Atoll determines the switching point between the ICIC and the non-ICIC parts of the frame using the ICIC ratios.
The switching points between the ICIC and non-ICIC parts of the frame of the victim and interfering cells, TXi(ic)
and TXj(jc) respectively, are calculated as follows:
TX  ic  r DL – CE TX  jc  r DL – CE
i j
SP = ----------------------------------------------------------------------------
- and SP = ----------------------------------------------------------------------------
-
TX i  ic  TX i  ic  N TX j  jc  TX j  jc  N
r DL – CE +  1 – r DL – CE  ----------------- r DL – CE +  1 – r DL – CE  -----------------
FB – CE FB – CE
  TX i  ic    TX j  jc 
N FB N FB
Where, SP is the switching point between the ICIC and the non-ICIC parts of the frame, and r DL – CE is the downlink
cell-edge traffic ratios of the cells.
513
The ICIC ratio is used to partition the total downlink traffic load into ICIC and non-ICIC
parts of the frame. Therefore, the switching point formula is derived from the equation:
r DL – CE  TL DL  1 – r DL – CE   TL DL
---------------------------------------------------------- = -----------------------------------------------
-
N FB – CE  1 – SP   W Channel
SP  WChannel  -----------------
N FB
With cells using static downlink ICIC, there can be four different interference scenarios.
i. Between the ICIC part of the victim and the ICIC part of the interferer.
ii. Between the ICIC part of the victim and the non-ICIC part of the interferer.
iii. Between the non-ICIC part of the victim and the ICIC part of the interferer.
iv. Between the non-ICIC part of the victim and the non-ICIC part of the interferer.
Therefore, Atoll calculates the probabilities of collision for each scenario and weights the total interference
according to the total collision probability. The probability of collision p Coll for each scenario is:
Interfered cell Interfering cell

Case p Coll
Common
N FB – CE
i ICIC ICIC ---------------------
TX i  ic 
N FB – CE
ii ICIC Non ICIC 1

Common
N FB – CE
iii Non ICIC ICIC ---------------------
TX i  ic 
N FB
iv Non ICIC Non ICIC 1
TX i  ic 
Common
Where, N FB – CE is the number of cell-edge frequency blocks common in TXi(ic) and TXj(jc), and N FB – CE is the
number of cell-edge frequency blocks in the cell TXi(ic).
For a pixel, subscriber, or mobile Mi in the cell-edge of TXi(ic), the total collision probability for the pixel,
 TX j  jc  TX i  ic 
 i
p Coll If SP  SP

TX i  ic  – TX j  jc   TX  jc  TX  ic  TX  jc 
p Collision = i
+ p Coll   SP 
j ii i j
 p Coll  SP – SP  TX j  jc  TX i  ic 
 -------------------------------------------------------------------------------------------------------------
TX i  ic 
If SP  SP
 SP

For a pixel, subscriber, or mobile Mi in the cell centre of TXi(ic), the total collision probability for the pixel,
 iv
TX j  jc  TX i  ic 
 p Coll If SP  SP

TX i  ic  – TX j  jc   TX j  jc  TX  jc  TX  ic 
=  p iv   + p iii   SP j – SP i 
Coll   1 – SP
p Collision
 ---------------------------------------------------------------------------------------------------------------------------- If SP j  SP i
 Coll  TX  jc  TX  ic 
 TX  ic 
  1 – SP i 
  
• Other combinations of ICIC modes

Separate probabilities of collisions, p Collision , are calculated for cell-centre and cell-edge cases as follows:
TX i  ic  – TX j  jc  Common
N FB – CC
Cell centre: p Collision = --------------------
-
TX i  ic 
N FB – CC
514
TX  ic  – TX  jc  Common
i j N FB – CE
Cell-edge: p Collision = --------------------
-
TX  ic 
i
N FB – CE
Common Common
Where, N FB – CC is the number of common frequency blocks in TXi(ic) and TXj(jc) in cell centre, N FB – CE is the
number of common frequency blocks in TXi(ic) and TXj(jc) on cell-edge, N FB – CC and N FB – CE are respectively the
numbers of frequency blocks in cell centre and cell-edge of TXi(ic).

ICIC mode
No FFR N FB N FB
N FB N FB – CEx
 
Interference reduction due to subframe collision probabilities:

The interference reduction factor due to downlink subframe collision probabilities f ABS – DL is calculated as
explained in "Subframe Pattern Collision Calculation" on page 497.
Calculation of the Downlink Inter-technology Interference
The downlink inter-technology interference is calculated as follows:
TX
 k
P DL – Rec 
 ---------------------------------------

Inter – Tech
I DL =
 F  TX  ic  TX 
i k
TX k  ICP DL 
TX
k
Here P DL – Rec is the received downlink power from an interfering cell TXk belonging to another technology, and
F  TX i  ic  TX k 
ICPDL is the inter-technology downlink channel protection ratio for a frequency offset F between the interfered
and interfering frequency channels of TXi(ic) and TXk.
TX k
P DL – Rec is calculated based on the EIRP from GSM cells, total power from UMTS, CDMA2000, and TD-SCDMA cells, maximum
power from LTE cells, preamble power from WiMAX cells, and downlink cell power from Wi-Fi cells.
Output
TX j  jc 
•  DLRS : Interfering energy per resource element (dBm/Sym) received over downlink reference signals from any cell
TXj(jc) at a pixel, subscriber, or mobile Mi covered by a cell TXi(ic).
TX  jc 
j
•  SS PBCH : Interfering energy per resource element (dBm/Sym) received over the SS and the PBCH from any cell TXj(jc)
at a pixel, subscriber, or mobile Mi covered by a cell TXi(ic).
TX j  jc 
•  PDSCH : Interfering energy per resource element (dBm/Sym) received over the PDSCH from any cell TXj(jc) at a pixel,
subscriber, or mobile Mi covered by a cell TXi(ic).
TX j  jc 
•  PDCCH : Interfering energy per resource element (dBm/Sym) received over the PDCCH from any cell TXj(jc) at a pixel,
subscriber, or mobile Mi covered by a cell TXi(ic).
515
TX  jc 
j
•  RSSI : Interfering energy per frequency block (dBm/RB) received from any cell TXj(jc) at a pixel, subscriber, or mobile
Mi over 1 frequency block during an OFDM symbol carrying reference signals.
Inter – Tech
• I DL : Downlink inter-technology interference.
6.4.4.4 C/N Calculation (DL)

Input
TX i  ic 
or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 501.
TX i  ic 
• E SS : Received SS energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as
calculated in "Signal Level Calculation (DL)" on page 501.
TX i  ic 
• E PBCH : Received PBCH energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as
TX i  ic 
• E PDCCH : Received PDCCH energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as
TX i  ic 
• E PDSCH : Received PDSCH energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as
TX  ic 
i
• n Sym : Downlink noise for one subcarrier for the cell TXi(ic) as calculated in "Noise Calculation (DL)" on page 505.
TX i  ic 
• CINR DLRS : Downlink reference signal C/(I+N) from cell TXi(ic) at pixel, subscriber, or mobile Mi as calculated in "C/
(I+N) and Bearer Calculation (DL)" on page 518.
Mi
• T SU – MIMO – DL : SU-MIMO threshold defined in the reception equipment used by Mi’s terminal.
Mi
• T B : Bearer selection thresholds of the bearers defined in the reception equipment used by Mi’s terminal.
Mi
• B DL – Highest  Service  : Highest downlink bearer defined in the properties of the service used by the pixel, subscriber,
or mobile Mi.
Mi
• B DL – Lowest  Service  : Lowest downlink bearer defined in the properties of the service used by the pixel, subscriber,
or mobile Mi.
TX i  ic 
Mi
• N Ant – RX : Number of reception (downlink) antenna ports defined for the terminal used by the pixel, subscriber, or
mobile Mi.
• Mobility  M i  : Mobility used for the calculations.
M
BLER  BDL : Downlink block error rate read from the graphs available in the reception equipment assigned to the
i
•
terminal used by the pixel, subscriber, or mobile Mi.
Mi
• G Div – PBCH : PBCH diversity gain defined in the reception equipment of the terminal used by the pixel, subscriber, or
mobile Mi.
Mi
• G Div – PDCCH : PDCCH diversity gain defined in the reception equipment of the terminal used by the pixel, subscriber,
or mobile Mi.
DL
• G Div : Additional downlink diversity gain defined for the clutter class where the pixel, subscriber, or mobile Mi is
located.
Calculations
The C/N for cell TXi(ic) are calculated as follows for any pixel, subscriber, or mobile Mi:
CNR DLRS = E DLRS – n Sym
516
i i i
CNR SS = E SS – n Sym
i i i
CNR PBCH = E PBCH – n Sym
i i i
CNR PDCCH = E PDCCH – n Sym
CNR PDSCH = E PDSCH – n Sym
Bearer Determination:
The bearers available for selection in the pixel, subscriber, or mobile Mi’s reception equipment are the ones:
• Which are common between Mi’s and TXi(ic)’s equipment (bearer indexes for which selection thresholds are
defined in both equipment), if the corresponding option has been set in the Atoll.ini file. For more information,
• Whose indexes are within the range defined by the lowest and the highest bearer indexes defined for the service
being accessed by Mi.
Mi TX i  ic 
• Whose selection thresholds are less than the PDSCH C/N at Mi: T B  CNR PDSCH
If the cell supports MIMO, and according to the Mi diversity mode depending on the SU-MIMO and MU-MIMO
Mi
thresholds and criteria, transmit diversity, SU-MIMO diversity, or MU-MIMO diversity gain, G Div – DL ,
corresponding to the bearer is applied to its selection threshold. The gain is read from the properties of the
TX i  ic  Mi
reception equipment assigned to the pixel, subscriber, or mobile Mi for N Ant – TX , N Ant – RX , Mobility  M i  ,
M
BLER  B DL .
i
 
DL
The additional downlink diversity gain defined for the clutter class of the pixel, subscriber, or mobile Mi G Div is
also applied. Therefore, the bearers available for selection are all the bearers defined in the reception equipment
for which the following is true:
M M TX  ic 
i i DL i
T B – G Div – DL – G Div  CNR PDSCH
The bearer selected for data transfer depends on the bearer selection criterion of the scheduler used by the cell TXi(ic).
• Bearer Index
From among the bearers available for selection, the selected bearer is the one with the highest index.
• Peak RLC Throughput
From among the bearers available for selection, the selected bearer is the one with the highest downlink peak RLC
channel throughput as calculated in "Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and
Per-user Throughput Calculation" on page 547.
• Effective RLC Throughput
From among the bearers available for selection, the selected bearer is the one with the highest downlink effective
RLC channel throughput as calculated in "Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput,
and Per-user Throughput Calculation" on page 547.
MIMO Diversity Gain:
With MIMO, the PBCH and PDCCH C/N become:
DL
CNR PBCH = CNR PBCH + G Div – PBCH + G Div
TX i  ic  TX i  ic  Mi DL
CNR PDCCH = CNR PDCCH + G Div – PDCCH + G Div
The PBCH and PDCCH diversity gains are applied to the C/N when the cell and the terminal both support any form of
MIMO in downlink. The additional downlink diversity gain defined per clutter is also applied.
Once the bearer is known, the PDSCH C/N calculated above becomes:
DL
CNR PDSCH = CNR PDSCH + G Div – DL + G Div
517
M
i
Where G Div – DL is the transmit diversity, SU-MIMO diversity, or MU-MIMO diversity gain applied if the cell supports
MIMO and depending on the Mi diversity mode, the SU-MIMO and MU-MIMO thresholds and criteria.
Output
TX i  ic 
• CNR DLRS : Downlink reference signal C/N from cell TXi(ic) at pixel, subscriber, or mobile Mi.
TX  ic 
i
• CNR SS : SS C/N from cell TXi(ic) at pixel, subscriber, or mobile Mi.
TX i  ic 
• CNR PBCH : PBCH C/N from cell TXi(ic) at pixel, subscriber, or mobile Mi.
TX i  ic 
• CNR PDCCH : PDCCH C/N from cell TXi(ic) at pixel, subscriber, or mobile Mi.
TX i  ic 
• CNR PDSCH : PDSCH C/N from cell TXi(ic) at pixel, subscriber, or mobile Mi.
6.4.4.5 C/(I+N) and Bearer Calculation (DL)

The carrier signal to interference and noise ratio is calculated in three steps. First Atoll calculates the received signal level from
the studied cell (as explained in "Signal Level Calculation (DL)" on page 501) at the pixel, subscriber, or mobile under study.
Next, Atoll calculates the interference received at the same studied pixel, subscriber, or mobile from all the interfering cells
(as explained in "Interference Calculation (DL)" on page 505). Interference from each cell is weighted according to the co- and
adjacent channel overlap between the studied and the interfering cells, the traffic loads of the interfering cells, and the
probability of collision in case ICIC is used by the cells. Finally, Atoll takes the ratio of the signal level and the sum of the total
interference from other cells and the downlink noise (as calculated in "Noise Calculation (DL)" on page 505).
The receiver terminal is always considered to be oriented towards its best server, except when the "Lock Status" is set to
"Server+Orientation" for a subscriber in a subscriber list and its azimuth and tilt manually edited. In the case of NLOS between
the receiver and the best server, Atoll does not try to find the direction of the strongest signal, the receiver is oriented towards
the best server just as in the case of LOS.
Input

TX i  ic 
• D CP : Cyclic prefix duration defined in TXi(ic) frame configuration or, otherwise, in the global network settings.
TX i  ic 
• N FB : Number of frequency blocks, defined in the frequency bands table, for the channel bandwidth used by the
cell TXi(ic).
TX i  ic 
TX i  ic 
bands.
FDD 10 0
DSUUU-DSUUU 2 2
DSUUD-DSUUD 4 2
DSUDD-DSUDD 6 2
DSUUU-DSUUD 3 2
DSUUU-DDDDD 6 1
518
Configuration i i
N SF – DL N TDD – SSF
DSUUD-DDDDD 7 1
DSUDD-DDDDD 8 1
TX i  ic 
or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 501.
TX i  ic 
• E SS : Received SS energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as
TX i  ic 
• E PBCH : Received PBCH energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as
TX i  ic 
• E PDCCH : Received PDCCH energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as
TX i  ic 
• E PDSCH : Received PDSCH energy per resource element from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as
TX  ic 
i
• N Sym – PDCCH : Number of PDCCH resource elements as calculated in "Downlink Transmission Power Calculation" on
page 485.
TX  ic 
i
• N Sym – PDSCH : Number of PDSCH resource elements as calculated in "Downlink Transmission Power Calculation" on
page 485.
TX  ic 
i
• n Sym : Downlink noise for one subcarrier for the cell TXi(ic) as calculated in "Noise Calculation (DL)" on page 505.
TX j  jc 
•  DLRS : Interfering energy per resource element (dBm/Sym) received over downlink reference signals from any cell
TXj(jc) at a pixel, subscriber, or mobile Mi covered by a cell TXi(ic) as calculated in "Interference Calculation (DL)" on
page 505.
TX j  jc 
•  SS PBCH : Interfering energy per resource element (dBm/Sym) received over the SS and the PBCH from any cell TXj(jc)
at a pixel, subscriber, or mobile Mi covered by a cell TXi(ic) as calculated in "Interference Calculation (DL)" on page 505.
TX j  jc 
•  PDSCH : Interfering energy per resource element (dBm/Sym) received over the PDSCH from any cell TXj(jc) at a pixel,
subscriber, or mobile Mi covered by a cell TXi(ic) as calculated in "Interference Calculation (DL)" on page 505.
TX j  jc 
•  PDCCH : Interfering energy per resource element (dBm/Sym) received over the PDCCH from any cell TXj(jc) at a pixel,
subscriber, or mobile Mi covered by a cell TXi(ic) as calculated in "Interference Calculation (DL)" on page 505.
TX j  jc 
•  RSSI : Interfering energy per frequency block (dBm/RB) received over 1 frequency block during an OFDM symbol
carrying reference signals from any cell TXj(jc) at a pixel, subscriber, or mobile Mi as calculated in "Interference
Calculation (DL)" on page 505.
Inter – Tech
• NR DL : Inter-technology downlink noise rise.
TX i  ic 
• CNR DLRS : Downlink reference signal C/N from cell TXi(ic) at pixel, subscriber, or mobile Mi as calculated in "C/N
Mi
• T SU – MIMO – DL : SU-MIMO threshold defined in the reception equipment used by Mi’s terminal.
Mi
• T B : Bearer selection thresholds of the bearers defined in the reception equipment used by Mi’s terminal.
M
i
or mobile Mi.
Mi
or mobile Mi.
TX  ic 
i
519
M
i
• N Ant – RX : Number of reception (downlink) antenna ports defined for the terminal used by the pixel, subscriber, or
mobile Mi.
M
BLER  BDL : Downlink block error rate read from the graphs available in the reception equipment assigned to the
i
•
Mi
• G Div – PBCH : PBCH diversity gain defined in the reception equipment of the terminal used by the pixel, subscriber, or
mobile Mi.
Mi
• G Div – PDCCH : PDCCH diversity gain defined in the reception equipment of the terminal used by the pixel, subscriber,
or mobile Mi.
DL
located.
Inter – Tech
• I DL : Downlink inter-technology interference as calculated in "Interference Calculation (DL)" on page 505.
Calculations
The downlink reference signal C/(I+N) for cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
    j  n Sym 
i 
TX i  ic     -------------------
DLRS -
TX i  ic  --------------------- 
CINR DLRS = E DLRS –  10  Log   10 10  + I Inter – Tech + 10 10  + NR Inter – Tech
   DL  DL 
  All TXj  jc     
     
The SS C/(I+N) for cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
    j  n Sym 
i 
TX  ic  TX  ic     -------------------------
SS PBCH-
--------------------- 
–  10  Log    10  +I  + NR 
i i 10 Inter – Tech 10 Inter – Tech
CINR SS = E SS + 10
  DL  DL 
  All TXj  jc     
     
The PBCH C/(I+N) for cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
    j  n Sym 
i 
TX i  ic     -------------------------
SS PBCH
TX i  ic  - --------------------- 
CINR PBCH = E PBCH –  10  Log   10 10  + I Inter – Tech + 10 10  + NR Inter – Tech

  DL  DL 
  All TXj  jc     
     
The PDCCH C/(I+N) for cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
    j  n
i  
TX i  ic  TX i  ic     --------------------
PDCCH 
- ---------------------
Sym 
   10  10  Inter – Tech

Inter – Tech
CINR PDCCH = E PDCCH – 10  Log 10 +I +10 + NR DL
    DL  
  All TXj  jc     
     
The PDSCH C/(I+N) for cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
    j  n
i  
TX i  ic     --------------------
PDSCH 
--------------------- 
Sym
TX i  ic  -
= E PDSCH –  10  Log   10
  +I  + NR 
CINR PDSCH +10
    DL  DL 
  All TXj  jc     
     
The downlink reference signal received quality (RSRQ) for cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile
Mi:
TX i  ic  TX i  ic  TX  ic  TX i  ic 
RSRQ = 10  Log  N FB  + E i – RSSI
  DLRS
520
i i
Where E DLRS is the cell’s RSRP and RSSI is the received signal strength indicator, i.e., the received signals from the
server (TXi(ic)), and all the interfering cells (TXj(jc)), calculated as follows:
For cells using more than 1 antenna port, the encircled 10 in the formulas below is
replaced with 8.
   j  n Sym
i 
TX  ic   TX  ic   -------------------
RSSI -
---------------------  TX  ic 
  10 
 12 + NR DL + 10  Log  N FB 

i i Inter – Tech 10 Inter – Tech i
RSSI = 10  Log   RSSI + 10 + + 10
  I DL
 All TX j  jc    
   
The downlink reference signal total noise (I+N) for cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
   j  n Sym 
i
TX i  ic    -------------------
DLRS 
- --------------------- TX  ic 
= 10  Log    10  +I  + NR Inter – Tech + 10  Log  2  N i 
 I + N  DLRS + 10
  DL  DL  FB 
 All TXj  jc    
   
The SS and PBCH total noise (I+N) for cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
   j  n Sym 
i
TX i  ic    -------------------------
SS PBCH
- ---------------------
 I + N  SS PBCH = 10  Log    10 10  + I Inter – Tech + 10 10  + NRInter – Tech + 10  Log  N
SCa – FB  N FB – SS PBCH 
  DL  DL
 All TX j  jc    
   
The PDSCH and PDCCH total noise (I+N) for cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
Method 1: Synchronised Transmission and Reception
For details, see "Interference Calculation (DL)" on page 456.
   j  n Sym 
i
TX i  ic 
TX i  ic    --------------------
PDCCH
- ---------------------  N Sym – PDCCH 
  10  10  

Inter – Tech – Tech
 I + N  PDCCH = 10  Log 10 +I + 10 - + NR Inter
+ 10  Log ----------------------------------------------
   DL   TX i  ic  TX i  ic   DL
 All TXj  jc      N SF – DL + N TDD – SSF
   
   j  n
i  TX  ic 
TX i  ic    --------------------
PDSCH- Sym -
--------------------  N Sym – PDSCH 
i
  10  10  

Inter – Tech – Tech
 I + N  PDSCH = 10  Log 10 + I DL + 10 - + NR Inter
+ 10  Log ----------------------------------------------
     TXi  ic  TX  ic  
i
DL
 All TXj  jc      N SF – DL + N TDD – SSF
   
Method 2: Non-synchronised Transmission and Reception

For details, see "Interference Calculation (DL)" on page 456.
   j  n Sym 
i
  --------------------
PDCCH
- ---------------------  N TXi  ic  TX  ic 
i 
Sym – PDSCH + N Sym – PDCCH 
TX i  ic 
  10  10  

Inter – Tech –
 I + N  PDCCH = 10  Log 10 + I DL + 10 - + NR Inter
+ 10  Log -------------------------------------------------------------------
     TX  ic  
i
DL
 All TXj  jc      N SD  Slot  N Slot  SF  N SF – DL
   
   j  n Sym 
i
  --------------------
PDSCH-
---------------------  N TXi  ic  TX i  ic 

Sym – PDSCH + N Sym – PDCCH 
TX i  ic 
  10  10  

Inter – Tech –
 I + N  PDSCH = 10  Log  -------------------------------------------------------------------- + NR Inter
 + I DL  + 10  Log 
+ 10
 10 TX i  ic   DL
 All TXj  jc      N SD  Slot  N Slot  SF  N SF – DL
   
With N SCa – FB calculated as follows:
W FB
N SCa – FB = ---------
-
F
521
The bearers available for selection in the pixel, subscriber, or mobile Mi’s reception equipment are the ones:
Mi TX i  ic 
• Whose selection thresholds are less than the PDSCH C/(I+N) at Mi: T B  CINR PDSCH
Mi
thresholds and criteria, transmit diversity, SU-MIMO diversity, or MU-MIMO diversity gain, G Div – DL ,
TX i  ic  Mi
reception equipment assigned to the pixel, subscriber, or mobile Mi for N Ant – TX , N Ant – RX , Mobility  M i  ,
M
BLER  B DL .
i
 
DL
The additional downlink diversity gain defined for the clutter class of the pixel, subscriber, or mobile Mi G Div is
also applied. Therefore, the bearers available for selection are all the bearers defined in the reception equipment
for which the following is true:
M M TX  ic 
i i DL i
T B – G Div – DL – G Div  CINR PDSCH
• Bearer Index
From among the bearers available for selection, the selected bearer is the one with the highest downlink peak RLC
With MIMO, the PBCH and PDCCH C/(I+N) become:
DL
CINR PBCH = CINR PBCH + G Div – PBCH + G Div
CINR PDCCH = CINR PDCCH + G Div – PDCCH + G Div
The PBCH and PDCCH diversity gains are applied to the C/(I+N) when the cell and the terminal both support any form
of MIMO in downlink. The additional downlink diversity gain defined per clutter is also applied.
Once the bearer is known, the PDSCH C/(I+N) calculated above becomes:
CINR PDSCH = CINR PDSCH + G Div – DL + G Div
Mi
Where G Div – DL is the transmit diversity, SU-MIMO diversity, or MU-MIMO diversity gain applied if the cell supports
MIMO and depending on the Mi diversity mode, the SU-MIMO and MU-MIMO thresholds and criteria.
Output
TX i  ic 
• CINR DLRS : Downlink reference signal C/(I+N) from cell TXi(ic) at pixel, subscriber, or mobile Mi.
TX i  ic 
• CINR SS : SS C/(I+N) from cell TXi(ic) at pixel, subscriber, or mobile Mi.
TX  ic 
i
• CINR PBCH : PBCH C/(I+N) from cell TXi(ic) at pixel, subscriber, or mobile Mi.
522
TX  ic 
i
• CINR PDCCH : PDCCH C/(I+N) from cell TXi(ic) at pixel, subscriber, or mobile Mi.
TX  ic 
i
• CINR PDSCH : PDSCH C/(I+N) from cell TXi(ic) at pixel, subscriber, or mobile Mi.
TX  ic 
i
• RSRQ : Downlink reference signal received quality from cell TXi(ic) at pixel, subscriber, or mobile Mi.
TX i  ic 
• RSSI : Received signal strength indicator, i.e., the received signals from the server (TXi(ic)), and all the interfering
cells (TXj(jc)), at pixel, subscriber, or mobile Mi.
TX  ic 
i
•  I + N  DLRS : Downlink reference signals total noise from the interfering cells TXj(jc) at the pixel, subscriber, or mobile
Mi covered by a cell TXi(ic).
TX i  ic 
•  I + N  SS PBCH : SS and PBCH total noise from the interfering cells TXj(jc) at the pixel, subscriber, or mobile Mi covered
by a cell TXi(ic).
TX i  ic 
•  I + N  PDCCH : PDCCH total noise from the interfering cells TXj(jc) at the pixel, subscriber, or mobile Mi covered by a
cell TXi(ic).
TX i  ic 
•  I + N  PDSCH : PDSCH total noise from the interfering cells TXj(jc) at the pixel, subscriber, or mobile Mi covered by a
cell TXi(ic).
Mi
• B DL : Bearer assigned to the pixel, subscriber, or mobile Mi in the downlink.
6.4.4.6 Signal Level Calculation (UL)

Input
TX i  ic 
• CINR PUSCH – Max : Maximum PUSCH C/(I+N) defined for the cell TXi(ic).
TX  ic 
i
• NR UL : Uplink noise rise of the cell TXi(ic). This value can be user-defined or calculated as explained in "Interference
Calculation (UL)" on page 525.
TX  ic 
i
• NR UL – ICIC : ICIC uplink noise rise of the cell TXi(ic). This value can be user-defined or calculated as explained in
"Interference Calculation (UL)" on page 525.
TX i  ic 
• n PUSCH PUCCH : Uplink noise for the PUSCH and the PUCCH for the cell TXi(ic).
TX i  ic 
cell TXi(ic).
TX  ic 
i
•  FPC : Fractional uplink power control factor defined for the cell TXi(ic).
Mi
• P Max : Maximum transmission power of the terminal used by the pixel, subscriber, or mobile Mi.
Mi
• P Eff : Effective transmission power of the terminal used by the pixel, subscriber, or mobile Mi after power control
adjustment as calculated in "C/(I+N) and Bearer Calculation (UL)" on page 532.
TX i
• G Ant : Transmitter antenna gain for the antenna used by the transmitter TXi.
TX i TX i
• L : Total transmitter losses for the transmitter TXi ( L = L Total – UL ).
TX i
• L Path : Path loss ( L Path = L Model + L Ant ).
• L Total : Total loss calculated as explained in "Signal Level Calculation (DL)" on page 501.
TX
i
is selected.
523
M
i
M
i
M
i
Mi
Mi
Mi
TX i  ic 
Calculations
Atoll first calculates the allowed maximum transmission power for the terminal used by the pixel, subscriber, or mobile Mi.
This power is calculated by performing fractional power control.
Fractional Power Control:
Fractional power control imposes a limitation on the maximum transmission power of the terminal. A nominal PUSCH
power is indicated by the cell to all the pixels, subscribers, or mobiles. This nominal PUSCH power is calculated as
follows:
TX i  ic  TX i  ic  TX i  ic  TX  ic  TX  ic 
+ n PUSCH PUCCH – 10  Log  N FB  for cell centre.
i i
P O_PUSCH = CINR PUSCH – Max + NRUL 
P O_PUSCH = CINR PUSCH – Max + NRUL – ICIC + n PUSCH PUCCH – 10  Log  NFB  for cell-edge.
i i i i i

Where n PUSCH PUCCH – 10  Log  N FB  corresponds to the uplink noise over 1 frequency block.
 
Next, the maximum allowed transmission power for the terminal used by the pixel, subscriber, or mobile Mi is
Mi  Mi TX i  ic  TX i  ic  TX i  ic  
P Allowed = Min  P Max 10  Log  NFB  + P O_PUSCH +  FPC  L Total 
 
 
Once the maximum allowed power has been calculated, it is used as an upper limit for transmission power in all the remaining
calculations.
The received PUSCH and PUCCH signal level (dBm) from a pixel, subscriber, or mobile Mi at its serving cell TXi(ic) is calculated
as follows:
Mi Mi TX i TX i Mi Mi
C PUSCH PUCCH = EIRP PUSCH PUCCH – L Path – M Shadowing – Model – L Indoor + G Ant – L – L Ant – L Body + f CP
Where EIRP is the effective isotropic radiated power of the terminal calculated as follows:
Mi Mi Mi Mi
EIRP PUSCH PUCCH = P +G –L
Mi Mi Mi Mi
With P = P Allowed without power control adjustment at the start of the calculations, and is P = P Eff after power control
adjustment.
f CP is the cyclic prefix factor, i.e., the ratio of the useful symbol energy to the total symbol energy.
The total symbol duration of a modulation symbol comprises the useful symbol duration, carrying the actual data bits, and a
cyclic prefix, added to the useful data bits as padding against multi-path to avoid inter-symbol interference. Hence, the total
energy within a modulation symbol belongs in part to the useful data bits and in part to the cyclic prefix. Once a modulation
524
symbol is received, only the energy of the useful data bits can be used for extracting the data. The energy belonging to the
cyclic prefix is lost once it has served its purpose of combatting inter-symbol interference. Therefore, f CP implies that the
energy belonging to the cyclic prefix is excluded from the useful signal level.
 TX  ic 
 10  Log  7  7.5  If D CPi = Normal

f CP =  TX i  ic 
 10  Log  6  7.5  If D CP = Extended

 0 If M i is an interferer
The cyclic prefix energy and the useful data bits energy are both taken into account when calculating interfering signal levels.
Output
Mi
• C PUSCH PUCCH : Received PUSCH and PUCCH signal level from the pixel, subscriber, or mobile Mi at a cell TXi(ic).
Mi
• P Allowed : Maximum allowed transmission power for the terminal used by the pixel, subscriber, or mobile Mi.
6.4.4.7 Noise Calculation (UL)

For determining the C/N and C/(I+N), Atoll calculates the uplink noise over the channel bandwidth used by the cell. The used
bandwidth depends on the number of used subcarriers.
The uplink noise comprises thermal noise and the noise figure of the equipment. The thermal noise density depends on the
temperature, i.e., it remains constant for a given temperature. However, the value of the thermal noise varies with the used
bandwidth.
Input
• W FB : Width of a frequency block in the frequency domain (180 kHz).
TX i  ic 
cell TXi(ic).
TX i  ic 
• nf : Noise figure of the cell TXi(ic).
Calculations

n 0 = 10  Log  K  T  1000  = – 174 dBm/Hz
The thermal noise for the PUSCH and the PUCCH is calculated as:
n 0 – PUSCH PUCCH = n 0 + 10  Log  N FB  W FB  1000
The uplink noise is the sum of the thermal noise and the noise figure of the cell TXi(ic).
n PUSCH PUCCH = n 0 – PUSCH PUCCH + nf
Output
TX i  ic 
• n PUSCH PUCCH : Uplink noise for the PUSCH and the PUCCH for the cell TXi(ic).
6.4.4.8 Interference Calculation (UL)

The PUSCH and PUCCH interference is only calculated during Monte Carlo simulations. In coverage predictions, the uplink
noise rise values already available in simulation results or in the Cells table are used.
The interference received by a cell TXi(ic) from an interfering mobile covered by a cell TXj(jc) can be defined as the PUSCH and
PUCCH signal level received from the interfering mobile Mj depending on the overlap that exists between the channels used
by the cells TXi(ic) and TXj(jc) and on the traffic load of the interfering mobile Mj.
525
The calculation of uplink interference can be divided into two parts:

• Calculation of the uplink interference from each individual interfering mobile as explained in "Interfering Signal Level
• Calculation of the uplink noise rise which represents the total uplink interference from all interfering mobiles as
explained in "Noise Rise Calculation (UL)" on page 528.
6.4.4.8.1 Interfering Signal Level Calculation (UL)

Input
i j
Mj
• C PUSCH PUCCH : PUSCH and PUCCH signal level received at a cell TXi(ic) from an interfering mobile Mj covered by a cell
TXj(jc) as calculated in "Signal Level Calculation (UL)" on page 523.
In Monte Carlo simulations, interfering signal levels already include M Shadowing – Model , as explained in "Signal Level
information, see "Shadow Fading Model" on page 90). As the interfering signal levels already include
M Shadowing – Model , M Shadowing – C  I is added to the received interfering signal levels in order to achieve the ratio
M Shadowing – Model – M Shadowing – C  I :
M M
j j
C PUSCH PUCCH = C PUSCH PUCCH + M Shadowing – C  I
is selected.
Mj
• TL UL : Uplink traffic load of the interfering mobile Mj.
Traffic loads are calculated during Monte Carlo simulations as explained in "Scheduling and Radio Resource
Allocation" on page 552.
Calculations
The uplink interference received at a cell TXi(ic) from an interfering mobile Mj covered by a cell TXj(jc) is calculated as follows:
Mj Mj TX i  ic  – TX j  jc  Mj TX i  ic  – TX j  jc  TX i  ic  – TX j  jc 
I PUSCH PUCCH = C PUSCH PUCCH + f O + f TL – UL + f ICIC – UL + f ABS – UL
Mj
Where f TL – UL is an interference reduction factor due to the uplink traffic load of the interfering mobile Mj, calculated as
follows:
M M
f TL – UL = 10  Log  TL UL
j j
 
526
TX  ic  – TX  jc 
i j
Calculations for the interference reduction factors due to channel overlapping ( f O ) and static uplink ICIC using
TX  ic  – TX  jc 
i j
fractional frequency reuse ( f ICIC – UL ) are explained below:
as follows:
fO = 10  Log  r O 

Interference reduction due to static uplink ICIC using fractional frequency reuse:
If the cell supports static ICIC in uplink, it means that a part of the LTE frame may use a fraction of the channel
bandwidth. The interference reduction factor due to static uplink ICIC using fractional frequency reuse is calculated as
follows:
f ICIC – UL = 10  Log  p Collision 
 
Where p Collision is the collision probability between the subcarriers used by the interfered and interfering cells.
It is determined during Monte Carlo simulations as follows:
N FB – CC
Cell centre: p Collision = --------------------
-
TX i  ic 
N FB – CC
N FB – CE
Cell-edge: p Collision = --------------------
-
TX  ic 
i
N FB – CE
Common Common
Where, N FB – CC is the number of common frequency blocks in TXi(ic) and TXj(jc) in cell centre, N FB – CE is the
number of common frequency blocks in TXi(ic) and TXj(jc) on cell-edge, N FB – CC and N FB – CE are respectively the
numbers of frequency blocks in cell centre and cell-edge of TXi(ic).

ICIC mode
No FFR N FB N FB
Soft FFR i i i
N FB – N FB – CEx N FB – CEx
–  N FB – CE0 + N FB – CE1 + N FB – CE2
N FB N FB – CEx
i i i i
Whether a pixel, subscriber, or mobile is located in the cell-edge is determined as explained in "Best Server
Interference reduction due to subframe collision probabilities:
The interference reduction factor due to uplink subframe collision probabilities f ABS – UL is calculated as
explained in "Subframe Pattern Collision Calculation" on page 497.
In Monte Carlo simulations, Atoll calculates two separate noise rise values; for the mobiles located in the cell-edge of the
interfered cell Atoll calculates the ICIC UL Noise Rise, and for the mobiles located in the cell centre of the interfered cell Atoll
calculates the UL Noise Rise.
527
In coverage predictions, point analysis, and calculations on subscriber lists, according to the zone, cell centre or cell-edge,
where the pixel, receiver, or subscriber is located, Atoll uses either the ICIC UL Noise Rise or the UL Noise Rise to calculate the
PUSCH and PUCCH C/(I+N). For more information on the calculation of the uplink noise rise, see "Noise Rise Calculation (UL)"
on page 528.
Output
Mj
• I PUSCH PUCCH : PUSCH and PUCCH interference signal level received at a cell TXi(ic) from an interfering mobile Mj
covered by a cell TXj(jc).
6.4.4.8.2 Noise Rise Calculation (UL)

The uplink noise rise is defined as the ratio of the total uplink interference received by any cell TXi(ic) from all interfering
mobiles Mj present in the coverage areas of all other cells TXj(jc) to the uplink noise of the cell TXi(ic). In other words, it is the
ratio (I+N)/N.
Input
Mj
• I PUSCH PUCCH : PUSCH and PUCCH interference signal levels received at a cell TXi(ic) from interfering mobiles Mj
covered by other cells TXj(jc) as calculated in "Interfering Signal Level Calculation (UL)" on page 526.
TX i  ic 
• n PUSCH PUCCH : Uplink noise for the PUSCH and the PUCCH for the cell TXi(ic) as calculated in "Noise Calculation (UL)"
on page 525.
Inter – Tech
• NRUL : Inter-technology uplink noise rise.
Calculations
For any mobile Mi in the cell centre of the interfered cell TXi(ic), Atoll calculates the UL Noise Rise as follows:
M
 j
 I PUSCH PUCCH  TX  ic  
   non-ICIC M i n PUSCH PUCCH
i
TX i  ic    -----------------------------------------------------------------------------
- ------------------------------------------- TX  ic 
= 10  Log    + NR Inter – Tech – n i
10 10
NR UL  10  + 10  UL PUSCH PUCCH
 All M j  
 
 
 All TXj  jc    
For any pixel, subscriber, or mobile Mi in the cell centre of the interfered cell TXi(ic), Atoll calculates the PUSCH and PUCCH
total noise (I+N) as follows:
 I + N  PUSCH PUCCH = NR UL + n PUSCH PUCCH
For any mobile Mi in the cell-edge of the interfered cell TXi(ic), Atoll calculates the ICIC UL Noise Rise as follows:
M
 j
   ICIC M i n PUSCH PUCCH
TX i  ic    ------------------------------------------------------------------
- ------------------------------------------- TX  ic 
= 10  Log    + NR Inter – Tech – n i
10 10
NR UL – ICIC
  10  + 10
 UL PUSCH PUCCH
 All Mj  
 
 
 All TXj  jc    
For any pixel, subscriber, or mobile Mi in cell-edge of the interfered cell TXi(ic), Atoll calculates the PUSCH and PUCCH total
noise (I+N) as follows:
i i i
 I + N  PUSCH PUCCH = NR UL – ICIC + n PUSCH PUCCH
Output
TX i  ic 
• NRUL : Uplink noise rise for the cell TXi(ic).
TX i  ic 
• NRUL – ICIC : ICIC uplink noise rise for the cell TXi(ic).
TX i  ic 
•  I + N  PUSCH PUCCH : PUSCH and PUCCH total noise for a cell TXi(ic) calculated for any pixel, subscriber, or mobile Mi.
528
6.4.4.9 C/N Calculation (UL)

Input
M
i
• C PUSCH PUCCH : Received PUSCH and PUCCH signal level from the pixel, subscriber, or mobile Mi at its serving cell TXi(ic)
as calculated in "Signal Level Calculation (UL)" on page 523.
TX i  ic 
• n PUSCH PUCCH : PUSCH and PUCCH noise for the cell TXi(ic) as calculated in "Noise Calculation (UL)" on page 525.
TX i  ic 
• T SU – MIMO – UL : SU-MIMO threshold defined in the reception equipment of the cell TXi(ic).
TX  ic 
i
• N FB : Total number of frequency blocks defined in the frequency bands table for the channel bandwidth used by
the cell TXi(ic).
TX i  ic 
• N FB – CE0 : Number of cell-edge frequency blocks defined in the frame configurations table for the channel bandwidth
used by the cell TXi(ic) and PSS ID 0.
TX  ic 
i
TX i  ic 
TX i  ic 
• T B – Lowest : Bearer selection threshold of the lowest bearer in the reception equipment assigned to the cell TXi(ic).
Mi
• P Allowed : Maximum allowed transmission power of the terminal used by the pixel, subscriber, or mobile Mi as
calculated in "Signal Level Calculation (UL)" on page 523.
Mi
• P Min : Minimum transmission power of the terminal used by the pixel, subscriber, or mobile Mi.
• M PC : Power control adjustment margin defined in the global network settings.
Mi
• T B : Bearer selection thresholds of the bearers defined in the reception equipment used bythe cell TXi(ic).
Mi
• B UL – Highest  Service  : Highest downlink bearer defined in the properties of the service used by the pixel, subscriber,
or mobile Mi.
Mi
• B UL – Lowest  Service  : Lowest downlink bearer defined in the properties of the service used by the pixel, subscriber,
or mobile Mi.
M
i
• N Ant – TX : Number of transmission (uplink) antenna ports defined for the terminal used by the pixel, subscriber, or
mobile Mi.
TX i  ic 
• N Ant – RX : Number of reception (uplink) antenna ports defined for the cell TXi(ic).
M
BLER  B UL : Uplink block error rate read from the graphs available in the reception equipment assigned to the cell
i
•
 
TXi(ic).
Calculations
The PUSCH and PUCCH C/N from a pixel, subscriber, or mobile Mi at its serving cell TXi(ic) is calculated as follows:
CNR PUSCH PUCCH = C PUSCH PUCCH – n PUSCH PUCCH
The bearers available for selection in the cell TXi(ic)’s reception equipment are the ones:
529
M M
i i
• Whose selection thresholds are less than the PUSCH and PUCCH C/N at Mi: T B  CNR PUSCH PUCCH
TX  ic 
i
thresholds and criteria, receive diversity, SU-MIMO diversity, or MU-MIMO diversity gain, G Div – UL ,
M TX  ic  M
reception equipment assigned to the cell TXi(ic) for N Ant – TX , N Ant – RX , Mobility  M i  , BLER  B UL .
i i i
UL
The additional uplink diversity gain defined for the clutter class of the pixel, subscriber, or mobile Mi G Div is also
applied. Therefore, the bearers available for selection are all the bearers defined in the reception equipment for
which the following is true:
Mi TX i  ic  Mi
UL
T B – G Div – UL – G Div  CNR PUSCH PUCCH
• Bearer Index
From among the bearers available for selection, the selected bearer is the one with the highest uplink peak RLC
From among the bearers available for selection, the selected bearer is the one with the highest uplink effective
Once the bearer is known, the PUSCH and PUCCH C/N calculated above become:
UL
CNR PUSCH PUCCH = CNR PUSCH PUCCH + G Div – UL + G Div
TX i  ic 
Where G Div – UL is the receive diversity, SU-MIMO diversity, or MU-MIMO diversity gain applied if the cell supports
MIMO and depending on the Mi diversity mode, the SU-MIMO and MU-MIMO threshold and criteria.
Uplink Bandwidth Allocation (No. of Used Frequency Blocks):

The uplink bandwidth allocation depends on the target defined for the scheduler used by the cell TXi(ic). The PUSCH
and PUCCH C/N calculated above is given for the total number of frequency blocks associated with the channel
TX i  ic 
bandwidth of the cell, i.e., N FB . Bandwidth allocation is performed for all the pixels, subscribers, or mobiles in the
uplink, and may reduce the number of used frequency blocks in order to satisfy the selected target.
• Full Bandwidth
Full channel width is used by each mobile in the uplink. As there is no reduction in the bandwidth used for
transmission, there is no gain in the PUSCH and PUCCH C/N.
• Maintain Connection
The bandwidth used for transmission by a mobile is reduced only if the PUSCH and PUCCH C/N is not enough to
even access the lowest bearer. For example, as a mobile moves from good to bad radio conditions, the number of
frequency blocks used by it for transmission in uplink are reduced one by one in order to improve the PUSCH and
PUCCH C/N. The calculation of the gain introduced by the bandwidth reduction is explained below.
• Best Bearer
The bandwidth used for transmission by a mobile is reduced in order to improve the PUSCH and PUCCH C/N
enough to access the best bearer. For example, if using 5 frequency blocks, a mobile is able to access the best
bearer, and using 6 it would only get access to the second best, it will be assigned 5 frequency blocks as the used
uplink bandwidth. Although using 4 frequency blocks, its PUSCH and PUCCH C/N will be better than when using 5,
the uplink bandwidth is not reduced to 4 because it does not provide any gain in terms of the bearer, i.e., the
mobile already has the best bearer using 5 frequency blocks. The calculation of the gain introduced by the
bandwidth reduction is explained below.
530
The definition of the best bearer depends on the bearer selection criterion of the scheduler used by the cell TXi(ic),
i.e., bearer with the highest index, with the highest peak RLC throughput, or with the highest effective RLC
throughput.
The uplink bandwidth allocation may result in the use of a number of frequency blocks which is less than the number
of frequency blocks associated with the channel bandwidth of the cell. The gain related to this bandwidth reduction
is applied to the PUSCH and PUCCH C/N:
Mi Mi  N TX i  ic  
= CNR PUSCH PUCCH + 10  Log  -----------------
-
FB
CNR PUSCH PUCCH
 Mi 
Final All FB
 N FB – UL
Mi TX i  ic 
Min
Where N FB – UL  Service   N FB – UL  N FB – CC for any pixel, subscriber, or mobile Mi in the cell centre of the interfered
Mi TX i  ic 
Min
cell TXi(ic), and N FB – UL  Service   N FB – UL  N FB – CE for any pixel, subscriber, or mobile Mi in the cell-edge of the
interfered cell TXi(ic).

ICIC mode
No FFR i i
N FB N FB
N FB N FB – CEx
Uplink Power Control Adjustment:

Once the bandwidth allocation is performed, Atoll continues to work with the C/N given by the bandwidth allocation,
Mi Mi
i.e., CNR PUSCH PUCCH = CNR PUSCH PUCCH .
Final
The pixel, subscriber, or mobile Mi reduces its transmission power so that the PUSCH and PUCCH C/N from it at its cell
is just enough to get the selected bearer.
Mi Mi Mi TX i  ic  TX i  ic 
If with P = P Allowed AND CNR PUSCH PUCCH  T Mi + M PC , where T Mi is the bearer selection threshold, from
B UL B UL
the reception equipment assigned to the cell TXi(ic), for the bearer selected for the pixel, subscriber, or mobile Mi.
The transmission power of Mi is reduced to determine the effective transmission power from the pixel, subscriber, or
mobile Mi as follows:
Mi  Mi  Mi  TXi  ic    Mi 
P Eff = Max  P Allowed –  CNR PUSCH PUCCH –  T M + M PC   P Min
   B i  
UL
Mi Mi
CNR PUSCH PUCCH is calculated again using P Eff .
Output
Mi
• CNR PUSCH PUCCH : PUSCH and PUCCH C/N from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic).
531
6.4.4.10 C/(I+N) and Bearer Calculation (UL)

The carrier signal to interference and noise ratio is calculated in three steps. First, Atoll calculates the received signal level
from each pixel, subscriber, or mobile at its serving cell using the effective power of the terminal used by the pixel, subscriber,
or mobile as explained in "Signal Level Calculation (UL)" on page 523. Next, Atoll calculates the uplink carrier to noise ratio as
explained in "C/N Calculation (UL)" on page 529. Finally, determines the uplink C/(I+N) by dividing the previously calculated
uplink C/N by the uplink noise rise value of the cell as calculated in "Noise Rise Calculation (UL)" on page 528.
The uplink noise rise can be set by the user manually for each cell or calculated using Monte Carlo simulations.
Input
Mi
• CNR PUSCH PUCCH : PUSCH and PUCCH C/N from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic) as calculated
in "C/N Calculation (UL)" on page 529.
TX i  ic 
• NRUL : Uplink noise rise for the cell TXi(ic) as calculated in "Noise Rise Calculation (UL)" on page 528.
TX i  ic 
• NRUL – ICIC : ICIC uplink noise rise for the cell TXi(ic) as calculated in "Noise Rise Calculation (UL)" on page 528.
TX i  ic 
• T SU – MIMO – UL : SU-MIMO threshold defined in the reception equipment of the cell TXi(c).
TX i  ic 
the cell TXi(ic).
TX i  ic 
TX  ic 
i
TX i  ic 
TX i  ic 
• T B – Lowest : Bearer selection threshold of the lowest bearer in the reception equipment assigned to the cell TXi(ic).
Mi
• P Allowed : Maximum allowed transmission power of the terminal used by the pixel, subscriber, or mobile Mi as
calculated in "Signal Level Calculation (UL)" on page 523.
Mi
• M PC : Power control adjustment margin defined in the global network settings.
Mi
• T B : Bearer selection thresholds of the bearers defined in the reception equipment used bythe cell TXi(ic).
Mi
or mobile Mi.
Mi
or mobile Mi.
Mi
• N Ant – TX : Number of transmission (uplink) antenna ports defined for the terminal used by the pixel, subscriber, or
mobile Mi.
TX i  ic 
• N Ant – RX : Number of reception (uplink) antenna ports defined for the cell TXi(ic).
M
BLER  BUL : Uplink block error rate read from the graphs available in the reception equipment assigned to the cell
i
•
 
TXi(ic).
532
Calculations
For any pixel, subscriber, or mobile Mi in the cell centre of the interfered cell TXi(ic), Atoll calculates the PUSCH and PUCCH C/
(I+N) as follows:
M M TX  ic 
i i i
CINR PUSCH PUCCH = CNR PUSCH PUCCH – NR UL
For any pixel, subscriber, or mobile Mi in the cell-edge of the interfered cell TXi(ic), Atoll calculates the PUSCH and PUCCH C/
(I+N) as follows:
CINR PUSCH PUCCH = CNR PUSCH PUCCH – NR UL – ICIC
The bearers available for selection in the cell TXi(ic)’s reception equipment are the ones:
Mi Mi
• Whose selection thresholds are less than the PUSCH and PUCCH C/(I+N) at Mi: T B  CINR PUSCH PUCCH
TX i  ic 
thresholds and criteria, receive diversity, SU-MIMO diversity, or MU-MIMO diversity gain, G Div – UL ,
M TX  ic  M
reception equipment assigned to the cell TXi(ic) for N Ant – TX , N Ant – RX , Mobility  M i  , BLER  BUL .
i i i
 
UL
The additional uplink diversity gain defined for the clutter class of the pixel, subscriber, or mobile Mi G Div is also
applied. Therefore, the bearers available for selection are all the bearers defined in the reception equipment for
M TX  ic  M
i i UL i
T B – G Div – UL – G Div  CINR PUSCH PUCCH
• Bearer Index
From among the bearers available for selection, the selected bearer is the one with the highest uplink peak RLC
Once the bearer is known, the PUSCH and PUCCH C/(I+N) calculated above become:
UL
CINR PUSCH PUCCH = CINR PUSCH PUCCH + G Div – UL + G Div
TX i  ic 
Where G Div – UL is the receive diversity, SU-MIMO diversity, or MU-MIMO diversity gain applied if the cell supports
MIMO and depending on the Mi diversity mode, the SU-MIMO and MU-MIMO threshold and criteria.
Uplink Bandwidth Allocation (No. of Used Frequency Blocks):

The uplink bandwidth allocation depends on the target defined for the scheduler used by the cell TXi(ic). The PUSCH
and PUCCH C/(I+N) calculated above is given for the total number of frequency blocks associated with the channel
533
TX  ic 
i
bandwidth of the cell, i.e., N FB . Bandwidth allocation is performed for all the pixels, subscribers, or mobiles in the
uplink, and may reduce the number of used frequency blocks in order to satisfy the selected target.
• Full Bandwidth
transmission, there is no gain in the PUSCH and PUCCH C/(I+N).
The bandwidth used for transmission by a mobile is reduced only if the PUSCH and PUCCH C/(I+N) is not enough
to even access the lowest bearer. For example, as a mobile moves from good to bad radio conditions, the number
of frequency blocks used by it for transmission in uplink are reduced one by one in order to improve the PUSCH
and PUCCH C/(I+N). The calculation of the gain introduced by the bandwidth reduction is explained below.
• Best Bearer
The bandwidth used for transmission by a mobile is reduced in order to improve the PUSCH and PUCCH C/(I+N)
enough to access the best bearer. For example, if using 5 frequency blocks, a mobile is able to access the best
bearer, and using 6 it would only get access to the second best, it will be assigned 5 frequency blocks as the used
uplink bandwidth. Although using 4 frequency blocks, its PUSCH and PUCCH C/(I+N) will be better than when using
5, the uplink bandwidth is not reduced to 4 because it does not provide any gain in terms of the bearer, i.e., the
mobile already has the best bearer using 5 frequency blocks. The calculation of the gain introduced by the
bandwidth reduction is explained below.
i.e., bearer with the highest index, with the highest peak RLC throughput, or with the highest effective RLC
throughput.
The uplink bandwidth allocation may result in the use of a number of frequency blocks which is less than the number
of frequency blocks associated with the channel bandwidth of the cell. The gain related to this bandwidth reduction
is applied to the PUSCH and PUCCH C/(I+N):
Mi Mi  N TX i  ic  
CINR PUSCH PUCCH = CINR PUSCH PUCCH + 10  Log  -----------------
-
FB
Mi 
Final All FB
 N FB – UL
Mi TX i  ic 
Min
Where N FB – UL  Service   N FB – UL  N FB – CC for any pixel, subscriber, or mobile Mi in the cell centre of the interfered
Mi TX i  ic 
Min
cell TXi(ic), and N FB – UL  Service   N FB – UL  N FB – CE for any pixel, subscriber, or mobile Mi in the cell-edge of the
interfered cell TXi(ic).

ICIC mode
No FFR N FB N FB
Hard FFR i i
N FB – CEx N FB – CEx
N FB N FB – CEx
 
Uplink Power Control Adjustment:

Once the bandwidth allocation is performed, Atoll continues to work with the C/(I+N) given by the bandwidth
Mi Mi
allocation, i.e., CINR PUSCH PUCCH = CINR PUSCH PUCCH .
Final
The pixel, subscriber, or mobile Mi reduces its transmission power so that the PUSCH and PUCCH C/(I+N) from it at its
cell is just enough to get the selected bearer.
534
M
i M
i
M
i TX  ic  TX  ic 
i i
If with P = P Allowed AND CINR PUSCH PUCCH  T M + M PC , where T M is the bearer selection threshold, from
i i
B B
UL UL
the reception equipment assigned to the cell TXi(ic), for the bearer selected for the pixel, subscriber, or mobile Mi.
Mi  Mi  Mi  TX i  ic    Mi 
P Eff = Max  P Allowed –  CINR PUSCH PUCCH –  T M + M PC   P Min
   B i  
UL
M M
i i
CINR PUSCH PUCCH is calculated again using P Eff .
Output
Mi
• CINR PUSCH PUCCH : PUSCH and PUCCH C/(I+N) from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic).
Mi
• N FB – UL : Number of frequency blocks used by the pixel, subscriber, or mobile Mi after uplink bandwidth allocation.
Mi
• P Eff : Effective transmission power of the terminal used by the pixel, subscriber, or mobile Mi.
Mi
• B UL : Bearer assigned to the pixel, subscriber, or mobile Mi in the uplink.
6.4.5 Best Server Determination

In LTE, best server refers to a cell ("serving transmitter"-"reference cell" pair) that best covers a pixel, subscriber, or mobile
Mi and provides the best service.
Input
TX i  ic 
• C DLRS : Downlink reference signal level received from any cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated
Mi Mi Mi Mi
in "Signal Level Calculation (DL)" on page 501 using the terminal and service parameters ( L ,G , L Ant , and L Body )
of Mi.
TX  ic 
i
• E DLRS : Received downlink reference signal energy per resource element (RSRP) from any cell TXi(ic) at a pixel,
subscriber, or mobile Mi as calculated in "Signal Level Calculation (DL)" on page 501.
TX i  ic 
• T RSRP : Minimum RSRP defined for the cell TXi(ic).
TX  ic 
i
• T Selection : Cell selection threshold defined for the cell TXi(ic).
TX i  ic 
• O Individual : Cell individual offset defined for the cell TXi(ic).
TX i  ic 
• M HO : Handover margin defined for the cell TXi(ic).
TX i  ic 
• p Layer : Priority defined for the layer assigned to for any cell TXi(ic).
Max – DL
• N SCell : Maximum number of downlink secondary cells defined for the terminal used by the pixel, subscriber, or
mobile Mi.
Max – UL
• N SCell : Maximum number of uplink secondary cells defined for the terminal used by the pixel, subscriber, or
mobile Mi.
Calculations
The serving cell selected for coverage predictions is based on the Standard serving cell selection method. The serving cell
selected for Monte Carlo simulations can also be based on the Random method instead of the Standard method. If no serving
cell is found for a mobile Mi, it is rejected for “No Coverage”.
535
The best server selection for any pixel, subscriber, or mobile Mi BSM is performed as follows:
i
• Standard cell selection based on 3GPP specifications for connected mode mobility:
a. Qualification:
To qualify as potential servers, cells must fulfill the following requirements:
• The cells’ layers must be supported by the pixel, subscriber, or mobile Mi’s service and terminal.
• The speed defined in the pixel, subscriber, or mobile Mi’s mobility type must be less than or equal to the
maximum speed supported by the cells’ layers.
• The pixel, subscriber, or mobile Mi must be located within the round-trip time distances corresponding to the
cells’ PRACH preamble format.
Cyclic
PRACH prefix + Window Guard
Cyclic prefix Preamble sequence RTT distance
preamble preamble size period
format sequence
Tsa Sec. Ts Sec. Sec. Sec. Sec. Metres
0 3168 0.00010 24576 0.00080 0.00090 0.00100 0.00010 14521
1 21024 0.00068 24576 0.00080 0.00148 0.00200 0.00052 77290
2 6240 0.00020 49152 0.00160 0.00180 0.00200 0.00020 29511
3 21024 0.00068 49152 0.00160 0.00228 0.00300 0.00072 107269
4 448 0.00001 4096 0.00013 0.00015 0.00017 0.00002 2811
a. The basic unit of time in LTE: Ts = 1/(15000 x 2048) seconds.
• For potential serving cells that belong to layers of higher priorities, the RSRP received at the pixel, subscriber,
or mobile Mi must be higher than or equal to the cells’ Min RSRP plus the cell selection threshold:
E DLRS  T RSRP + Max  0 T Selection .
i i i
• For the potential serving cells that belong to the layer of the lowest priority, the RSRP received at the pixel,
subscriber, or mobile Mi must be higher than or equal to the cells’ Min RSRP: E DLRS  T RSRP .
b. Preselection:
From the list of cells that qualify as potential servers in step a., the cell that fulfills the following conditions is
preselected as the serving cell ( S 0 ):
• The cell belonging to the highest priority layer, and

TX i  ic 
• From which the pixel, subscriber, or mobile Mi receives the highest reference signal level or RSRP ( C DLRS or
TX i  ic 
E DLRS ) according to the defined best server selection criterion.
c. Final selection:
Among the cells that qualify as potential servers, other than the preselected server ( S 0 ), and fulfill the following
condition:
TX i  ic  TX i  ic  S0 S0 S0
E DLRS + O Individual  E DLRS + O Individual + M HO
Atoll selects as the best server the cell from which the pixel, subscriber, or mobile Mi receives the highest
TX i  ic  S0 TX i  ic  S0
reference signal level or RSRP plus the cell individual offset ( C DLRS + O Individual or E DLRS + O Individual ),
according to the defined best server selection criterion.
If no cell fulfills the above condition, then the preselcted server ( S 0 ) is selected as the best server.
• Random cell selection:

a. Qualification:
To qualify as potential servers, cells must fulfill the following requirements:
• The cells’ layers must be supported by the pixel, subscriber, or mobile Mi’s service and terminal.
536
• The cells’ frequency band must be supported by the pixel, subscriber, or mobile Mi’s terminal.
• The speed defined in the pixel, subscriber, or mobile Mi’s mobility type must be less than or equal to the
maximum speed supported by the cells’ layers.
• The pixel, subscriber, or mobile Mi must be located within the round-trip time distances corresponding to the
cells’ PRACH preamble format (see table above).
• The RSRP received at the pixel, subscriber, or mobile Mi must be higher than or equal to the cells’ Min RSRP:
i i
E DLRS  T RSRP .
b. Final selection:
From the list of cells that qualify as potential servers in step a., Atoll keeps only one potential server per layer, i.e.,
per layer the cell from which the pixel, subscriber, or mobile Mi receives the highest reference signal level or RSRP,
and then selects among these cells one cell as the best server at random.
For carrier aggregation, Atoll selects multiple servers by processing lists of potential servers according to the Standard or
Random cell selection method:
LTE users:
a. A list of potential serving cells whose cell type includes “LTE”
LTE-A users:
b. A list of potential primary serving cells whose cell type includes “LTE” and “LTE-A PCell”
c. A list of potential secondary serving cells whose may include “LTE-A SCell DL” and “LTE-A SCell UL”
Atoll selects the serving cell for LTE users from the list a. and a primary serving cell for LTE-A users from the remaining
list b.
Once a primary serving cell has been selected, Atoll eliminates the selected cell as well as any other co-channel cell
from list c. Here, co-channel cells are cells whose channels overlap the channel being used the primary serving cell.
In intra-eNode-B carrier aggregation, at this stage Atoll also eliminates cells belonging to
other eNode-Bs than that of the selected primary cell.
In group-based carrier aggregation, at this stage Atoll also eliminates cells not belonging
to the carrier aggregation groups to which the selected primary cell belongs.
You can switch between carrier aggregation modes, using the Atoll.ini file. For more
information, see the Administrator Manual.
For LTE-A users with a primary serving cell of type “LTE-A PCell” selected from list b., Atoll selects secondary serving
cells from list c.
This step is carried out until either list c. is empty, or the numbers of downlink or uplink secondary serving cells
Max – DL
assigned to the user become equal to the maximum numbers defined in the terminal properties ( N SCell and
Max – UL
N SCell ). Secondary cells are selected based on the reference signal level or RSRP, according to the defined best
server selection criterion. Only secondary cells whose PDSCH C/(I+N) is higher than or equal to the secondary cell
DL
activation threshold defined in the terminal reception equipment properties ( T SCell ) are activated for aggregation in
downlink. Similarly, only secondary cells whose PDSCH C/(I+N) and PUSCH C/(I+N) are both higher than or equal to the
DL
secondary cell activation threshold defined in the terminal and cell reception equipment properties ( T SCell and
UL
T SCell ), respectively, are activated for aggregation in uplink.
The primary and secondary serving cells once assigned to a mobile do not change during a Monte Carlo simulation.
Atoll determines whether the pixel, subscriber, or mobile Mi is in the cell-edge or cell centre of TXi(ic) by calculating the
difference between the path loss from the second best server and the best server, and comparing it with the delta path loss
threshold defined for the best server of the pixel, subscriber, or mobile Mi.
Therefore, a pixel, subscriber, or mobile Mi is considered to be a cell edge if
2ndBS M BS M – 2ndBS M BS M BS M BS M
Log  r O
i i i i i i
L Total + 10  – L Total  L Path , and it is considered to be in cell centre otherwise. Here, L Total is the
 
2ndBS
Mi
total loss from Mi’s best server and L Total is the total loss from Mi’s second best server calculated as explained in "Signal
Level Calculation (DL)" on page 453. The second best server for a pixel, subscriber, or mobile Mi is calculated as follows:
537
2ndBS M = TX i  ic 
i TX  ic   TX i  ic  
i
C = 2ndBest  C 
DLRS All TX  ic   DLRS 
i
BS – 2ndBS
M M
i i
rO is the total channel overlap ratio between the best server and the second best server as calculated in "Co- and
BS
Mi
Adjacent Channel Overlaps Calculation" on page 493. L Path is the delta path loss threshold defined for the best server of
the pixel, subscriber, or mobile Mi.
Output
• BS M : Best serving cell of the pixel, subscriber, or mobile Mi.

i
6.4.6 Throughput Calculation

Throughputs are calculated in two steps.
• Calculation of uplink and downlink total resources in a cell as explained in "Calculation of Total Cell Resources" on
page 538.
• Calculation of uplink and downlink UE capacities as explained in "Calculation UE Capacities" on page 545.
• Calculation of throughputs as explained in "Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and
6.4.6.1 Calculation of Total Cell Resources

The total amount of resources in a cell is the number of modulation symbols that can be used for data transfer in each frame.
The total cell resources can be calculated separately for the downlink and uplink as described in:
• "Calculation of Downlink Cell Resources" on page 538.
• "Calculation of Uplink Cell Resources" on page 543.
6.4.6.1.1 Calculation of Downlink Cell Resources

Input

TX  ic 
i
TX i  ic 
• N SD – PDCCH : Number of PDCCH symbol durations per subframe defined in TXi(ic) frame configuration or, otherwise,
in the global network settings.
TX i  ic 
the cell TXi(ic).
TX i  ic 
TX  ic 
i
TX i  ic 
TX i  ic 
538
TX  ic 
i
bands.
i i
FDD 10 0
DSUUU-DSUUU 2 2
DSUUD-DSUUD 4 2
DSUDD-DSUDD 6 2
DSUUU-DSUUD 3 2
DSUUU-DDDDD 6 1
DSUUD-DDDDD 7 1
DSUDD-DDDDD 8 1
TX i  ic 
Calculations
calculations below.
The number of modulation symbols (resource elements) per scheduler resource block is calculated as follows:
W FB
N SCa – FB = ---------
-
F
The number of modulation symbols (resource elements) corresponding to the DwPTS per scheduler resource block in the TDD
special subframes is calculated as follows:
DwPTS DwPTS
N Sym  SSF = N SCa – FB  N SD  SSF
DwPTS
Where N SD  SSF is the number of DwPTS symbol durations (OFDM symbols) per special subframe, determined from the TDD
special subframe configuration according to the 3GPP specifications as follows:
Cyclic Prefix = Normal Cyclic Prefix = Extended
Special
Subframe DwPTS GP UpPTS DwPTS GP UpPTS
Configuration DwPTS
N SD  SSF
GP
N SD  SSF
UpPTS
N SD  SSF
DwPTS
N SD  SSF
GP
N SD  SSF
UpPTS
N SD  SSF
0 3 10 3 8
1 9 4 8 3
1
2 10 3 1 9 2
3 11 2 10 1
4 12 1 3 7
5 3 9 8 2 2
6 9 3 9 1
2
7 10 2
8 11 1
The total number of modulation symbols (resource elements) in downlink is calculated as follows:
N Sym – DL = N FB  N Sym  SRB  N SF – DL + N Sym – DwPTS
539
i i i DwPTS
Where N Sym – DwPTS = N FB  N TDD – SSF  N Sym  SSF
TX  ic 
i
The total downlink cell resources, i.e., R DL , are calculated as follows:
i i i i i i
R DL = N Sym – DL – O DLRS – O PSS – O SSS – O PBCH – O PDCCH – O DMRS
TX i  ic 
The downlink DwPTS resources, i.e., R DwPTS , are calculated as follows:
R DwPTS = N Sym – DwPTS – O DLRS  DwPTS – O PDCCH  DwPTS
TX i  ic 
Where O DLRS is the overhead corresponding to the downlink reference signals, O PSS is the overhead corresponding to the
TX i  ic 
primary synchronisation signals, O SSS is the overhead corresponding to the secondary synchronisation signals, O PBCH is the
TX i  ic 
overhead corresponding to the physical broadcast channel, and O PDCCH is the overhead corresponding to the physical
TX i  ic 
downlink control channel. O DMRS is the overhead corresponding to the UE-specific reference signals transmitted on the
logical antenna port 5 or the demodulation reference signals transmitted using antenna ports 7 and 8 or 7 through 14.
These overheads are calculated as follows:
Downlink reference signal overhead
The downlink reference signal overhead depends on the number of transmission antenna ports:
i i i i i
O DLRS = N FB  N DLRS  SRB  N SF – DL + O DLRS  DwPTS
Where O DLRS  DwPTS = N FB  N DLRS  DwPTS  N TDD – SSF ,

 TX  ic 
i
 8  

TX  ic 
i  TX  ic 
N DLRS  SRB =  16 if  N Ant – TX = 2 ,
i
  

 TX  ic 
if  N Ant – TX = 4 or 8
i
 24

TX  ic 
i
And N DLRS  DwPTS is determined from the table below:

Subframe DwPTS TX i  ic  TX i  ic  DwPTS TX i  ic  TX i  ic 
1 2 1 2
2 4 2 4
0 3 3
4 8 4 8
8 8 8 8
1 6 1 6
2 12 2 12
1 9 8
4 20 4 20
8 20 8 20
1 6 1 6
2 12 2 12
2 10 9
4 20 4 20
8 20 8 20
540

i i i i
1 6 1 8
2 12 2 16
3 11 10
4 20 4 24
8 20 8 24
1 8 1 2
2 16 2 4
4 12 3
4 24 4 8
8 24 8 8
1 2 1 6
2 4 2 12
5 3 8
4 8 4 20
8 8 8 20
1 6 1 6
2 12 2 12
6 9 9
4 20 4 20
8 20 8 20
1 6
2 12
7 10
4 20
8 20
1 6
2 12
8 11
4 20
8 20
PSS and SSS overhead

The primary and secondary synchonisation signals are transmitted on 1 symbol duration each in the 1st and the 6th
downlink subframes, over the centre 6 frequency blocks. Therefore,
O PSS = 2  N FB – SS PBCH  N SCa – FB = 144 symbols
O SSS = 2  N FB – SS PBCH  N SCa – FB = 144 symbols
PBCH overhead
The physical broadcast channel is transmitted on four symbol durations in the 1st downlink subframe over the center
6 frequency blocks. The physical broadcast channel overlaps with the downlink reference signals, therefore, some
downlink reference signal modulation symbols are subtracted:
216 for extended cyclic prefix
240 for normal cyclic prefix
PDCCH overhead
The physical downlink control channel can be transmitted over up to 4 symbol durations in each subframe. The
number of symbol durations for the PDCCH is defined in the global network settings. The PDCCH overlaps some
downlink reference signal symbols. These downlink reference signal symbols are subtracted from the PDCCH
overhead:
TX  ic 
if  N SD – PDCCH = 0 :
i
TX i  ic 
O PDCCH = 0
541
i i
O PDCCH =  N SD – PDCCH  N SCa – FB – 4  N FB
i i i i i
i N i  i i
 SD – PDCCH  N SCa – FB – 4  N FB
Where O PDCCH  DwPTS =  N TDD – SSF
Otherwise:
TX  ic  TX  ic  TX  ic  TX  ic  TX i  ic  TX i  ic 
O PDCCH =  N SD – PDCCH  N SCa – FB – 2  Min  4 N Ant – TX   N FB
i i i i
  
TX  ic  TX  ic  TX  ic  TX  ic  TX i  ic 
Where O PDCCH  DwPTS =  Min  2 N SD – PDCCH  N SCa – FB – 2  Min  4 N Ant – TX   N FB
i i i i
 N TDD – SSF
    
UE-specific and demodulation reference signal overhead

UE-specific reference signals are transmitted on the logical antenna port 5, DMRS are transmitted on antenna ports 7
and 8, or on 7 through 14.
TX i  ic 
Without smart antennas and MIMO: O DMRS = 0
With smart antennas and without MIMO: O DMRS = 12  N FB  N SF – DL
With smart antennas and with MIMO: O DMRS = 24  N FB  N SF – DL
Without smart antennas and with SU-MIMO or MU-MIMO and N Ant – TX  4 : O DMRS = 24  N FB  N SF – DL
TX  ic 
i
Once R DL is known, it is scaled down according to the ICIC mode used by the cell TXi(ic) depending on whether the
downlink cell resources are being calculated for a cell-centre or cell-edge pixel, subscriber, or mobile.
ICIC ABS
R DL = R DL  f Scaling  f Scaling
ICIC
f Scaling is calculated as follows for the different ICIC modes:
ICIC
ICIC cell resource scaling factor f Scaling for
ICIC mode
No FFR 1 1
TX i  ic 
N FB – CEx
Time-switched FFR 1 --------------------
TX i  ic 
N FB
Hard FFR -------------------- --------------------
N FB N FB
i i i
Soft FFR ------------------------------------------ --------------------
i i
N FB N FB
N FB –  N FB – CE0 + N FB – CE1 + N FB – CE2 N FB – CEx
Partial soft FFR ------------------------------------------------------------------------------------------------------ --------------------
N FB N FB
i i i i
542
ABS
f Scaling is calculated as follows:
• Method 1: ABS Patterns Used Only at Cell Edges

 1 Cell centre
 TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX i  ic 
   R DL – R DwPTS + SFP SSF   R DwPTS
ABS
f Scaling =  
SFP DL
    
 -------------------------------------------------------------------------------------------------------------------------------------------------
1 1 - Cell edge
 TX i  ic 
 80  R DL

• Method 2: ABS Patterns Used Throughout the Cell

TX i  ic  TX i  ic  TX  ic  TX i  ic  TX  ic 
 SFPDL   R DL – R DwPTS +  SFPSSF   R DwPTS
i i
ABS
f Scaling = -------------------------------------------------------------------------------------------------------------------------------------------------
1 1 -
TX i  ic 
80  R DL
For more information on SFP DL and SFP SSF , see "Subframe Pattern Collision Calculation" on page 497.
Output
TX  ic 
i
• R DL : Amount of downlink resources in the cell TXi(ic).
6.4.6.1.2 Calculation of Uplink Cell Resources

Input

TX i  ic 
TX i  ic 
• N FB – PUCCH : Average number of PUCCH frequency blocks per frame defined in TXi(ic) frame configuration or,
otherwise, in the global network settings.
TX i  ic 
the cell TXi(ic).
TX i  ic 
TX i  ic 
TX i  ic 
TX  ic 
i
• N SF – UL : Number of uplink subframes in the frame for the cell TXi(ic). It is equal to 10 for FDD frequency bands, and
TX  ic 
i
bands.
N SF – UL and N TDD – SSF are determined as follows:
543
Configuration i i
N SF – UL N TDD – SSF
FDD 10 0
DSUUU-DSUUU 6 2
DSUUD-DSUUD 4 2
DSUDD-DSUDD 2 2
DSUUU-DSUUD 5 2
DSUUU-DDDDD 3 1
DSUUD-DDDDD 2 1
DSUDD-DDDDD 1 1
UpPTS is used for SRS (sounding reference signals) if the UpPTS duration is 1 OFDM
symbol, and for SRS and PRACH if the UpPTS duration is 2 OFDM symbols. Therefore, the
uplink cell capacity can be determined without considering the UpPTS symbols.
Calculations
calculations below.
The number of modulation symbols (resource elements) per resource block is calculated as follows:
W FB
N SCa – FB = ---------
-
F
The total number of modulation symbols (resource elements) in uplink is calculated as follows:
N Sym – UL =  N FB – N FB – PUCCH  N Sym  SRB  N SF – UL
i i i i
TX i  ic 
The total uplink cell resources, i.e., R UL , are calculated as follows:
R UL = N Sym – UL – O ULSRS – O ULDRS
Where O ULSRS is the overhead corresponding to the uplink sounding reference signals, and O ULDRS is the overhead
corresponding to the uplink demodulation reference signals. These control channel overheads are calculated as follows:
Calculations of uplink control channel overheads
The uplink sounding reference signals are transmitted on 1 symbol duration in each uplink subframe. Therefore,
TX i  ic  N SCa – FB TX i  ic 
-  N Sym
O ULSRS = --------------------- – UL
N Sym  SRB
The uplink demodulation reference signals are transmitted on two symbol durations in each uplink subframe.
Therefore,
TX i  ic  N SCa – FB TX i  ic 
O ULDRS = 2  ---------------------
-  N Sym – UL
N Sym  SRB
TX i  ic 
Once R UL is known, it is scaled down according to the ICIC mode used by the cell TXi(ic) depending on whether the uplink
cell resources are being calculated for a cell-centre or cell-edge pixel, subscriber, or mobile.
ICIC ABS
R UL = R UL  f Scaling  f Scaling
ICIC
f Scaling is calculated as follows for the different ICIC modes:
544
ICIC
ICIC cell resource scaling factor f Scaling for
ICIC mode
No FFR 1 1
TX  ic 
i
N FB – CEx
Time-switched FFR 1 --------------------
TX i  ic 
N FB
Hard FFR -------------------- --------------------
N FB N FB
Soft FFR ------------------------------------------ --------------------
N FB N FB
TX  ic  TX  ic  TX  ic  TX  ic  TX i  ic 
i i i i
N FB N FB – CEx
Partial soft FFR ------------------------------------------------------------------------------------------------------ --------------------
N FB N FB
ABS
f Scaling is calculated as follows:
• Method 1: ABS Patterns Used Only at Cell Edges
 1 Cell centre

 TX i  ic 

ABS
f Scaling =  SFP UL
 1
 ------------------------------ Cell edge
 80
• Method 2: ABS Patterns Used Throughout the Cell

TX  ic 
 SFPUL
i
ABS
f Scaling = -----------------------------
1 -
80
TX i  ic 
For more information on SFP UL , see "Subframe Pattern Collision Calculation" on page 497.
Output
TX i  ic 
• R UL : Amount of uplink resources in the cell TXi(ic).
6.4.6.2 Calculation UE Capacities

The UE category parameters define the maximum throughput that can be supported by a UE in downlink and uplink. The UE
capacities are calculated for the downlink and uplink as described in:
• "Calculation of Downlink UE Capacity" on page 545.
• "Calculation of Uplink UE Capacity" on page 546.
6.4.6.2.1 Calculation of Downlink UE Capacity

Input
• D Frame : Frame duration.

Max – DL
• N TBB  TTI : Maximum number of transport block bits per TTI (subframe) in downlink defined for a UE category.
TX i  ic 
545
TX  ic 
i
bands.
i i
FDD 10 0
DSUUU-DSUUU 2 2
DSUUD-DSUUD 4 2
DSUDD-DSUDD 6 2
DSUUU-DSUUD 3 2
DSUUU-DDDDD 6 1
DSUUD-DDDDD 7 1
DSUDD-DDDDD 8 1
Calculations
In LTE, the maximum throughput that can be supported by a user equipment is defined through its UE category parameter
Transport Block Size. This is the maximum number of transport block bits that the UE can carry per subframe.
The downlink UE capacity in terms of the maximum throughput supported by a UE in downlink is calculated as follows:
N i i 
Max Max – DL  SF – DL + N TDD – SSF
TP UE – DL = N TBB  TTI  ---------------------------------------------------
-
D Frame
The maximum transport block sizes defined by the 3GPP for different UE categories correspond to the following maximum
throughput capacities in FDD:
UE Category 1 2 3 4 5 6 7 8
Max – DL
N TBB  TTI (bits/TTI) 10296 51024 102048 150752 299552 301504 301504 2998560
Max
TP UE – DL (Mbps) 10.296 51.024 102.048 150.752 299.552 301.504 301.504 2998.560
Output
Max
• TP UE – DL : Maximum downlink throughput capacity of a UE category.
6.4.6.2.2 Calculation of Uplink UE Capacity

Input

Max – UL
• N TBB  TTI : Maximum number of transport block bits per TTI (subframe) in uplink defined for a UE category.
TX  ic 
i
• N SF – UL : Number of uplink subframes in the frame for the cell TXi(ic). It is equal to 10 for FDD frequency bands, and
TX  ic 
i
bands.
N SF – UL and N TDD – SSF are determined as follows:
Configuration N SF – UL N TDD – SSF
FDD 10 0
DSUUU-DSUUU 6 2
546
Configuration i i
N SF – UL N TDD – SSF
DSUUD-DSUUD 4 2
DSUDD-DSUDD 2 2
DSUUU-DSUUD 5 2
DSUUU-DDDDD 3 1
DSUUD-DDDDD 2 1
DSUDD-DDDDD 1 1
Calculations
In LTE, the maximum throughput that can be supported by a user equipment is defined through its UE category parameter
Transport Block Size. This is the maximum number of transport block bits that the UE can carry per subframe.
The uplink UE capacity in terms of the maximum throughput supported by a UE in uplink is calculated as follows:
TX  ic 
i
Max Max – UL N SF – UL
TP UE – UL = N TBB  TTI  -----------------
D Frame
The maximum transport block sizes defined by the 3GPP for different UE categories correspond to the following maximum
throughput capacities in FDD:
UE Category 1 2 3 4 5 6 7 8
Max – UL
N TBB  TTI (bits/TTI) 5160 25456 51024 51024 75376 51024 102048 1497760
Max
TP UE – UL (Mbps) 5.16 25.456 51.024 51.024 75.376 51.024 102.048 1497.760
Output
Max
• TP UE – UL : Maximum uplink throughput capacity of a UE category.
6.4.6.3 Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput,

and Per-user Throughput Calculation
Channel throughputs are calculated for the entire channel resources allocated to the pixel, subscriber, or mobile Mi. Cell
capacities are similar to channel throughputs but upper-bound by the maximum downlink and uplink traffic loads. Allocated
bandwidth throughputs are calculated for the number of used frequency blocks in uplink allocated to the pixel, subscriber, or
mobile Mi. Per-user throughputs are calculated by dividing the downlink cell capacities or uplink allocated bandwidth
throughputs by the average number of downlink or uplink users defined for the cell, respectively.
Input
TX i  ic 
• TL DL – Max : Maximum downlink traffic load for the cell TXi(ic).
TX  ic 
i
• TL UL – Max : Maximum uplink traffic load for the cell TXi(ic).
TX i  ic 
• R DL : Amount of downlink resources in the cell TXi(ic) as calculated in "Calculation of Total Cell Resources" on
page 538.
TX i  ic 
• R UL : Amount of uplink resources in the cell TXi(ic) as calculated in "Calculation of Total Cell Resources" on
page 538.
•  M : Bearer efficiency (bits/symbol) of the bearer assigned to the pixel, subscriber, or mobile Mi in the downlink in
i
B DL
"C/(I+N) and Bearer Calculation (DL)" on page 518.

•  M : Bearer efficiency (bits/symbol) of the bearer assigned to the pixel, subscriber, or mobile Mi in the uplink in "C/
i
B UL
(I+N) and Bearer Calculation (UL)" on page 532.

TX i  ic 
• T SU – MIMO – UL : SU-MIMO threshold defined in the reception equipment of the cell TXi(ic).
547
M
i
• T SU – MIMO – DL : SU-MIMO threshold defined in the reception equipment of the pixel, subscriber, or mobile Mi.
TX  ic 
i
• T MU – MIMO – UL : MU-MIMO threshold defined in the reception equipment of the cell TXi(ic).
TX  ic 
i
• T MU – MIMO – DL : MU-MIMO threshold defined in the reception equipment of the pixel, subscriber, or mobile Mi.
TX i  ic 
• G MU – MIMO – UL : Average number of co-scheduled MU-MIMO users in uplink for the cell TXi(ic).
TX i  ic 
• G MU – MIMO – DL : Average number of co-scheduled MU-MIMO users in downlink for the cell TXi(ic).
M TX  ic 
BLER  BDL : Downlink block error rate read from the BLER vs. CINR PDSCH graph available in the reception equipment
i i
•
assigned to the terminal used by the pixel, subscriber, or mobile Mi.
M M
BLER  BUL : Uplink block error rate read from the BLER vs. CINR PUSCH PUCCH graph available in the reception
i i
•
 
equipment assigned to the cell TXi(ic).
DL
• T SCell : Secondary cell activation threshold of the reception equipment assigned to the pixel, subscriber, or mobile Mi.
UL
• T SCell : Secondary cell activation threshold of the reception equipment assigned to the cell TXi(ic).
Mi
• f TP – Scaling : Throughput scaling factor defined in the properties of the service used by the pixel, subscriber, or mobile
Mi.
M
i
• TP Offset : Throughput offset defined in the properties of the service used by the pixel, subscriber, or mobile Mi.
TX i  ic 
cell TXi(ic).
Mi
• N FB – UL : Number of frequency blocks used by the pixel, subscriber, or mobile Mi after uplink bandwidth allocation as
calculated in "C/(I+N) and Bearer Calculation (UL)" on page 532.
TX  ic 
i
• N Users – DL : Number of users connected to the cell TXi(ic) in downlink.
TX i  ic 
• N Users – UL : Number of users connected to the cell TXi(ic) in uplink.
Calculations
Downlink:
TX  ic 
i
R DL  Mi
Mi B DL
• Peak RLC Channel Throughput: CTP P – DL = ---------------------------------
D Frame
In the above formula, the actual value of D Frame is used to calculate the channel throughput for coverage predictions,
while D Frame = 1 sec for Monte Carlo simulations.
TX i  ic 
For proportional fair schedulers, the channel throughput is increased by the multi-user diversity gain G MUG – DL read
Mi
from the scheduler properties for the bearer B DL , Mobility  M i  , and the number of users connected to the cell in
downlink.
TX i  ic 
R DL  Mi
Mi B TX  ic 
DL i
CTP P – DL = ---------------------------------  G MUG – DL
D Frame
TX i  ic  Mi
Max
G MUG – DL = 1 if CINR PDSCH  CINR MUG
If the multi-user diversity gain for the exact value of the number of connected users is not available in the graph, it is
interpolated from the gain values available for the numbers of users just less than and just greater than the actual
number of users.
MIMO – SU-MIMO Gain:
548
If the cell supports MIMO, and according to the Mi diversity mode depending on the SU-MIMO threshold and criterion,
Max – M
i
the SU-MIMO gain, G SU – MIMO – DL , corresponding to the bearer is applied to its efficiency. The gain is read from the
TX  ic  M
i i
properties of the reception equipment assigned to the pixel, subscriber, or mobile Mi for N Ant – TX , N Ant – RX ,
M
Mobility  M i  , BLER  B DL .
i
Atoll also takes into account the SU-MIMO Gain Factor f SU – MIMO defined for the clutter class where the pixel,
subscriber, or mobile Mi is located.
Max – M i
 =    1 + f SU – MIMO  G SU – MIMO – DL – 1 
Mi Mi   
B DL B DL
If the Max SU-MIMO Gain for the exact value of the C/(I+N) is not available in the table, it is interpolated from the gain
values available for the C/(I+N) just less than and just greater than the actual C/(I+N).
MIMO – MU-MIMO Gain (for throughput coverage predictions):
If the cell supports MU-MIMO, and according to the Mi diversity mode depending on the MU-MIMO threshold and
TX i  ic 
criterion, the MU-MIMO gain, G MU – MIMO – DL , which is the average number of co-scheduled users, is applied to the
channel throughput.
CTP P – DL = CTP P – DL  G MU – MIMO – DL
M M M
Effective RLC Channel Throughput: CTP E – DL = CTP P – DL   1 – BLER  B DL 
i i i
•
  
Mi
• Application Channel Throughput: CTP A – DL = CTP E – DL  ------------------------
- – TP Offset
100
• Peak RLC Cell Capacity: Cap P – DL = CTP P – DL  TL DL – Max
M M M
Effective RLC Cell Capacity: Cap E – DL = Cap P – DL   1 – BLER  B DL 
i i i
•
  
Mi
Mi f TP – Scaling
Mi Mi
• Application RLC Capacity: Cap A – DL = Cap E – DL  ------------------------
- – TP Offset
100
Mi
Mi Cap P – DL
• Peak RLC Throughput per User: PUTP P – DL = -----------------------
-
TX i  ic 
N Users – DL
Mi
i
M Cap E – DL
• Effective RLC Throughput per User: PUTP E – DL = -----------------------
-
TX i  ic 
N Users – DL
Mi
i
M
i
M f TP – Scaling M
i
• Application Throughput per User: PUTP A – DL = PUTP E – DL  ------------------------
- – TP Offset
100
549
Carrier Aggregation:
Aggregated throughputs are calculated by summing the throughputs from each serving
cell taking part in carrier aggregation for any LTE-A pixel, subscriber, or mobile. If the sum
of the throughputs exceeds the maximum throughput supported by the UE category, the
aggregated throughput is scaled down by the following ratio:
 Mi 
Min  TP UE – DL  CTP P – DL
Max
 TX i  ic  
r = ---------------------------------------------------------------------------
Mi
 CTP P – DL
TX i  ic 
Only secondary cells whose PDSCH C/(I+N) is higher than or equal to the secondary cell
DL
activation threshold ( T SCell ) defined in the terminal reception equipment properties are
activated for aggregation.
Uplink:
TX i  ic 
R UL  Mi
M B UL
i
• Peak RLC Channel Throughput: CTP P – UL = ---------------------------------
D Frame
TX i  ic 
For proportional fair schedulers, the channel throughput is increased by the multi-user diversity gain G MUG – UL read
Mi
from the scheduler properties for the bearer B UL , Mobility  M i  , and the number of users connected to the cell in
uplink.
TX i  ic 
R UL  Mi
Mi B TX  ic 
UL i
CTP P – UL = ---------------------------------  G MUG – UL
D Frame
TX i  ic  Mi
Max
G MUG – UL = 1 if CINR PUSCH PUCCH  CINR MUG
number of users.
If the cell supports MIMO, and according to the Mi diversity mode depending on the SU-MIMO threshold and criterion,
Max – TX i  ic 
the SU-MIMO gain, G SU – MIMO – UL , corresponding to the bearer is applied to its efficiency. The gain is read from the
M TX  ic  M
properties of the reception equipment assigned to the TXi(ic) for N Ant – RX , N Ant – TX , Mobility  M i  , BLER  B UL .
i i i
Max – TX i  ic 
 =    1 + f SU – MIMO  G SU – MIMO – UL – 1 
Mi Mi   
B UL B UL
550
MIMO – MU-MIMO Gain (for throughput coverage predictions):

If the cell supports MU-MIMO, and according to the Mi diversity mode depending on the MU-MIMO threshold and
TX  ic 
i
criterion, the MU-MIMO gain, G MU – MIMO – UL , which is the average number of co-scheduled users, is applied to the
channel throughput.
M M TX  ic 
i i i
CTP P – UL = CTP P – UL  G MU – MIMO – UL
M M M
Effective RLC Channel Throughput: CTP E – UL = CTP P – UL   1 – BLER  B UL 
i i i
•
  
Mi
Mi f TP – Scaling
Mi Mi
• Application Channel Throughput: CTP A – UL = CTP E – UL  ------------------------
- – TP Offset
100
M M TX  ic 
i i i
• Peak RLC Cell Capacity: Cap P – UL = CTP P – UL  TL UL – Max
M M M
Effective RLC Cell Capacity: Cap E – UL = Cap P – UL   1 – BLER  B UL 
i i i
•
  
Mi
Mi f TP – Scaling
Mi Mi
• Application Cell Capacity: Cap A – UL = Cap E – UL  ------------------------
- – TP Offset
100
M
i
Mi Mi N FB – UL
• Peak RLC Allocated Bandwidth Throughput: ABTP P – UL = CTP P – UL  -----------------
TX  ic 
-
i
N FB
M M M
Effective RLC Allocated Bandwidth Throughput: ABTP E – UL = ABTP P – UL   1 – BLER  B UL 
i i i
•
  
Mi
• Application Allocated Bandwidth Throughput: ABTPA – UL = ABTP E – UL  ------------------------
- – TPOffset
100
M  Cap M i M 
Peak RLC Throughput per User: PUTP P – UL = Min  -----------------------
P – UL
- ABTPP –i UL
i
•
 TX
i
 ic  
Mi  Cap Mi M 
• Effective RLC Throughput per User: PUTP E – UL = Min  -----------------------
E – UL
 TX
i
 ic  
Mi
Mi f TP – Scaling
Mi Mi
• Application Throughput per User: PUTP A – UL = PUTP E – UL  ------------------------
- – TP Offset
100
cell taking part in carrier aggregation for any LTE-A pixel, subscriber, or mobile. If the sum
of the throughputs exceeds the maximum throughput supported by the UE category, the
aggregated throughput is scaled down by the following ratio:
 Mi 
Min  TP UE – UL  CTP P – UL
Max
 
 TX i  ic  
r = ---------------------------------------------------------------------------
Mi
 CTP P – UL
TX i  ic 
Only secondary cells whose PDSCH C/(I+N) is higher than or equal to the secondary cell
DL
activation threshold ( T SCell ) defined in the terminal reception equipment properties and
UL
PUSCH C/(I+N) is higher than or equal to the secondary cell activation threshold ( T SCell )
defined in the cell reception equipment properties are activated for aggregation.
551
Output
M
i
• CTP P – DL : Downlink peak RLC channel throughput at the pixel, subscriber, or mobile Mi.
M
i
• CTP E – DL : Downlink effective RLC channel throughput at the pixel, subscriber, or mobile Mi.
Mi
• CTP A – DL : Downlink application channel throughput at the pixel, subscriber, or mobile Mi.
M
i
• Cap P – DL : Downlink peak RLC cell capacity at the pixel, subscriber, or mobile Mi.
Mi
• Cap E – DL : Downlink effective RLC cell capacity at the pixel, subscriber, or mobile Mi.
Mi
• Cap A – DL : Downlink application cell capacity at the pixel, subscriber, or mobile Mi.
Mi
• PUTP P – DL : Downlink peak RLC throughput per user at the pixel, subscriber, or mobile Mi.
Mi
• PUTP E – DL : Downlink effective RLC throughput per user at the pixel, subscriber, or mobile Mi.
M
i
• PUTP A – DL : Downlink application throughput per user at the pixel, subscriber, or mobile Mi.
Mi
• CTP P – UL : Uplink peak RLC channel throughput at the pixel, subscriber, or mobile Mi.
Mi
• CTP E – UL : Uplink effective RLC channel throughput at the pixel, subscriber, or mobile Mi.
Mi
• CTP A – UL : Uplink application channel throughput at the pixel, subscriber, or mobile Mi.
Mi
• Cap P – UL : Uplink peak RLC cell capacity at the pixel, subscriber, or mobile Mi.
M
i
• Cap E – UL : Uplink effective RLC cell capacity at the pixel, subscriber, or mobile Mi.
Mi
• Cap A – UL : Uplink application cell capacity at the pixel, subscriber, or mobile Mi.
Mi
• ABTP P – UL : Uplink peak RLC allocated bandwidth throughput at the pixel, subscriber, or mobile Mi.
Mi
• ABTP E – UL : Uplink effective RLC allocated bandwidth throughput at the pixel, subscriber, or mobile Mi.
Mi
• ABTP A – UL : Uplink application allocated bandwidth throughput at the pixel, subscriber, or mobile Mi.
Mi
• PUTP P – UL : Uplink peak RLC throughput per user at the pixel, subscriber, or mobile Mi.
Mi
• PUTP E – UL : Uplink effective RLC throughput per user at the pixel, subscriber, or mobile Mi.
Mi
• PUTP A – UL : Uplink application throughput per user at the pixel, subscriber, or mobile Mi.

Atoll LTE module includes a number of scheduling methods which can be used for scheduling and radio resource allocation
during Monte Carlo simulations. These resource allocation algorithms are explained in "Scheduling and Radio Resource
Allocation" on page 552 and the calculation of user throughputs is explained in "User Throughput Calculation" on page 561.
6.4.7.1 Scheduling and Radio Resource Allocation

Input
TX  ic 
i
TX i  ic 
TX i  ic 
• N Users – Max : Maximum number of users defined for the cell TXi(ic).
Mi
• p QCI : QCI priority of the service accessed by a mobile Mi.
Mi
• p Service : User-defined priority of the service accessed by a mobile Mi.
552
M
i
• TPD Min – DL : Downlink minimum throughput demand for the service accessed by a mobile Mi.
M
i
• TPD Min – UL : Uplink minimum throughput demand for the service accessed by a mobile Mi.
M
i
• TPD Max – DL : Downlink maximum throughput demand for the service accessed by a mobile Mi.
Mi
• TPD Max – UL : Uplink maximum throughput demand for the service accessed by a mobile Mi.
TX i  ic 
• G MU – MIMO – UL : Average number of co-scheduled MU-MIMO users in uplink for the cell TXi(ic).
TX i  ic 
• G MU – MIMO – DL : Average number of co-scheduled MU-MIMO users in downlink for the cell TXi(ic).
•  Mi : Bearer efficiency (bits/symbol) of the bearer assigned to the mobile Mi in the downlink in "C/(I+N) and Bearer
B DL

•  M : Bearer efficiency (bits/symbol) of the bearer assigned to the mobile Mi in the uplink in "C/(I+N) and Bearer
i
B UL

•  M : Bearer efficiency (bits/symbol) of the highest bearer of the service being used by the mobile Mi in the
i
B DL – Highest
downlink.
•  M : Bearer efficiency (bits/symbol) of the highest bearer of the service being used by the mobile Mi in the
i
B UL – Highest
uplink.
M TX  ic 
BLER  B DL : Downlink block error rate read from the BLER vs. CINR PDSCH graph available in the reception equipment
i i
•
 
assigned to the terminal used by the mobile Mi.
M M
BLER  B UL : Uplink block error rate read from the BLER vs. CINR PUSCH PUCCH graph available in the reception
i i
•
 
DL
• T SCell : Secondary cell activation threshold of the reception equipment assigned to the mobile Mi.
UL
• T SCell : Secondary cell activation threshold of the reception equipment assigned to the cell TXi(ic).
Mi
• f TP – Scaling : Throughput scaling factor defined in the properties of the service used by the mobile Mi.
M
i
• TP Offset : Throughput offset defined in the properties of the service used by the mobile Mi.
Mi
• CTP P – DL : Downlink peak RLC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
page 538.
Mi
• CTP E – DL : Downlink effective RLC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
page 538.
Mi
• CTP P – UL : Uplink peak RLC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
page 538.
Mi
• CTP E – UL : Uplink effective RLC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
page 538.
Mi
• ABTP P – UL : Uplink peak RLC allocated bandwidth throughput at the mobile Mi as calculated in "Throughput
Mi
• ABTP E – UL : Uplink effective RLC allocated bandwidth throughput at the mobile Mi as calculated in "Throughput
Max
• TP UE – DL : Maximum downlink throughput capacity of the UE category of the mobile Mi as calculated in "Calculation
of Downlink UE Capacity" on page 545.
Max
• TP UE – UL : Maximum uplink throughput capacity of the UE category of the mobile Mi as calculated in "Calculation of
Uplink UE Capacity" on page 546.
553
Calculations
The following calculations are described for any cell TXi(ic) containing the users Mi for which it is the best server.
Mobile Selection:
TX  ic 
i
The scheduler selects N Users mobiles for the scheduling and RRM process. If the Monte Carlo user distribution has generated
TX i  ic 
a number of users which is less than N Users – Max , the scheduler keeps all the mobiles generated for the cell TXi(ic).
N Users = Min  N Users – Max N Users – Generated
i i i
 
TX  ic 
Sel i
For a cell, mobiles M i  N Users are selected for RRM by the scheduler.
Calculation of Actual Minimum and Maximum Throughput Demands:

If the service maximum throughput demand downgrading is active (for more information, see the Administrator Manual), the
maximum throughput demand of each user will be downgraded as follows:
  Sel 
Mi
Sel
Mi  Sel
Mi
Sel
Mi B DL 
Downlink: TPD Max – DL = Max TPD Min – DL TPD Max – DL  ------------------------------

  Sel 
 Mi 
 B DL – Highest
  Sel 
Mi
Sel
Mi  Sel
Mi
Sel
Mi B UL 
Uplink: TPD Max – UL = Max TPD Min – UL TPD Max – UL  -------------------------------

  Sel 
 Mi 
 B UL – Highest
Then, depending on the selected target throughput of the scheduler assigned to the cell TXi(ic), the actual minimum and
maximum throughput demands can be considered as the peak RLC, effective RLC, or application throughput. Therefore:
• Target Throughput = Peak RLC Throughput
Sel Sel
M M
i i
Downlink: TPD Min – DL , TPD Max – DL
Sel Sel
Mi  Mi Mi 
Uplink: TPD Min – UL , Min  TPD Max – UL ABTP P – UL
 
• Target Throughput = Effective RLC Throughput

Sel Sel
Sel Mi Sel Mi
Mi TPD Min – DL Mi TPD Max – DL
Downlink: TPD Min – DL = ---------------------------------------------
- , TPD Max – DL = ---------------------------------------------
-
Sel Sel
  Mi     Mi  
 1 – BLER  B DL    1 – BLER  BDL  
     
Sel
Sel  Mi Mi 
Sel i
M Sel Min  TPD Max – UL ABTP P – UL
Mi TPD Min – UL Mi  
Uplink: TPD Min – UL = ---------------------------------------------- , TPD Max – UL = -------------------------------------------------------------------------
Sel Sel
  Mi     Mi  
 1 – BLER  B UL    1 – BLER  B UL  
     
• Target Throughput = Application Throughput

Sel Sel
Sel Mi Mi Sel Mi Mi
TPD Min – DL + TP Offset TPD Max – DL + TP Offset
M M
i i
Downlink: TPD Min – DL = ------------------------------------------------------------------------------ , TPD Max – DL = -----------------------------------------------------------------------------
-
Sel Sel
  Mi   Mi   Mi   Mi
 1 – BLER  B DL    f TP – Scaling  1 – BLER  B DL    f TP – Scaling
     
554
Sel
Sel M M
i i
TPD Min – UL + TP Offset
M
i
Uplink: TPD Min – UL = ------------------------------------------------------------------------------ ,
Sel
  Mi   Mi
 1 – BLER  BUL    f TP – Scaling
  
Sel
 Mi Mi  Mi
Sel Min  TPD Max – UL ABTP P – UL + TP Offset
Mi  
TPD Max–UL = --------------------------------------------------------------------------------------------------
Sel
  Mi   Mi
 1 – BLER  B UL    f TP – Scaling
  
The Min() function selects the lower of the two values. This calculation is performed in order to limit the maximum uplink
throughput demand to the maximum throughput that a user can get in uplink using the allocated bandwidth (number of
frequency blocks) calculated for it in "C/(I+N) and Bearer Calculation (UL)" on page 532.
Resource Allocation for Minimum Throughput Demands:
• For their minimum throughput demands, LTE-A users are only scheduled on their
primary serving cells.
Sel Sel
TX i  ic  Mi Mi
Sel
1. Atoll sorts the M i  N Users in order of decreasing effective service priority (combination of p QCI and p Service ).
The mobiles are sorted first in the order of decreasing QCI priority (as listed in the table below) and then in the order
of decreasing user-defined service priority within a QCI. For example:
QoS class identifier 1 2 3 4 5 6 7 8 9
QCI priority 2 4 3 5 1 6 7 8 9
Sel Sel
Sel Mi Mi
Mi p QCI p Service
1 1 i
2 :
3 0
: 2 i
: :
: 0
: 3 i
: :
: 0
: 4 i
: :
: 0
: 5 i
: :
: 0
: 6 i
: :
: 0
: 7 i
: :
: 0
: 8 i
555
Sel Sel
Sel M M
Mi p QCI
i i
p Service
: :
: 0
: 9 i
: :
: 0
: NULL i
: :
N 0
TX  ic 
Sel Sel i
2. Starting with M i = 1 up to M i = N Users , Atoll allocates the downlink and uplink resources required to satisfy
each user’s minimum throughput demands in downlink and uplink as follows:
Sel Sel
Sel Mi Sel Mi
Mi TPD Min – DL Mi TPD Min – UL
R Min – DL = --------------------------
- and R Min – UL = --------------------------
-
Sel Sel
Mi Mi
CTP P – DL CTP P – UL
3. Atoll stops the resource allocation in downlink or uplink,

Sel
Mi TX i  ic 
• When/If in downlink  R Min – DL = TL DL – Max , i.e., the resources available in downlink have been used up for
Sel
Mi
satisfying the minimum throughput demands of the mobiles.

Sel
Mi TX i  ic 
• When/If in uplink  R Min – UL = TL UL – Max , i.e., the resources available in uplink have been used up for
Sel
M
i
4. Mobiles which are active DL+UL must be able to get their minimum throughput demands in both UL and DL in order
to be considered connected DL+UL. If an active DL+UL mobile is only able to get its minimum throughput demand in
one direction, it is rejected, and the resources that were allocated to it in the one direction in which it was able to get
a throughput are allocated to other mobiles.
Sel Max
TP UE – DL Mi
5. Mobiles with minimum throughput demands higher than their UE capacities, i.e., R Min – DL  --------------------
Sel
- or
Mi
CTP P – DL
Sel Max
Mi TP UE – UL
R Min – UL  --------------------
Sel
- , are rejected due to No Service.
Mi
CTP P – UL
6. Mobiles which are active UL and whose minimum throughput demand in UL is higher than the uplink allocated
Sel Sel
Mi Mi
bandwidth throughput ( TPD Min – UL  ABTP P – UL ) are rejected due to Resource Saturation.
Sel Sel
Mi TX i  ic  Mi TX i  ic 
7. If  R Min – DL  TL DL – Max or  R Min – UL  TL UL – Max , and all the minimum throughput resources demanded by
Sel Sel
Mi Mi
the mobiles have been allocated, Atoll goes to the next step for allocating resources to satisfy the maximum
throughput demands.
Backhaul Saturation:
If at this stage, a site’s downlink or uplink effective RLC aggregate throughput exceeds its maximum downlink or uplink S1
interface throughput, respectively, mobiles are rejected one by one due to Backhaul Saturation, starting from the mobile with
the lowest priority service, among all the cells of the site in order to reach a downlink or uplink effective RLC aggregate site
throughput ≤ the site’s maximum downlink or uplink S1 interface throughput.
Resource Allocation for Maximum Throughput Demands:
For each cell, the remaining cell resources available are:
556
Sel

i i i
Downlink: R Rem – DL = TL DL – Max – R Min – DL
Sel
M
i
Sel

i i i
Uplink: R Rem – UL = TL UL – Max – R Min – UL
Sel
M
i
For each mobile, the remaining throughput demands are either the maximum UE capacities or the difference between the
maximum and the minimum throughput demands, whichever is smaller:
Sel Sel Sel

M
i  M
i
M
i Max 
Downlink: TPD Rem – DL = Min  TPD Max – DL – TPD Min – DL TP UE – DL
 
Sel Sel Sel

Mi  Mi Mi
Max 
Uplink: TPD Rem – UL = Min  TPDMax – UL – TPD Min – UL TP UE – UL
 
For their maximum throughput demands, LTE-A users are scheduled separately on each of their serving cells, primary and
Sel Sel
Mi Mi
secondary. Each user’s remaining throughput demand ( TPD Rem – DL and TPD Rem – UL ) is distributed over each of its serving
cells proportionally to the resources available on each serving cell and to the user’s downlink effective RLC channel throughput
or uplink effective RLC allocated bandwidth throughput on each of its serving cell:
TX i  ic  Mi
Sel
M
Sel
M R Rem – DL  CTP E – DL
i i Server n Server n
Downlink: TPD Rem – DL = TPD Rem – DL  --------------------------------------------------------------------------------------------------
5
-
Server n
TX i  ic  Mi
  RRem – DL Server n
 CTP E – DL
Server n

n=1
TX i  ic  Mi
Sel
Mi
Sel
Mi R Rem – UL  ABTP E – UL
Server n Server n
Uplink: TPD Rem – UL = TPD Rem – UL  ------------------------------------------------------------------------------------------------------
5
-
Server n
TX i  ic  Mi
  RRem – UL Server n
 ABTP E – UL
Server n

n=1
You can add an option in the Atoll.ini file to have each user’s remaining throughput demand distributed over each of its serving
cells proportionally only to the resources available on each serving cell:
TX i  ic 
Sel Sel R Rem – DL
Mi Mi
Server n
Downlink: TPD Rem – DL = TPD Rem – DL  -----------------------------------------------------
5
-
Server n
TX i  ic 
  RRem – DL Server n

n=1
TX i  ic 
Sel Sel R Rem – UL
Mi Mi
Server n
Uplink: TPD Rem – UL = TPD Rem – UL  -----------------------------------------------------
5
-
Server n
TX  ic 
  RRem – UL 
i
Server n
n=1
For more information, see the Administrator Manual.

Only secondary cells whose PDSCH C/(I+N) is higher than or equal to the secondary cell activation threshold defined in the
DL
terminal reception equipment properties ( T SCell ) are activated for aggregation in downlink. Similarly, only secondary cells
whose PDSCH C/(I+N) and PUSCH C/(I+N) are both higher than or equal to the secondary cell activation threshold defined in
DL UL
the terminal and cell reception equipment properties ( T SCell and T SCell ), respectively, are activated for aggregation in uplink.
Within each serving cell, resource allocation for the maximum throughput demands is carried out according to the scheduler
used by that particular cell. For the remaining throughput demands of the mobiles, the following resource allocation methods
are available:
• Proportional Fair:
The goal of this scheduling method is to distribute resources among users fairly in such a way that, on the average,
each user gets the highest possible throughput that it can get under the radio conditions at its location.
557
Sel
Let the total number of users be N  M i .
i i
a. Each user’s channel throughput is increased by the multi-user diversity gain G MUG – DL or G MUG – UL read from the
Sel Sel
M M
i i Sel
scheduler properties for the downlink or uplink bearer ( B DL or B UL ), Mobility  M i  assigned to mobile
Sel
Mi , and the number of connected users, DL or UL, in the cell TXi(ic) in the iteration k-1.
Sel Sel Sel Sel

Mi Mi TX i  ic  Mi Mi TX i  ic 
CTP P – DL = CTP P – DL  G MUG – DL and CTP P – UL = CTP P – UL  G MUG – UL
Without MUG Without MUG
Sel Sel
Max Max
G MUG – DL = 1 if CINR PDSCH  CINR MUG and G MUG – UL = 1 if CINR PUSCH PUCCH  CINR MUG .
If the multi-user diversity gain for the exact value of the number of connected users is not available in the graph,
it is interpolated from the gain values available for the numbers of users just less than and just greater than the
actual number of users.
b. Atoll divides the remaining resources in the cell into equal parts for each user:
R Rem – DL R Rem – UL
--------------------- and --------------------
-
N N
c. Atoll converts the remaining throughput demands of all the users to their respective remaining resource
demands:
Sel Sel
Sel Mi Sel Mi
Mi TPD Rem – DL Mi TPD Rem – UL
RD Rem – DL = ---------------------------
- and RD Rem – UL = ---------------------------
-
Sel Sel
Mi Mi
Remaining resource demands of a user are given by the ratio between its remaining throughput demands and the
peak channel throughputs at the user’s location.
d. The resources allocated to each user by the Proportional Fair scheduling method for satisfying its maximum
throughput demands are:
Sel Sel TX i  ic  Sel Sel TX i  ic 

M
i  Mi R Rem – DL M
i  Mi R Rem – UL
R Max – DL = Min  RD Rem – DL --------------------
- and R Max – UL = Min  RD Rem – UL ---------------------
 N   N 
Each user gets either the resources it needs to achieve its maximum throughput demands or an equal share from
the remaining resources of the cell, whichever is smaller.
e. Atoll stops the resource allocation in downlink or uplink,
Sel
Mi TX i  ic 
• When/If in downlink  R Max – DL = R Rem – DL , i.e., the resources available in downlink have been used up
Sel
Mi
for satisfying the maximum throughput demands of the mobiles.

Sel
Mi TX i  ic 
• When/If in uplink  R Max – UL = R Rem – UL , i.e., the resources available in uplink have been used up for
Sel
Mi
satisfying the maximum throughput demands of the mobiles.

f. If the resources allocated to a user satisfy its maximum throughput demands, this user is removed from the list of
remaining users.
g. Atoll recalculates the remaining resources as follows:
Sel Sel
TX  ic  TX  ic  M M
 
i i i i
R Rem – DL = TL DL – Max – R Min – DL – R Max – DL and
Sel Sel
Mi Mi
Sel Sel
TX i  ic  TX i  ic  Mi Mi
R Rem – UL = TL UL – Max –  R Min – UL –  R Max – UL
Sel Sel
M M
i i
558
h. Atoll repeats the all the above steps for the users whose maximum throughput demands have not been satisfied
i i
until either R Rem – DL = 0 and R Rem – UL = 0 , or all the maximum throughput demands are satisfied.
• Round Robin:
The goal of this scheduling method is to allocate equal resources to users fairly.
Sel
Let the total number of users be N  M i .
a. Atoll divides the remaining resources in the cell into equal parts for each user:
--------------------- and --------------------
-
N N
b. Atoll converts the remaining throughput demands of all the users to their respective remaining resource
demands:
Sel Sel
Sel M Sel M
i i
RD Rem – DL = ---------------------------
- and RD Rem – UL = ---------------------------
-
Sel Sel
Mi Mi
c. The resources allocated to each user by the Round Robin scheduling method for satisfying its maximum

Mi  Mi R Rem – DL Mi  Mi R Rem – UL
 N   N 
d. Atoll stops the resource allocation in downlink or uplink,
Sel
Mi TX i  ic 
Sel
M
i
Sel
Mi TX i  ic 
Sel
Mi

e. If the resources allocated to a user satisfy its maximum throughput demands, this user is removed from the list of
remaining users.
f. Atoll recalculates the remaining resources as follows:
Sel Sel
R Rem – DL = TL DL – Max –  R Min – DL –  R Max – DL and
Sel Sel
Mi Mi
Sel Sel
R Rem – UL = TL UL – Max –  R Min – UL –  R Max – UL
Sel Sel
Mi Mi
g. Atoll repeats the all the above steps for the users whose maximum throughput demands have not been satisfied
• Proportional Demand:
The goal of this scheduling method is to allocate resources to users weighted according to their remaining throughput
demands. Therefore, the user throughputs for users with high throughput demands will be higher than those with low
559
throughput demands. In other words, this scheduler distributes channel throughput between users proportionally to
their demands.
a. Atoll converts the remaining throughput demands of all the users to their respective remaining resource
demands:
Sel Sel
Sel M Sel M
i i
i
M TPD Rem – DL M
i TPD Rem – UL
RD Rem – DL = ---------------------------
- and RD Rem – UL = ---------------------------
-
Sel Sel
Mi Mi
b. Atoll calculates the amount of effective remaining resources of the cell to distribute among the users as follows:
 Sel   Sel 
TX i  ic   TXi  ic  Mi  TX i  ic   TX i  ic  Mi 
R Eff – Rem – DL = Min  R Rem – DL
  
RD Rem – DL and R Eff – Rem – UL = Min  R Rem – UL

RD Rem – UL
 
 Sel
Mi   Sel
Mi 
c. The resources allocated to each user by the Proportional Demand scheduling method for satisfying its maximum
Sel Sel
Sel Mi Sel Mi
Mi TX i  ic  RD Rem – DL Mi TX i  ic  RD Rem – UL
R Max – DL = R Eff – Rem – DL  ----------------------------------
Sel
- and R Max – UL = R Eff – Rem – UL  ----------------------------------
Sel
-
Mi Mi
 RDRem – DL  RDRem – UL
Sel Sel
Mi Mi
• Max C/I:
The goal of this scheduling method is to achieve the maximum aggregate throughput for the cells. This is done by
allocating as much resources as needed to mobiles with high C/(I+N) conditions. As mobiles with high C/(I+N) can get
higher bearers, and therefore require less amount of resources, more mobiles can therefore be allocated resources
in the same frame, and the end-throughput for each cell will be the highest compared to other types of schedulers.
TX i  ic 
Sel
a. Atoll sorts the M i  N Users in order of decreasing PDSCH, or PUSCH and PUCCH C/(I+N), depending on whether
the allocation is being performed for the downlink or for the uplink.
b. Starting with the mobile with the highest rank, Atoll allocates the downlink and uplink resources required to
satisfy each user’s remaining throughput demands in downlink and uplink as follows:
Sel Sel
Sel M Sel M
i i
i
M TPD Rem – DL M
i TPD Rem – UL
R Max – DL = ---------------------------
- and R Max – UL = ---------------------------
-
Sel Sel
Mi Mi
c. Atoll stops the resource allocation in downlink or uplink,

Sel
M TX  ic 

i i
• When/If in downlink R Max – DL = R Rem – DL , i.e., the resources available in downlink have been used up
Sel
Mi

Sel
Mi TX i  ic 
Sel
Mi

Spatial Multiplexing with Multi-User MIMO:
MU-MIMO lets the system/scheduler work with parallel LTE frames. Therefore, many users can be co-scheduled on the same
resources. MU-MIMO can be used if the cell supports MU-MIMO, if the calculated value for the MU-MIMO criterion is higher
than the MU-MIMO threshold T MU – MIMO – DL or T MU – MIMO – UL , and the number of antenna ports is equal to 2 or more.
MU – MIMO
MU – MIMO Mi
In both downlink and uplink, each mobile Mi has a corresponding traffic load TL . However, the actual
MU – MIMO
resource consumptions of a mobile Mi are given by:
560
MU – MIMO MU – MIMO
MU – MIMO M MU – MIMO M
i i
i
M TL DL M
i TL UL
RC DL = -----------------------------------
- and RC UL = ------------------------------------
i i
G MU – MIMO – DL G MU – MIMO – UL
MU – MIMO – UL MU – MIMO – DL
M TX  ic  M TX  ic 
 
i i i i
Saturation occurs when RC UL = TL UL – Max or RC DL = TL DL – Max .
Backhaul Capacity Limitation:

Backhaul overflow ratios are calculated for each site as follows:
  Mi
Sel
Mi 
Sel 
 
  
 Sel
M i  Site

= Max  1 ------------------------------------------------------------------------------------------------------ and
Site
BHOFDL
Sel Sel
  Mi Mi  

Site
 TP S1 – DL –  R Min – DL  CTP E – DL 
  
 Sel
M i  Site 
  Mi
Sel
Mi 
Sel 
 
   R Max – UL  CTP E – UL
  
 Sel
 Site

= Max  1 ------------------------------------------------------------------------------------------------------

Site M
i
BHOFUL
Sel Sel
  Mi Mi  

Site
 TP S1 – UL –  R Min – UL  CTP E – UL 
  
 Sel
M i  Site 
Total Amount of Resources Assigned to Each Selected Mobile:

Sel
Atoll calculates the amounts of downlink and uplink resources allocated to each individual mobile M i (which can also be
referred to as the traffic loads of the mobiles) as follows:
Sel Sel
M M
 Mi
Sel
Mi 
Sel R i i 
 R Min – DL  CTP P – DL +  ------------------------------------------------ Site 
Sel
Mi
Sel
Mi    BHOF DL 
Downlink: TL DL = R DL = ------------------------------------------------------------------------------------------------------------------
Sel
Mi
CTP P – DL
Sel Sel
Mi Mi
 Mi
Sel
Mi 
Sel R 
 R Min – UL  CTP P – UL +  ------------------------------------------------ Site
-

Sel
Mi
Sel
Mi    BHOF UL 
Uplink: TL UL = R UL = ------------------------------------------------------------------------------------------------------------------
Sel
M
i
CTP P – UL
Output
Sel Sel
Mi Mi Sel
• TL DL = R DL : Downlink traffic load or the amount of downlink resources allocated to the mobile M i .
Sel Sel
Mi Mi
Sel
• TL UL = R UL : Uplink traffic load or the amount of uplink resources allocated to the mobile M i .
6.4.7.2 User Throughput Calculation

User throughputs are calculated for the percentage of resources allocated to each mobile selected by the scheduling for RRM
Sel
during the Monte Carlo simulations, M i .
cell taking part in carrier aggregation for any LTE-A mobile, limited by the maximum
throughput supported by the UE category.
561
Input
Sel
M
i Sel
• R DL : Amount of downlink resources allocated to the mobile M i as calculated in "Scheduling and Radio Resource
Sel
M
i Sel
• R UL : Amount of uplink resources allocated to the mobile M i as calculated in "Scheduling and Radio Resource
Sel
Mi
Sel
• CTP P – DL : Downlink peak RLC channel throughput at the mobile M i as calculated in "Throughput Calculation" on
page 538.
Sel
Mi
Sel
• CTP P – UL : Uplink peak RLC channel throughput at the mobile M i as calculated in "Throughput Calculation" on
page 538.
Sel
 Mi  TX i  ic 
• BLER  BDL  : Downlink block error rate read from the BLER vs. CINR PDSCH graph available in the reception
 
Sel
equipment assigned to the terminal used by the mobile M i .
Sel
 Mi  Mi
• BLER  BUL  : Uplink block error rate read from the BLER vs. CINR PUSCH PUCCH graph available in the reception
 
Sel
Mi
Sel
• f TP – Scaling : Throughput scaling factor defined in the properties of the service used by the mobile M i .
Sel
Mi
Sel
• TP Offset : Throughput offset defined in the properties of the service used by the mobile M i .
Calculations
Downlink:
Sel Sel Sel
Mi Mi Mi
• Peak RLC User Throughput: UTP P – DL = R DL  CTP P – DL
Sel Sel Sel
Mi Mi   Mi  
• Effective RLC User Throughput: UTP E – DL = UTP P – DL   1 – BLER  B DL  
  
Sel
Sel Sel Mi Sel
• Application User Throughput: UTP A – DL = UTP E – DL  ------------------------
- – TP Offset
100
Uplink:
Sel Sel Sel
Mi Mi Mi
• Peak RLC User Throughput: UTP P – UL = R UL  CTP P – UL
Sel Sel Sel
Mi Mi   Mi  
• Effective RLC User Throughput: UTP E – UL = UTP P – UL   1 – BLER  B UL  
  
Sel
Sel Sel Mi Sel
• Application User Throughput: UTP A – UL = UTP E – UL  ------------------------
- – TP Offset
100
Output
Sel
Mi Sel
• UTP P – DL : Downlink peak RLC user throughput at the pixel, subscriber, or mobile M i .
Sel
Mi
Sel
• UTP E – DL : Downlink effective RLC user throughput at the pixel, subscriber, or mobile M i .
Sel
Mi
Sel
• UTP A – DL : Downlink application user throughput at the pixel, subscriber, or mobile M i .
Sel
Mi
Sel
• UTP P – UL : Uplink peak RLC user throughput at the pixel, subscriber, or mobile M i .
562
Sel
M
i Sel
• UTP E – UL : Uplink effective RLC user throughput at the pixel, subscriber, or mobile M i .
Sel
M
i Sel
• UTP A – UL : Uplink application user throughput at the pixel, subscriber, or mobile M i .
6.5 Automatic Planning Algorithms

The following sections describe the algorithms for:
• "Automatic Neighbour Planning" on page 563.
• "Automatic Inter-technology Neighbour Planning" on page 567.
• "Automatic Frequency Planning Using the AFP" on page 570.
• "Automatic Physical Cell ID Planning Using the AFP" on page 572.
• "Automatic PRACH RSI Planning Using the AFP" on page 576.
6.5.1 Automatic Neighbour Planning

The intra-technology neighbour planning algorithm takes into account the cells of all the TBC transmitters. It means that the
cells of all the TBC transmitters of your ATL document are potential neighbours.
The cells to be allocated will be called TBA cells. They must fulfil the following conditions:
• They belong to the folder on which allocation has been executed. This can be the Transmitters folder or a group of
transmitters (subfolder).
Only TBA cells are assigned neighbours.
zone.
We assume a reference cell TXi(ic) and a candidate neighbour cell TXj(jc). When automatic allocation starts, Atoll checks the
1. The distance between both cells must be less than the user-definable maximum inter-site distance. If the distance
discarded.
Atoll calculates the effective distance between the reference cell and its candidate neighbour from the real distance
between them and the azimuths of their antennas:
Dist  CellA CellB  = D   1 + x  cos  – x  cos  
Where x = 0.3% so that the maximum variation in D does not to exceed 1%. D is stated in m.
Figure 6.3: Inter-Transmitter Distance Calculation
The formula above implies that two cells facing each other have a smaller effective distance than the actual distance.
Candidate neighbours are ranked in the order of increasing effective distance from the reference cell.
2. The calculation options,
563
• Force Co-site Cells as Neighbours: If selected, Atoll adds all the cells located on the same site as the reference cell
to the candidate neighbour list. The weight of this constraint can be defined. It is used to calculate the rank of each
neighbour, and its importance.
• Force Adjacent Cells as Neighbours: If selected, Atoll adds all the cells geographically adjacent to the reference
cell to the candidate neighbour list. The weight of this constraint can be defined. It is used to calculate the rank of
each neighbour, and its importance.
Determination of Adjacent Cells: Geographically adjacent cells are determined on the basis of their best server
coverage areas. A candidate neighbour cell TXi(ic) is considered adjacent to the reference cell TXi(ic) if there exists
at least one pixel of TXj(jc)’s best server coverage area where TXi(ic) is the second best server. The ranking of
adjacent neighbour cells increases with the number of such pixels. Adjacent cells are sorted in the order of
decreasing rank.
Figure 6.4: Determination of Adjacent Cells
• Force Adjacent Layers as Neighbours: If selected, Atoll adds all the cells adjacent across network layers to the
reference cell to the candidate neighbour list. The weight of this constraint is always the average of the Min and
Max values defined for the adjacency factor. This weight is used to calculate the rank of each neighbour and its
importance. Cells are considered adjacent across layers if they belong to different layers and have a coverage
overlap of at least one pixel.
• Force Symmetry: If selected, Atoll adds the reference cell to the candidate neighbour list of the its candidate
neighbour.
A symmetric neighbour relation is allowed only if the neighbour list of the reference cell is not already full. If TXj(jc)
is a neighbour of TXi(ic) but TXi(ic) is not a neighbour of TXj(jc), there can be two possibilities:
i. The neighbour list of TXj(jc) is not full, Atoll will add TXi(ic) to the end of the list.
ii. The neighbour list of TXj(jc) is full, Atoll will not be able to add TXi(ic) to the list, so it will also remove TXj(jc)
from the neighbour list of TXi(ic).
[Neighbours]
• Force Exceptional Pairs: This option enables you to force/forbid some neighbour relations. Exceptional pairs are
pairs of cells which will always or never be neighbours of each other.
If you select "Force exceptional pairs" and "Force symmetry", Atoll considers the constraints between exceptional
pairs in both directions so as to respect symmetry condition. On the other hand, if neighbourhood relationship is
forced in one direction and forbidden in the other, symmetry cannot be respected. In this case, Atoll displays a
warning in the Event viewer.
• Delete Existing Neighbours: If selected, Atoll deletes all the current neighbours and carries out a new neighbour
allocation. If not selected, the existing neighbours are kept in the list.
3. If the Use Coverage Conditions check box is selected, the coverage areas of TXi(ic) and TXj(jc) must have an overlap.
Otherwise, only the distance criterion is taken into account.
The overlapping zone ( S TX  ic   S TX  jc  ) is defined as follows
i j
• Here S TX  ic  is the surface area covered by the cell TXi(ic) that comprises all the pixels where:
i
564
• The distance to the cell TXi(ic) is less than or equal to the round-trip time distance corresponding to the cell’s
PRACH preamble format.
i i
• The received RSRP is greater than or equal to the cell’s Min RSRP: E DLRS  T RSRP .
i i i i i i
• The received RSRP is within E DLRS + O Individual + M HO and E DLRS + O Individual + M HO + M End .
TX i  ic 
M HO is the handover margin defined for the cell TXi(ic). When a global handover start value is used,
TX i  ic 
M HO = M Start . M Start and M End are global handover start and handover end values.
• S TX  jc  is the surface area covered by the cell TXj(jc) that comprises all the pixels where:
j
• The distance to the cell TXj(jc) is less than or equal to the round-trip time distance corresponding to the cell’s
PRACH preamble format.
j j
• The received RSRP is greater than or equal to the cell’s Min RSRP: E DLRS  T RSRP .
• The received RSRP with offset ( E DLRS + O Individual ) is the highest.
TX i  ic 
• If a global value of the minimum RSRP threshold ( T RSRP ) is set in the coverage
conditions dialogue, for each cell, Atoll uses the higher of the two values, i.e., global
value and the value defined for that cell.
• For calculating the overlapping coverage areas, Atoll uses the service with the lowest
body loss, the terminal that has the highest difference between gain and losses, and
the shadowing margin calculated using the defined cell edge coverage probability, if
the option is selected. The service and terminal are selected such that the selection
gives the largest possible coverage areas for the cells.
S TX  ic   S TX  jc 
i j
When the above conditions are met, Atoll calculates the percentage of the coverage area overlap ( ---------------------------------------  100 ),
S TX  ic 
i
and compares this value with the % Min Covered Area. TXj(jc) is considered a neighbour of TXi(ic) if
S TX  ic   S TX  jc 
i j
-  100  % Min Coverage Area .
--------------------------------------
S TX  ic 
i
Figure 6.5: Overlapping Zones
Next, Atoll calculates the importance of the automatically allocated neighbours. Atoll sorts the neighbours by decreasing
importance in order to keep the ones with high importance. If the maximum number of neighbours to be allocated to each
cell is exceeded, Atoll keeps the ones with high importance.
565

Only if the Delete Existing Neighbours option is not selected
Exceptional pair Only if the Force Exceptional Pairs option is selected 100 %
Only if the Force Co-site Cells as Neighbours option is
selected
Only if the Force Adjacent Layers as Neighbours option is
selected
Only if the Force Adjacent Cells as Neighbours option is
selected
Only if the % Min Covered Area is exceeded Importance Function (IF)
Only if the Force Neighbour Symmetry option is selected Importance Function (IF)
relationship
The importance is evaluated using an Importance Function (IF), which takes into account the following factors:
transmitter.
d-
 Di  = 1 – ----------
d max
d is the effective distance (in m). It corresponds to the real inter-transmitter distance ( D in m) weighted by the
azimuths of antennas.



Coverage 10%+20%{10%(Di)+90%(O)}+1%+9%(Di)
Adjacent cells
Co-site cells
Where:
566
co-site neighbours, adjacent layers, adjacent cells, and neighbours allocated based on
coverage overlapping.
causes.
In the results, Atoll lists only the cells for which it finds new neighbours. Cells whose channels have the same centre frequency
are listed as intra-carrier neighbours. Otherwise, neighbour cells are listed as inter-carrier neighbours.
[Neighbours]
[Neighbours]
6.5.2 Automatic Inter-technology Neighbour Planning

The inter-technology neighbour planning algorithm takes into account all the TBC transmitters (if the other technology is
GSM) or the cells of all the TBC transmitters (for any other technology than GSM). This means that all the TBC transmitters
(GSM) or the cells of all the TBC transmitters (all other technologies) of the linked document are potential neighbours.
The cells to be allocated in the main document will be called TBA cells. They must fulfil the following conditions:
• They belong to the folder on which allocation has been executed. This can be the Transmitters folder or a group of
transmitters (subfolder).
Only TBA cells are assigned neighbours.
zone.
We assume a reference cell A and a candidate neighbour B. When automatic allocation starts, Atoll checks following
conditions:
1. The distance between reference cell and the candidate neighbour must be less than the user-definable maximum
inter-site distance. If the distance is greater than this value, the candidate neighbour is discarded.
Atoll calculates the effective distance between the reference cell and its candidate neighbour from the real distance
between them and the azimuths of their antennas:
Dist  CellA CellB  = D   1 + x  cos  – x  cos  
567
Where x = 0.3% so that the maximum variation in D does not to exceed 1%. D is stated in m.
Figure 6.6: Inter-Transmitter Distance Calculation
The formula above implies that two cells facing each other have a smaller effective distance than the actual distance.
Candidate neighbours are ranked in the order of increasing effective distance from the reference cell.
2. The calculation options:
• CDMA carriers: This option is available when an LTE network is being co-planned with a UMTS, CDMA, or TD-
SCDMA network. This option enables you to select the CDMA carrier(s) that you want Atoll to consider as potential
neighbours of LTE cells. You may choose one or more carriers. Atoll will allocate only the cells using the selected
carriers as neighbours.
• Force co-site cells as neighbours: If selected, Atoll adds all the transmitters/cells located on the same site as the
reference cell in its candidate neighbour list. The weight of this constraint can be defined. It is used to calculate
the rank of each neighbour and its importance.
• Force exceptional pairs: This option enables you to force/forbid some neighbour relations. Exceptional pairs are
pairs of cells which will always or never be neighbours of each other.
• Delete existing neighbours: If selected, Atoll deletes all the current neighbours and carries out a new neighbour
allocation. If not selected, the existing neighbours are kept in the list.
3. Neighbour relation criterion:
• Allocation based on distance:
The allocation algorithm is based on the effective distance between the reference cell and its candidate
neighbour.
• Algorithm based on coverage overlapping:
The coverage areas of the reference cell A and the candidate neighbour B must overlap ( S A  S B ).
Two cases may exist for SA:
• 1st case: SA is the area where the cell A is the best serving cell, with a 0 dB margin.
This means that the reference signal energy per resource element received from A is greater than the
minimum required (Min RSRP), and is the highest one. .
• 2nd case: The margin is other than 0 dB. SA is the area where:
The reference signal energy per resource element received from A exceeds the minimum required (Min RSRP)
and is within a margin from the highest signal level.
Two cases may exist for SB:
• 1st case: SB is the area where the candidate neighbour is the best server. In this case, the margin must be set
to 0dB.
The signal level received from B exceeds the minimum required, and is the highest one.
• 2nd case: The margin is other than 0dB. SB is the area where:
The signal level received from B exceeds the minimum required and is within a margin from the best signal
level.
SA  SB
-  100 ) and compares this value with the %
Atoll calculates the percentage of the coverage area overlap ( -----------------
SA
SA  SB
-  100  % Min Covered Area .
Min Covered Area. B is considered a neighbour of A if -----------------
SA
Candidate neighbours are ranked in the order of decreasing coverage area overlap percentages.
568
Next, Atoll calculates the importance of the automatically allocated neighbours. Atoll sorts the neighbours by decreasing
importance in order to keep the ones with high importance. If the maximum number of neighbours to be allocated to each
cell is exceeded, Atoll keeps the ones with high importance.
The importance (%) of neighbours depends on the distance and on the reason of allocation:
• For allocation based on distance:
Neighbour cause When Importance value

Co-site transmitter/cell If the Force co-site cells as neighbours option is selected 100 %
Neighbour relation that fulfils d-
distance conditions d max
d is the effective distance between the reference cell and the neighbour and d max is the maximum inter-site
distance.
• For allocation based on coverage overlapping:
Neighbour cause When Importance value

Co-site transmitter/cell If the Force co-site cells as neighbours option is selected IF
The importance is evaluated using an Importance Function (IF), which takes into account the following factors:
transmitter.
d-
 Di  = 1 – ----------
d max
antennas.


IF
Yes 60%+40%{1/7%(Di)+6/7%(O)}
Where
569
causes.
In the results, Atoll displays only the cells for which it finds new neighbours.
6.5.3 Automatic Frequency Planning Using the AFP

The role of an Automatic Frequency Planning (AFP) tool is to assign frequencies (channels) to cells of a network such that the
overall network performance is optimised. In other words, the interference within the network is reduced as much as possible.
Co-channel interference is the main reason for overall network quality degradation in LTE. In order to improve network
performance, the LTE AFP tries to minimise co- and adjacent channel interference as much as possible while respecting any
constraints input to it. The main constraints are the resources available for allocation, i.e., the number of frequencies with
which the AFP can work, and the relationships to take into account, i.e., interference matrices, neighbours, and distance
between transmitters.
The AFP is based on a cost function which represents the interference level in the network. The aim of the AFP is to minimise
the cost. The best, or optimum, frequency plan is the one which corresponds to the lowest cost.
The following describes the AFP’s automatic planning method for carriers in LTE networks, which takes into account
interference matrices, neighbour relations, and distance between transmitters.
The AFP takes into account the cells of all the TBC transmitters. The cells to be allocated will be called TBA cells. They must
fulfil the following conditions:
• Their channel allocation status is not set to locked,
• They are located inside the focus zone.
zone.
6.5.3.1 Constraint and Relationship Weights

The AFP is based on a cost function which takes into account channel separation constraints based on the channel overlap
ratio as calculated in "Co- and Adjacent Channel Overlaps Calculation" on page 493. Channel separation is studied between
each TBA cell and its related cells. Atoll calculates the cost between each individual TBA and related cell, and then the overall
cost for the TBA cell.
Related cells of a TBA cell are:
• Its neighbours, if the check box "Existing neighbours" is selected,
Default weight  Neighbour = 0.5
• Cells that are listed in the interference matrix of the TBA cell,
Default weight  IM = 0.3
• Cells within the cell’s (or the default) minimum reuse distance, if the check box "Reuse distance" is selected,
Default weight  Dis tan ce = 0.2
The sum of the weights assigned to the above relations is 1.
You can modify these weights in your LTE document. The absolute values of the constraint weights are calculated from the
relative weights (%) defined in the Constraint Weights dialogue as follows:
570
% Neighbour
 Neighbour = ----------------------------------------------------------------------------------
-
% Neighbour + % IM + % Dis tan ce
% IM
 IM = ----------------------------------------------------------------------------------
-
% Dis tan ce
 Dis tan ce = ----------------------------------------------------------------------------------
-
6.5.3.2 Cost Calculation

The cost of the relation between the TBA cell and its related cell is calculated as follows:
TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  TX  ic  – TX j  jc 
$ = rO    Neighbour   Neighbour +  Dis tan ce   Dis tan ce  +  i
 IM IM
Where r O is the channel overlap ratio as calculated in "Co- and Adjacent Channel Overlaps Calculation" on
page 493.
 Neighbour is the importance of the relationship between the TBA cell and its related neighbour cell.  Neighbour is
calculated during automatic neighbour planning by Atoll as explained in "Automatic Neighbour Planning" on page 563. For
manual neighbour planning, this value is equal to 1.
 IM is the importance of the relationship between the TBA cell and its related interfering cell calculated as follows:
TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc 
 IM = r CCO   IM – CC + r ACO   IM – AC
 IM – CC and  IM – CC are respectively the co- and adjacent channel interference probabilities calculated as
i j i j
explained in "Interference Matrix Calculation" on page 578. r CCO and r ACO are the co- and adjacent channel
overlap ratios as calculated in "Co- and Adjacent Channel Overlaps Calculation" on page 493.
 Dis tan ce is the importance of the relationship between the TBA and its related cell with respect to the distance between
them.  Dis tan ce is calculated as explained in "Distance Importance Calculation" on page 579.
Atoll calculates the quality reduction factor for the TBA cell and its related cell from the cost calculated above as follows:
QRF = 1–$
The quality reduction factor is a measure of the cost of an individual relation.
The total cost of the current frequency plan for any TBA cell is given as follows, considering all the cells with which the TBA
cell has relations:
TX i  ic  TX i  ic  – TX j  jc 
$ Total = 1 –  QRF
TX j  jc 
And, the total cost of the current frequency plan for the entire network is simply the sum of the total TBA cell costs calculated
above, i.e.,
TX i  ic 
$ Total =  $ Total
TX  ic 
i
6.5.3.3 AFP Algorithm

The AFP algorithm is an iterative algorithm which:
• Calculates the cost (as describe d above) of the initial frequency plan,
• Tries different frequency plans in order to reduce the cost,
• Memorises the different frequency plans in order to determine the best one, i.e., the frequency plan which provides
the lowest total cost,
• Stops when it is unable to improve the cost of the network, and proposes the last known best frequency plan as the
solution.
571
6.5.4 Automatic Physical Cell ID Planning Using the AFP

In LTE, 504 physical cell IDs are available, numbered from 0 to 503. There are as many pseudo-random sequences defined in
the 3GPP specifications. Physical cell IDs are grouped into 168 unique cell ID groups (called SSS IDs in Atoll), with each group
containing 3 unique identities (called PSS IDs in Atoll). An SSS ID is thus uniquely defined by a number in the range of 0 to 167,
and a PSS ID is defined by a number in the range of 0 to 2.
Each cell’s downlink reference signals transmit a pseudo-random sequence corresponding to the physical cell ID of the cell.
The SSS and PSS are transmitted over the centre six frequency blocks independent of the channel bandwidths used by cells.
Mobiles synchronise there transmission and reception frequency and time by listening first to the PSS. Once they know the
PSS ID of the cell, they listen to the SSS of the cell in order to know the SSS ID. The combination of these two IDs gives the
physical cell ID and the associated pseudo-random sequence that is transmitted over the downlink reference signals.
Once the physical cell ID and the associated pseudo-random sequence is known to the mobile, the cell is recognized by the
mobile based on the received downlink reference signals. Downlink channel quality measurements are also made on the
downlink reference signals.
As can be understood from the above description, if all the cells in the network transmit the same physical cell ID, it will be
impossible for a mobile to identify different cells. Cell search and selection will be impossible. Therefore, it is important to
intelligently allocate physical cell IDs to cells so as to allow easy recognition of cells by mobiles.
The following describes the AFP’s automatic planning method for physical cell IDs in an LTE network, which takes into account
interference matrices, neighbour relations (first-order neighbours, first-order neighbours of a common LTE cell, first-order
neighbours of a common GSM or UMTS cell in 3GPP multi-RAT documents and CDMA cell in 3GPP2 multi-RAT documents, and
optionally second-order neighbours), distance between transmitters, and the frequency plan of the network.
• Their PSS ID and SSS ID statuses are both not set to locked,
zone.

The AFP is based on a cost-based function which takes into account the following constraints, in the order of priority:
1. Physical cell ID,
Assigned weight  ID = 0.75
2. PSS ID,
Assigned weight  PSS = 0.02
3. SSS ID,
Assigned weight  SSS = 0.23
4. PCI Mod 6, for single-antenna port DL CRS),

Assigned weight  CRS = 0
5. PCI Mod 30, for UL DMRS sequence groups,

Assigned weight  ULDMRS = 0
6. PCI Mod (number of frequency blocks / 2), for PCFICH resource element groups,
Assigned weight  PCFICH = 0
The sum of the weights assigned to the above constraints is 1.
572
% ID
 ID = ---------------------------------------------------------------------------------------------------------------------------------------------
% ID + % PSS + % SSS + % CRS + % ULDMRS + % PCFICH
% PSS
 PSS = ---------------------------------------------------------------------------------------------------------------------------------------------
% SSS
 SSS = ---------------------------------------------------------------------------------------------------------------------------------------------
% CRS
 CRS = ---------------------------------------------------------------------------------------------------------------------------------------------
% ULDMRS
 ULDMRS = ---------------------------------------------------------------------------------------------------------------------------------------------
% PCFICH
 PCFICH = ---------------------------------------------------------------------------------------------------------------------------------------------
The above constraints are studied between each TBA cell and its related cells. Atoll calculates the cost between each
individual TBA and related cell, and then the overall cost for the TBA cell.
Assigned weight  Neighbour = 0.35
TBA cells which are first-order neighbours of a common cell are also related to each other through that cell. This
relation is also taken into account,
Assigned weight  Inter – Neighbour = 0.15
You can choose to not take into account the physical cell ID collision between neighbours of a common cell by adding
an option in the Atoll.ini file (see the Administrator Manual). If the collision between neighbours of a common cell is
not taken into account, the weight assigned to the direct first-order neighbour relation alone is  Neighbour = 0.5 and
that of the collision between neighbours of a common cell is of course  Inter – Neighbour = 0 .
By adding an option in the Atoll.ini file (see the Administrator Manual), second-order neighbours can also be taken
into account. In this case, the assigned weights are:  Neighbour = 0.25 ,  2nd – Neighbour = 0.10 , and
 Inter – Neighbour = 0.15 .
 Inter – Neighbour applies to the relation between neighbours of a common cell, which can be an LTE cell, a UMTS cell
or a GSM transmitter in 3GPP multi-RAT documents or an LTE or CDMA cell in 3GPP2 multi-RAT documents.
Figure 6.7 on page 573 depicts the different neighbour relations that may exist in LTE.
Figure 6.7: Neighbour Relations for Physical Cell ID Allocation
573
Assigned weight  IM = 0.3
• Cells within the cell’s (or the default) reuse distance, if the check box "Reuse distance" is selected,
Assigned weight  Dis tan ce = 0.2
% Neighbour
 Neighbour = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
% Neighbour + % Inter – Neighbour + % 2nd – Neighbour + % IM + % Dis tan ce
% Inter – Neighbour
 Inter – Neighbour = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
% 2nd – Neighbour
 2nd – Neighbour = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
% IM
 IM = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
% Dis tan ce
 Dis tan ce = -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-

Atoll calculates the constraint violation levels between the TBA cell TXi(ic) and its related cell TXj(jc) as follows:
TX  ic  – TX  jc 
i j ID SSS CRS ULDMRS PCFICH
VL 1 =  ID  p Coll +  SSS  p Penalty +  CRS  p Coll +  ULDMRS  p Coll +  PCFICH  p Coll
TX  ic  – TX  jc 
i j PSS
VL 2 =  PSS  p Coll
Where  ID ,  PSS , and  SSS are the weights assigned to the physical cell ID, PSS ID, and SSS ID constraints.
 TX  ic 
i
TX  jc 
j
ID ID  1 if ID  = ID 
p Coll is the physical cell ID collision probability given by p Coll =  .
 TX i  ic  TX j  jc 
 0 if ID   ID 
 TX i  ic  TX j  jc 
PSS PSS  1 if ID PSS = ID PSS
p Coll is the PSS ID collision probability given by p Coll =  .
 0 if ID PSS  ID PSS
 TX i  ic  TX i  ic  TX j  jc 
 1 if R Co-site  3 AND ID SSS  ID SSS
SSS SSS 
p Penalty is the SSS ID penalty given by p Penalty =  TX  ic 
i
TX  ic 
i
TX  jc 
j
TX  ic 
i
if the SSS ID
 1 if R Co-site  3 AND ID  – ID   R Co-site

 0 Otherwise
SSS
planning strategy is set to "Same per site", and by p Penalty = 0 if the SSS ID planning strategy is set to "Free". The SSS penalty
models the SSS ID allocation constraint.
CRS
p Coll is the single antenna downlink cell-specific reference signal collision probability given by
CRS  1 if ID  Mod6 = ID  Mod6
p Coll =  .
 0 if ID  Mod6  ID Mod6
574
 TX  ic 
i
TX  jc 
j
ULDRMS ULDMRS  1 if ID  Mod30 = ID  Mod30
p Coll is the UL DMRS collision probability given by p Coll =  .
 TX  ic  TX  jc 
i j
 0 if ID  Mod30  ID  Mod30
PCFICH
p Coll is the collision probability of the physical control format indicator channel resource element groups given by
 TX i  ic   N FB  TX j  jc   N FB 
 1 if ID  Mod  ---------------- = ID  Mod  -----------------
PCFICH   2   2 
p Coll =  .
 TX i  ic   N FB  TX j  jc   N FB 
 0 if ID  Mod  ----------------  ID  Mod  ----------------
-
  2   2 
Next, Atoll calculates the importance of the neighbour relations between the TBA cell and its related cell.
 Neighbours =  Neighbour   Neighbour +  Inter – Neighbour   Inter – Neighbour +  2nd – Neighbour   2nd – Neighbour
Where  Neighbour is the importance of the relationship between the TBA cell and its related neighbour cell.  Neighbour
is calculated during automatic neighbour planning by Atoll as explained in "Automatic Neighbour Planning" on page 563. For
manual neighbour planning, this value is equal to 1.
 Inter – Neighbour is calculated from the neighbour relationship importance values calculated during automatic neighbour
planning. If two cells are neighbours of a common cell and have the same physical cell ID assigned, the importance of the
physical cell ID collision is the average of their neighbour importance values with the common neighbour cell. If more than
one pair of neighbours of the TBA cell has the same physical cell ID assigned, then the importance is the highest value among
all the averages:
TX i  ic  – TX j1  j1c  TX i  ic  – TX j2  j2c 
  Neighbour +  Neighbour 
 Inter – Neighbour = Max  ---------------------------------------------------------------------------------
All Neighbour Pairs  2 
with ID Collisions
Where TX j1  j1c  and TX j2  j2c  are two neighbours of the TBA cell TX i  ic  that have the same physical cell ID assigned. The
above applies to intra-technology as well as inter-technology neighbours in 3GPP multi-RAT and 3GPP2 multi-RAT documents.
 2nd – Neighbour is calculated from the neighbour relationship importance values calculated during automatic neighbour
planning. If the TBA cell has the same physical cell ID assigned as one of its second-order neighbours, the importance of the
physical cell ID collision is the multiple of the importance values of the first order neighbour relations between the TBA cell
and its second order neighbour. If the TBA cell is related to its second order neighbour through more than one first order
neighbour, the importance is the highest value among all the multiples:
TX  ic  – TX  jc  TX j  jc  – TX k  kc 
 2nd – Neighbour = Max  i j
  Neighbour 
All Neighbour Pairs
 Neighbour 
with ID Collisions
Where TX k  kc  is the second-order neighbour of TX i  ic  through TX j  jc  .
Next, Atoll calculates the importance of the interference relations between the TBA cell and its related cell.
TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc 
 Interference =  IM   IM +  Dis tan ce   Dis tan ce  f Overlap
 IM = r CCO   IM – CC + r ACO   IM – AC if the frequency plan is taken into account
and  IM =  IM – CC otherwise.
explained in "Interference Matrix Calculation" on page 578. r O , r CCO , and r ACO are the total,
co-channel, and adjacent channel overlap ratios as calculated in "Co- and Adjacent Channel Overlaps Calculation" on
page 493.
575
TX  ic  – TX  jc 
i j
TX  ic  – TX  jc 
i j
TX  ic  – TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc 
i j i j i j
f Overlap = rO if the frequency plan is taken into account and f Overlap = 1 otherwise.
From the constraint violation levels and the importance values of the relations between the TBA and its related cell, Atoll
calculates the quality reduction factor for the pair as follows:
TX  ic  – TX  jc   TX  ic  – TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc  
= 1 –   VL 1   i
i j i j i j j i j i j i j
QRF + VL 2 + VL    f 
 Interference 1 Neighbours Overlap
 
The total cost of the current physical cell ID plan for any TBA cell is given as follows, considering all the cells with which the
TBA cell has relations:
TX j  jc 
And, the total cost of the current physical cell ID plan for the entire network is simply the sum of the total TBA cell costs
calculated above, i.e.,
TX i  ic 
TX i  ic 

• Calculates the cost (as described above) of the current physical cell ID plan,
• Tries different physical cell IDs to cells in order to reduce the costs,
• Memorises the different plans in order to determine the best plan, i.e., which provides the lowest total cost,
• Stops when it is unable to improve the cost of the network, and proposes the last known best physical cell ID plan as
the solution.
6.5.5 Automatic PRACH RSI Planning Using the AFP

The following describes the AFP’s automatic planning method for PRACH RSIs in an LTE network, which takes into account
interference matrices, neighbour relations (first-order neighbours and optionally second-order neighbours), and distance
between transmitters.
• Their PRACH RSI status is not set to locked,
zone.

The AFP is based on a cost-based function which takes into account various relations between each TBA cell and its related
cells. Atoll calculates the cost between each individual TBA and related cell, and then the overall cost for the TBA cell.
Assigned weight  Neighbour = 0.50
By adding an option in the Atoll.ini file (see the Administrator Manual), second-order neighbours can also be taken
into account. In this case, the assigned weights are:  Neighbour = 0.25 and  2nd – Neighbour = 0.10 .
576
Assigned weight  IM = 0.3
• Cells within the cell’s (or the default) reuse distance, if the check box "Reuse distance" is selected,
Assigned weight  Dis tan ce = 0.2
relative weights (%) defined in the Weights dialogue as follows:
% Neighbour
 Neighbour = ----------------------------------------------------------------------------------------------------------------------------------
-
% Neighbour + % 2nd – Neighbour + % IM + % Dis tan ce
% 2nd – Neighbour
 2nd – Neighbour = ----------------------------------------------------------------------------------------------------------------------------------
-
% IM
 IM = ----------------------------------------------------------------------------------------------------------------------------------
-
% Dis tan ce
 Dis tan ce = ----------------------------------------------------------------------------------------------------------------------------------
-

Atoll calculates the constraint violation levels between the TBA cell TXi(ic) and its related cell TXj(jc) as follows:
PRACH
VL = Min  1 p Coll 
PRACH ID N Common PRACH RSIs
p Coll is the PRACH RSI collision probability given by p Coll = ----------------------------------------- .
TX i  ic 
N Req PRACH RSIs
TX  ic  – TX  jc  TX  ic 
i j i
Where N Common PRACH RSIs is the number of PRACH RSIs common between cells TXi(ic) and TXj(jc), and N Req PRACH RSIs is the
number of PRACH RSIs required by the cell TXi(ic).
Next, Atoll calculates the importance of the neighbour relations between the TBA cell and its related cell.
 Neighbours =  Neighbour   Neighbour +  2nd – Neighbour   2nd – Neighbour
Where  Neighbour is the importance of the relationship between the TBA cell and its related neighbour cell.  Neighbour
is calculated during automatic neighbour planning by Atoll as explained in the Technical Reference Guide. For manual
neighbour planning, this value is equal to 1.
 2nd – Neighbour is calculated from the neighbour relationship importance values calculated during automatic neighbour
planning. If the TBA cell has the same PRACH RSI assigned as one of its second-order neighbours, the importance of the PRACH
RSI collision is the multiple of the importance values of the first order neighbour relations between the TBA cell and its second
order neighbour. If the TBA cell is related to its second order neighbour through more than one first order neighbour, the
importance is the highest value among all the multiples:
TX  ic  – TX  jc  TX j  jc  – TX k  kc 
 2nd – Neighbour = Max  i j
  Neighbour 
All Neighbour Pairs
 Neighbour 
with Collisions
Where TX k  kc  is the second-order neighbour of TX i  ic  through TX j  jc  .
Next, Atoll calculates the importance of the interference relations between the TBA cell and its related cell.
TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc 
 Interference =  IM   IM +  Dis tan ce   Dis tan ce  f Overlap
577
TX  ic  – TX  jc 
i j
TX  ic  – TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc 
i j i j i j i j i j
 IM = r CCO   IM – CC + r ACO   IM – AC if the frequency plan is taken into account
i j i j
and  IM =  IM – CC otherwise.
TX  ic  – TX  jc  TX  ic  – TX  jc  TX  ic  – TX  jc 
i j i j i j
explained in "Interference Matrix Calculation" on page 578. r O , r CCO , and r ACO are the total,
co-channel, and adjacent channel overlap ratios calculated as calculated in "Co- and Adjacent Channel Overlaps Calculation"
on page 493.
f Overlap = rO if the frequency plan is taken into account and f Overlap = 1 otherwise.
From the constraint violation levels and the importance values of the relations between the TBA and its related cell, Atoll
calculates the quality reduction factor for the pair as follows:
QRF = 1 – VL    Interference +  Neighbours  f Overlap 
 
The total cost of the current PRACH RSI plan for any TBA cell is given as follows, considering all the cells with which the TBA
cell has relations:
TX j  jc 
And, the total cost of the current PRACH RSI plan for the entire network is simply the sum of the total TBA cell costs calculated
above, i.e.,
TX i  ic 
TX  ic 
i

• Calculates the cost (as described above) of the current PRACH RSI plan,
• Tries different PRACH RSIs to cells in order to reduce the costs,
• Memorises the different plans in order to determine the best plan, i.e., which provides the lowest total cost,
• Stops when it is unable to improve the cost of the network, and proposes the last known best PRACH RSI plan as the
solution.
6.5.6 Appendices
6.5.6.1 Interference Matrix Calculation
The co-channel interference probability is calculated as follows:
S TX  ic 
i
 
 C Max + M Quality n
Sym 
TX i  ic   ----------------------------------------------------- --------------------- TX  ic   TX  ic  TX i  ic 
10 10 i   T i
– 10  Log  10 2N
Sym 
C + 10 –n
DLRS  FB   RSRP
 
TX i  ic  – TX j  jc   
 
 IM – CC = ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
S TX  ic 
i
The adjacent channel interference probability is calculated as follows:
578
S TX  ic 
i
TX  jc  TX  ic  TX  ic 
 j i i 
 C Max + M Quality + f ACS n
Sym 
TX  ic   ------------------------------------------------------------------------------
- --------------------
- TX  ic   TX  ic  TX  ic 
i  10 10 i   T i i
10  Log 10 2N
DLRS –   RSRP – n Sym 
C + 10
 FB
 
TX  ic  – TX  jc   
i j  
 IM – AC = ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
S TX  ic 
i
For frequencies farther than the adjacent channel, the interference probability is 0.
Here S TX  ic  is the best server coverage area of the cell TXi(ic), that comprises all the pixels where E DLRS  T RSRP as
i
calculated in "Best Server Determination" on page 535. S TX  ic  is the best server coverage area of the cell TXi(ic)
i Condition
where the given condition is true. C DLRS is the received downlink reference signal level from the cell TXi(ic). C Max is the
TX i  ic 
received maximum signal level from the cell TXj(jc) calculated using the Max Power defined for this cell. n Sym is the per-
TX  ic 
i
subcarrier noise for the cell TXi(ic) as calculated in "Noise Calculation (DL)" on page 505 and N FB is the total number of
frequency blocks defined in the frequency bands table for the channel bandwidth used by the cell. M Quality is the quality
TX i  ic 
margin used for the interference matrices calculation. And, f ACS is the adjacent channel suppression factor defined for the
frequency band of the cell TXi(ic).
6.5.6.2 Distance Importance Calculation

TX  ic  – TX  jc 
i j
The distance importance between two cells (  Dis tan ce ) is calculated as follows:
 TX  ic  – TX  jc 
 1 if D
i j
1

TX i  ic  – TX j  jc    D Reuse  2
 Dis tan ce =  Log   ---------------------------------  
  D  
 --------------------------------------------------------- Otherwise
 Log  D Reuse 
2

Where D Reuse is the minimum reuse distance, either defined for each TBA cell individually or set for all the TBA cells in the
AFP dialogue, and D is the weighted distance between the TBA cell TXi(ic) and its related cell TXj(jc) calculated as
follows:
D = d   1 + x   cos    – cos    – 2  
D is weighted according to the azimuths of the TBA cell and its related cell with respect to the straight line joining
them. d is the distance between the two cells considering any offsets with respect to the site locations. x is set
to 10 % so that the maximum variation in D due to the azimuths does not exceed 40 %.  and  are calculated
from the azimuths of the two cells as shown in Figure 6.8 on page 579.
Figure 6.8: Weighted Distance Between Cells
The above formula implies that two cells facing each other will have a shorter effective distance between them than the real
distance, and two cells pointing in opposite directions will have a greater effective distance.
579
The importance of the distance relation is explained in Figure 6.9 on page 580. This figure shows that cells that are located
near (based on the effective distance which is weighted by the orientations of the cells) have high importance, which is
interpreted as a high cost, and cells that are located far have low importance. Cells that are further than the reuse distance
do not have any cost related to the distance relation.
Figure 6.9: Importance Based on Distance Relation
580
Chapter 7
3GPP Multi-RAT
Networks
• "Multi-RAT Monte Carlo Simulations" on page 583
• "Multi-RAT Coverage Predictions" on page 585
© Forsk 2015
582
AT330_TRR_E1 Chapter 7: 3GPP Multi-RAT Networks
7 3GPP Multi-RAT Networks

This chapter describes the calculations specific to 3GPP multi-RAT documents. Multi-RAT calculations that are the same as
those in single-RAT documents can be found in:
• "GSM GPRS EDGE Networks" on page 125,
• "UMTS HSPA Networks" on page 211, and
• "LTE Networks" on page 445.
7.1 Definitions
This table lists the input to coverage prediction and simulation calculations.

UL
f act Service parameter None Uplink activity factor
DL
f act Service parameter None Downlink activity factor
TL DL – GSM Subcell parameter % Downlink traffic load (GSM)
7.2 Multi-RAT Monte Carlo Simulations

7.2.1 User Distribution

• "Simulations Based on User Profile Traffic Maps" on page 584.
[Simulation]
RandomTotalUsers=0
In 3GPP multi-RAT documents, services can be classified under constant bit rate and variable bit rate services, which can be
provided by one or more technology. These service categories comprise the following service types in different technologies:
Constant Bit Rate Services Variable Bit Rate Services
Circuit
GSM GPRS EDGE Packet (Max Bit Rate)
Packet (Constant Bit Rate)
Packet R99
Circuit R99
UMTS HSPA Packet HSDPA (Best Effort)
Packet HSPA (Constant Bit Rate)
Packet HSPA (Best Effort)
LTE Voice Data
583
Chapter 7: 3GPP Multi-RAT Networks ©Forsk 2015
Simulations Based on User Profile Traffic Maps
points.
service usage characteristics described in the user profiles, i.e., the number of service sessions, the average duration of each
constant bit rate service session, or the volume of the data transfer in the uplink and the downlink in each variable bit rate
service session as explained in:
• GSM and LTE: "Simulations Based on User Profile Traffic Maps and Subscriber Lists" on page 476
• UMTS: "Simulations Based on User Profile Traffic Maps" on page 225.
For any variable bit rate service (j), once several numbers of users with different activity statuses have been calculated
for different technologies, the final numbers of users are obtains as follows:
inactive inactive inactive inactive

Number of inactive users: n j = Average  n j  nj  nj 
GSM UMTS LTE
Number of users active on UL: n j  UL  = Average  n j  UL   n j  UL   n j  UL  

GSM UMTS LTE
Number of users active on DL: n j  DL  = Average  n j  DL   n j  DL   n j  DL  

GSM UMTS LTE
Number of users active on UL+DL: n j  UL + DL  = Average  n j  UL + DL   n j  UL + DL   n j  UL + DL  

GSM UMTS LTE
Simulations Based on Sector Traffic Maps
Atoll calculates the number of active users of each service UL and DL as follows:
• GSM and LTE: "Simulations Based on Sector Traffic Maps" on page 478
• UMTS: "Simulations Based on Sector Traffic Maps" on page 229.
Distribution of Terminals
Terminals assigned to users depend on the percentages defined per traffic map and the technologies supported by each
terminal. For example, if the percentages of terminals are defined as follows:
Terminal A (GSM): 30 %
Terminal B (GSM+UMTS): 50 %
Terminal C (GSM+UMTS+LTE): 20 %
For users of services that can be provided by GSM, UMTS, or LTE:
Terminal A: 30 %
Terminal B: 50 %
584
AT330_TRR_E1 Chapter 7: 3GPP Multi-RAT Networks
Terminal C: 20 %
For users of services that can be provided by UMTS or LTE:
Terminal B: 50/70 = 71.4 %
Terminal C: 20/70 = 28.6 %
For users of services that can be provided by LTE only, Terminal C will be assigned.
7.2.2 Simulation Process

Each Monte Carlo simulation is a snap-shot of the network where resource allocation is carried out. The steps of this algorithm
are listed below.
• Serving cell/technology selection
For each mobile, Atoll searches for a serving cell of each supported and available technology as follows:
• GSM: Atoll determines a best server based on the HCS layer/server selection algorithm. If no best server can be
found, the mobile will be considered rejected by GSM.
• UMTS: Atoll determines a best server based on Ec/Io. If no best server can be found, the mobile will be considered
rejected by UMTS.
• LTE: Atoll determines the best server based on RSRP or RS level and the serving cell selection method. If no best
server can be found, the mobile will be considered rejected by LTE.
Once the potential serving technologies have been identified, Atoll selects the highest priority as defined in the service
assigned to each mobile. The best server and technology assigned to each mobile remains unchanged for the rest of
the simulation.
• Technology-wise Monte Carlo simulations as explained in:
• GSM: "Radio Resource Management in GSM" on page 184
• UMTS: "Power Control Simulation" on page 231
• LTE: "Scheduling and Radio Resource Management" on page 552
7.3 Multi-RAT Coverage Predictions

Coverage predictions are calculated by determining the best server for each technology on each pixel and then determining
the selected display parameter within the best server’s calculation area. Each pixel within the calculation area is considered
a non-interfering receiver. The properties of the non-interfering probe receiver are set by selecting a terminal, a mobility type,
and a service.
The resolutions of coverage predictions do not depend on the resolutions of the path loss matrices or the geographic data and
can be defined separately for each coverage prediction. Coverage predictions are generated using a bilinear interpolation
3GPP multi-RAT coverage predictions are combinations of corresponding single-RAT coverage predictions with specific
parameter settings. Multi-RAT coverage predictions include:
• Effective Service Area Analysis (DL+UL)
Effective Service Area Analysis (DL+UL)
The 3GPP multi-RAT effective service area is the combination of single-RAT effective service areas:
• GSM Service Area Analysis (DL) is based on a coverage by coding scheme, as explained in "GPRS/EDGE Coverage Pre-
dictions" on page 140 or on a coverage by codec modes, as explained in "Circuit Quality Indicators Coverage Predic-
tions" on page 149, depending on the type of service.
Radio conditions are evaluated over the HCS server area with a margin of 4 dB, on all the interfered subcells. Codec
modes and coding schemes are obtained from these radio conditions based on C/I+N without ideal link adaptation (as
explained in "Throughput Calculation Based on Interpolation Between C/N and C/(I+N)" on page 138). This implies
that a frequency plan has to be defined in order to obtain this GSM/GPRS/EDGE coverage.
• UMTS Effective Service Area Analysis (Eb⁄Nt) (DL+UL) is based on a combination of downlink and uplink service area
predictions, as explained in "Downlink Service Area Analysis" on page 300 and "Uplink Service Area Analysis" on
page 302. In the case of HSPA services, the coverage is based on a combination of HSDPA et HSUPA service areas as
explained in "HSDPA Prediction Study" on page 304 and "HSUPA Prediction Study" on page 309.
• LTE Effective Service Area Analysis (DL+UL) is based on a combination of downlink and uplink service area predictions,
as explained in "Effective Signal Analysis Coverage Predictions" on page 472.
Two display options are available for this prediction:
585
Chapter 7: 3GPP Multi-RAT Networks ©Forsk 2015
• Technologies: Each pixel displays the colour representing the visible technology having the highest priority defined in
the selected service.
• Available Technologies: Pixels display the colour representing the combined areas over which a multi-technology ter-
minal can be served. For instance, the GSM+UMTS area shows the union between the GSM and the UMTS service
areas as explained above.
Coverage by Throughput (DL)
The 3GPP multi-RAT throughput prediction is the combination of single-RAT throughput predictions:
• GSM Packet Throughput Analysis (DL) explained in "Application Throughput Calculation" on page 139
The 3GPP multi-RAT effective RLC throughput is obtained from the maximum effective RLC throughput of the GSM
layer. The 3GPP multi-RAT application throughput from the maximum application throughput of the GSM layer.
• R99 Service Area Analysis (Eb⁄Nt) (DL) explained in "Downlink Service Area Analysis" on page 300 and HSDPA
Throughput Analysis (DL) explained in "HSDPA Prediction Study" on page 304
R99: The 3GPP multi-RAT effective RLC and application throughputs are respectively obtained from the effective RLC
and application throughputs of the R99 layer (see "Downlink Service Area Analysis" on page 300 for more
information).
HSDPA: The 3GPP multi-RAT effective RLC and application throughputs are respectively obtained from the effective
RLC and application throughputs of the HSDPA layer (see "HSDPA Prediction Study" on page 304 for more
information).
• LTE Coverage by Throughput (DL) explained in "C/(I+N)-based Coverage Predictions" on page 473
The 3GPP multi-RAT effective RLC and application throughputs are respectively obtained from the Effective RLC
Channel Throughput (DL) and the Application Channel Throughput (DL) (see "C/(I+N)-based Coverage Predictions" on
page 473 for more information).
Four display options are available for this prediction:
• Effective RLC Throughput: The throughput on the RLC layer that a cell can provide to the selected terminal per pixel
taking into account possible transmission errors (BLER) for the highest priority technology.
• Max Effective RLC Throughput: The maximum throughput on the RLC layer that a cell can provide to the selected
terminal per pixel taking into account possible transmission errors (BLER), considering all available technologies.
• Application Throughput: The throughput on the application layer that a cell can provide to the selected terminal per
pixel taking into account possible transmission errors (BLER) for the highest priority technology.
• Max Application Throughput: the maximum throughput on the application layer that a cell can provide to the selected
586
Chapter 8
3GPP2 Multi-RAT
Networks
• "Multi-RAT Monte Carlo Simulations" on page 589
• "Multi-RAT Coverage Predictions" on page 591
© Forsk 2015
588
AT330_TRR_E1 Chapter 8: 3GPP2 Multi-RAT Networks
8 3GPP2 Multi-RAT Networks

This chapter describes the calculations specific to 3GPP2 multi-RAT documents. Multi-RAT calculations that are the same as
those in single-RAT documents can be found in:
• "CDMA2000 Networks" on page 337, and
• "LTE Networks" on page 445
8.1 Definitions
This table lists the input to coverage prediction and simulation calculations.

UL
f act Service parameter None Uplink activity factor
DL
f act Service parameter None Downlink activity factor
8.2 Multi-RAT Monte Carlo Simulations

8.2.1 User Distribution

[Simulation]
RandomTotalUsers=0
In 3GPP2 multi-RAT documents, services can be classified under constant bit rate and variable bit rate services, which can be
provided by one or more technology. These service categories comprise the following service types in different technologies:
Constant Bit Rate Services Variable Bit Rate Services

Speech
1xRTT Data
1xEV-DO rev. A (Best Effort)
CDMA 1xEV-DO rev. 0
1xEV-DO rev. B (Best Effort)
1xEV-DO rev. A (Guaranteed Bit Rate)
1xEV-DO rev. B (Guaranteed Bit Rate)
LTE Voice Data
Simulations Based on User Profile Traffic Maps
589
Chapter 8: 3GPP2 Multi-RAT Networks ©Forsk 2015
points.
service usage characteristics described in the user profiles, i.e., the number of service sessions, the average duration of each
constant bit rate service session, or the volume of the data transfer in the uplink and the downlink in each variable bit rate
service session as explained in:
• LTE: "Simulations Based on User Profile Traffic Maps and Subscriber Lists" on page 476
• CDMA: "Simulations Based on User Profile Traffic Maps" on page 359.
For any variable bit rate service (j), once several numbers of users with different activity statuses have been calculated
for different technologies, the final numbers of users are obtains as follows:
inactive inactive inactive
Number of inactive users: n j = Average  n j  nj 
CDMA LTE
Number of users active on UL: n j  UL  = Average  n j  UL   n j  UL  

CDMA LTE
Number of users active on DL: n j  DL  = Average  n j  DL   n j  DL  

CDMA LTE
Number of users active on UL+DL: n j  UL + DL  = Average  n j  UL + DL   n j  UL + DL  

CDMA LTE
Simulations Based on Sector Traffic Maps
Atoll calculates the number of active users of each service UL and DL as follows:
• LTE: "Simulations Based on Sector Traffic Maps" on page 478
• CDMA: "Simulations Based on Sector Traffic Maps" on page 362.
8.2.2 Simulation Process

Each Monte Carlo simulation is a snap-shot of the network where resource allocation is carried out. The steps of this algorithm
are listed below.
• Serving cell/technology selection
For each mobile, Atoll searches for a serving cell of each supported and available technology as follows:
• CDMA: Atoll determines a best server based on Ec/Io. If no best server can be found, the mobile will be considered
rejected by CDMA.
• LTE: Atoll determines the best server based on RSRP or RS level and the serving cell selection method. If no best
server can be found, the mobile will be considered rejected by LTE.
Once the potential serving technologies have been identified, Atoll selects the highest priority as defined in the service
assigned to each mobile. The best server and technology assigned to each mobile remains unchanged for the rest of
the simulation.
• Technology-wise Monte Carlo simulations as explained in:
• CDMA: "Simulations" on page 358
• LTE: "Scheduling and Radio Resource Management" on page 552
590
AT330_TRR_E1 Chapter 8: 3GPP2 Multi-RAT Networks
8.3 Multi-RAT Coverage Predictions

Coverage predictions are calculated by determining the best server for each technology on each pixel and then determining
the selected display parameter within the best server’s calculation area. Each pixel within the calculation area is considered
a non-interfering receiver. The properties of the non-interfering probe receiver are set by selecting a terminal, a mobility type,
and a service.
The resolutions of coverage predictions do not depend on the resolutions of the path loss matrices or the geographic data and
can be defined separately for each coverage prediction. Coverage predictions are generated using a bilinear interpolation
3GPP2 multi-RAT coverage predictions are combinations of corresponding single-RAT coverage predictions with specific
parameter settings. Multi-RAT coverage predictions include:
• Effective Service Area Analysis (DL+UL)
Effective Service Area Analysis (DL+UL)
The 3GPP2 multi-RAT effective service area is the combination of single-RAT effective service areas:
• CDMA Effective Service Area Analysis (Eb⁄Nt) (DL+UL) is based on a combination of downlink and uplink service area
predictions, as explained in "Downlink Service Area Analysis" on page 413 and "Uplink Service Area Analysis" on
page 417.
• LTE Effective Service Area Analysis (DL+UL) is based on a combination of downlink and uplink service area predictions,
as explained in "Effective Signal Analysis Coverage Predictions" on page 472.
Two display options are available for this prediction:
• Technologies: Each pixel displays the colour representing the visible technology having the highest priority defined in
the selected service.
• Available Technologies: Pixels display the colour representing the combined areas over which a multi-technology ter-
minal can be served. For instance, the CDMA+LTE area shows the union between the CDMA and the LTE service areas
as explained above.
Coverage by Throughput (DL)
The 3GPP2 multi-RAT throughput prediction is the combination of single-RAT throughput predictions:
• Service Area Analysis (Eb⁄Nt) (DL) explained in "Downlink Service Area Analysis" on page 413
The 3GPP2 multi-RAT effective RLC and application throughputs are respectively obtained from the effective RLC and
application throughputs (see "Downlink Service Area Analysis" on page 300 for more information).
• LTE Coverage by Throughput (DL) explained in "C/(I+N)-based Coverage Predictions" on page 473
The 3GPP multi-RAT effective RLC and application throughputs are respectively obtained from the Effective RLC
Channel Throughput (DL) and the Application Channel Throughput (DL) (see "C/(I+N)-based Coverage Predictions" on
page 473 for more information).
Four display options are available for this prediction:
• Effective RLC Throughput: The throughput on the RLC layer that a cell can provide to the selected terminal per pixel
taking into account possible transmission errors (BLER) for the highest priority technology.
• Max Effective RLC Throughput: The maximum throughput on the RLC layer that a cell can provide to the selected
• Application Throughput: The throughput on the application layer that a cell can provide to the selected terminal per
pixel taking into account possible transmission errors (BLER) for the highest priority technology.
• Max Application Throughput: the maximum throughput on the application layer that a cell can provide to the selected
591
Chapter 8: 3GPP2 Multi-RAT Networks ©Forsk 2015
592
Chapter 9
TD-SCDMA Networks

• "Definitions and Formulas" on page 595
• "Signal Level Based Calculations" on page 602
• "Monte Carlo Simulations" on page 608
• "TD-SCDMA Prediction Studies" on page 626
• "Smart Antenna Modelling" on page 638
• "N-Frequency Mode and Carrier Allocation" on page 650
• "Neighbour Allocation" on page 651
• "Scrambling Code Allocation" on page 656
• "Automatic GSM/TD-SCDMA Neighbour Allocation" on
page 666
© Forsk 2015
594
AT330_TRR_E1 Chapter 9: TD-SCDMA Networks
9 TD-SCDMA Networks
This chapter describes in detail the algorithms, calculation parameters, and processes of the coverage predictions and the
simulations available in TD-SCDMA documents. The first part of this chapter lists all the input and output parameters in the
TD-SCDMA documents, their significance, location in the Atoll GUI, and their usage. Detailed explanation of the basic coverage
predictions, which do not require simulation results, is provided in the second part. The third part describes the traffic
scenario generation and Montel Carlo simulation algorithms including smart antenna modelling and dynamic channel
allocation. The next sections are dedicated to TD-SCDMA coverage predictions which can be based on results obtained from
simulations. The last three sections describe in detail the allocation of frequencies, i.e., master and slave carriers, the
allocation of neigbours, and the allocation of scrambling codes.
9.1 Definitions and Formulas

The tables in the following subsections list the input and output parameters and formulas used in simulations and other
computations.
9.1.1 Inputs
This table lists the inputs to computations, coverage predictions, and simulations.
R Ch Global parameter Mcps Chip rate (or Spreading rate) (1.28)

Spread
F Min Global parameter None Minimum spreading factor (1)
Spread
F Max Global parameter None Maximum spreading factor (16)
Proc
G P – CCPCH Global parameter None P-CCPCH processing gain (13.8 dB)
SF Number of timeslots per subframe

N TS Global parameter None
(7)
SF Global parameter ms Subframe duration (5)
D
Frame Global parameter ms Frame duration (10)
D
GP Number of guard period chips per
N Ch  TS Global parameter None
timeslot (16)
Data Number of data chips per timeslot
(704)
Midamble Number of midamble chips per
timeslot (144)
N Ch  PTS Global parameter None
pilot timeslot (96)
N Ch  DwPTS Global parameter None
DwPTS timeslot (32)
SYNC_DL Number of SYNC_DL chips per
N Ch  DwPTS Global parameter None
DwPTS timeslot (64)
Global parameter Total number of chips per DwPTS
Total
N Ch  DwPTS Total GP SYNC_DL None
N Ch  DwPTS = N Ch  DwPTS + N Ch  DwPTS timeslot (96)

N Ch  UpPTS Global parameter None
UpPTS timeslot (32)
SYNC_UL Number of SYNC_UL chips per UpPTS
N Ch  UpPTS Global parameter None
timeslot (128)
Global parameter Total number of chips per UpPTS
Total
N Ch  UpPTS Total GP SYNC_UL None
N Ch  UpPTS = N Ch  UpPTS + N Ch  UpPTS timeslot (160)
595
Chapter 9: TD-SCDMA Networks ©Forsk 2015

Calculated global parameter
Data
W N Ch  TS bps Chip rate (140800 bps)
W = ---------------
-
SF
D
Average frequency range of the
F Avg Frequency band parameter MHz
frequency band (2010)
Channel bandwidth of the carriers of
BW Frequency band parameter MHz
a frequency band (1.6)
F IRF Cell parameter None Interference reduction factor
TX
F JD Site equipment parameter None Joint Detection (JD) factor
TX
F MCJD Site equipment parameter None Multi-Cell Joint Detection factor
Transmitter parameter
TX (user-defined or calculated from transmitter equipment None BTS Noise Figure
NF
characteristics)
Transmitter loss
Transmitter parameter
TX (user-defined or calculated from transmitter equipment None L Tx = L Total – UL on uplink
L
characteristics) L Tx = L Total – DL on downlink
TX Percentage of BTS signal correctly

 BTS parameter None
transmitted
Max
P TCH Cell parameter W Maximum cell traffic timeslot power
P P – CCPCH Cell parameter W P-CCPCH power on TS0
P DwPCH Cell parameter W DwPCH power on DwPTS

Other common channel power on
P OCCH – TS0 Cell parameter W
TS0
P-CCPCH RSCP comparative
TComp P – CCPCH Cell parameter None
threshold for baton handover
Maximum difference between two
P Max Cell parameter None
transmitted powers
Req
RU UL Cell parameter None Required resource units in uplink
Req
RU DL Cell parameter None Required resource units in downlink
Available HS-PDSCH power available per

P HS – PDSCH Cell parameter W
downlink timeslot
P HR Cell parameter None Power headroom
HS-SCCH power per downlink
P HS – SCCH Cell parameter W
timeslot
N HS – SCCH Cell parameter None Number of HS-SCCH channels
N HS – SICH Cell parameter None Number of HS-SICH channels

Max
N HSDPA Cell parameter None Maximum number of HSDPA users
Min Minimum number of HS-PDSCH

N HS-PDSCH Codes Cell parameter None
codes
Max Maximum number of HS-PDSCH
N HS-PDSCH Codes Cell parameter None
codes
Max Maximum number of intra-
N Intra – Neigh Cell parameter None
technology neighbours
Max Maximum number of inter-
N Intra – Neigh Cell parameter None
technology neighbours
596

The minimum P-CCPCH RSCP
Min
RSCP P – CCPCH Cell parameter or Global parameter W required for a user to be connected
to the cell
P OCCH Timeslot parameter W Other common channel power
DL
P TCH Timeslot parameter W Downlink traffic power
Max Timeslot parameter Maximum percentage of downlink

%P DL None
(Simulation constraint) used power
Timeslot parameter
X UL None Uplink load factor
(Simulation result)
Max Timeslot parameter
X UL None Maximum uplink load factor
(Simulation constraint)
Available
P HS – PDSCH Timeslot parameter W HS-PDSCH power available
Min Minimum number of HS-PDSCH

N HS-PDSCH Codes Timeslot parameter None
codes
Max Maximum number of HS-PDSCH
N HS-PDSCH Codes Timeslot parameter None
codes
RU Overhead Timeslot parameter Overhead resource units
Body Service parameter None Body loss
L
Downlink activity factor for circuit-
Act
f DL Service parameter None switched services and the A-DPCH
activity factor for HSDPA services
Uplink activity factor for circuit-
Act
f UL Service parameter None switched services and the A-DPCH
activity factor for HSDPA services
Eff Downlink efficiency factor for circuit-
f DL Service parameter None
switched services
Eff Uplink efficiency factor for circuit-
f UL Service parameter None
switched services
F Scaling Service parameter None Application througput scaling factor
O TP Service parameter kbps Application throughput offset
UL Average number of packet calls on

N PacketCall Service parameter (packet session modelling) None
the uplink during a session
DL Average number of packet calls on
N PacketCall Service parameter (packet session modelling) None
the downlink during a session
UL Average time between two packet
T PacketCall Service parameter (packet session modelling) ms
calls on the uplink
DL Average time between two packet
T PacketCall Service parameter (packet session modelling) ms
calls on the downlink
UL Minimum packet call size on the
S Min – PacketCall Service parameter (packet session modelling) KBytes
uplink
DL Minimum packet call size on the
S Min – PacketCall Service parameter (packet session modelling) KBytes
downlink
UL Maximum packet call size on the
S Max – PacketCall Service parameter (packet session modelling) KBytes
uplink
DL Maximum packet call size on the
S Max – PacketCall Service parameter (packet session modelling) KBytes
downlink
UL Average time between two packets
T Packet Service parameter (packet session modelling) ms
on the uplink
DL Average time between two packets
T Packet Service parameter (packet session modelling) ms
on the downlink
597

UL
S Packet Service parameter (packet session modelling) Bytes Packet size on uplink
DL
S Packet Service parameter (packet session modelling) Bytes Packet size on downlink
Nom
R DL R99 bearer parameter kbps Downlink peak throughput
Nom
R UL R99 bearer parameter kbps Uplink peak throughput
W-
G DL
Proc R99 bearer parameter (Can be calculated as -----------
Nom
) None Downlink processing gain
R DL
W-
G UL
Proc R99 bearer parameter (Can be calculated as -----------
Nom
) None Uplink processing gain
R UL
Min Allowed minimum downlink traffic

P TCH – DL R99 bearer parameter W
channel power
Max Allowed maximum downlink traffic
P TCH – DL R99 bearer parameter W
channel power
TS
N DL R99 bearer parameter None Number of downlink timelots
TS
N UL R99 bearer parameter None Number of uplink timelots
E Req C Req
Req
Q TCH – UL R99 bearer parameter per mobility (  ----b- or  --- ) None Eb/Nt or C/I target on uplink
 N t TCH – UL  I  TCH – UL
E Req C Req
Req
Q TCH – DL R99 bearer parameter per mobility (  ----b- or  --- ) None Eb/Nt or C/I target on downlink
 N t TCH – DL  I  TCH – DL
Req
RSCP TCH – UL R99 bearer parameter per mobility W Target RSCP on uplink TCH
Req
RSCP TCH – DL R99 bearer parameter per mobility W Target RSCP on downlink TCH
Div
G DL R99 bearer parameter per mobility None Downlink diversity gain
Div
G UL R99 bearer parameter per mobility None Uplink diversity gain
Term
P Max Terminal parameter W Maximum terminal power
Term
P Min Terminal parameter W Minimum terminal power
P UpPCH Terminal parameter W UpPCH power

Term Terminal parameter None Terminal Noise Figure
NF
Term
F JD Terminal parameter None Joint Detection (JD) factor
Term Percentage of terminal signal

 Terminal parameter None
Term
G Terminal parameter None Terminal gain
Term Terminal parameter None Terminal loss

L
TAdd P – CCPCH Mobility parameter W Required RSCP T_Add for P-CCPCH
TDrop P – CCPCH Mobility parameter W Required RSCP T_Drop for P-CCPCH

Req
RSCP DwPCH Mobility parameter W Required RSCP threshold for DwPCH
Req
RSCP UpPCH Mobility parameter W Required RSCP threshold for UpPCH
E Req C Req
Mobility parameter (  ----b- or  ---
Req Required quality threshold for P-
Q P – CCPCH ) None
 N t P – CCPCH  I  P – CCPCH CCPCH
598
E Req
Mobility parameter (  ----c-
Req Required quality threshold for HS-
Q HS – SCCH ) None
 N t HS – SCCH SCCH
E P – CCPCH
Mobility parameter (  ----c-
Req Required quality threshold for P-
Q HS – SICH ) None
 N t HS – SICH CCPCH
C Req
Mobility parameter (  ---
Req Required quality threshold for
Q DwPCH ) None
 I  DwPCH DwPCH
Model
 Clutter class parameter None Model standard deviation
Eb/Nt
 P – CCPCH or P-CCPCH Eb/Nt or C/I standard
Clutter class parameter None
CI deviation
 P – CCPCH
Eb/Nt CI Downlink Eb/Nt or C/I standard

 DL or  DL Clutter class parameter None
deviation
Eb/Nt CI Uplink Eb/Nt or C/I standard
 UL or  UL Clutter class parameter None
deviation
L Indoor Clutter (and, optionally, frequency band) parameter None Indoor loss
Ortho
F DL Clutter class parameter None Downlink orthogonality factor
Ortho
F UL Clutter class parameter None Uplink orthogonality factor
 Spread Clutter class parameter ° Spreading angle
K 1.38 x 10-23 J/K Boltzman constant

N0
TX
NFTX  K  T  BW W Thermal noise at transmitter
N0
Term
NF Term  K  T  BW W Thermal noise at terminal
TX Antenna parameter None Transmitter antenna gain

G
L Path Propagation model result None Path loss
Model Result calculated from cell edge coverage probability and model Model shadowing margin used in
M Shadowing None
standard deviation coverage predictions
P – CCPCH Result calculated from cell edge coverage probability and P-CCPCH P-CCPCH Eb/Nt shadowing margin
M Shadowing None
Eb/Nt standard deviation used in coverage predictions
 Eb  Nt  DL Result calculated from cell edge coverage probability and DL Eb/Nt DL Eb/Nt shadowing margin used in
M Shadowing None
 Eb  Nt  UL Result calculated from cell edge coverage probability and UL Eb/Nt UL Eb/Nt shadowing margin used in
M Shadowing None
599

For RSCP calculation Transmitter-terminal total loss in
TX Term Body Model coverage predictions
Model L Path  L  L L  L Indoor  M Shadowing
LT = ----------------------------------------------------------------------------------------------------------------------
-
TX Term
G G In UL, only carrier power is
 Eb  Nt  UL
For P-CCPCH Eb/Nt calculation attenuated by M Shadowing .
TX Term Body P – CCPCH
 Eb  Nt  P L Path  L  L L  L Indoor  M Shadowing
LT = ----------------------------------------------------------------------------------------------------------------------
-
TX Term In DL, carrier power and intra-cell
G G
LT None interference are attenuated by
For DL Eb/Nt calculation  Eb  Nt  DL
P – CCPCH
TX Term Body  Eb  Nt  DL M Shadowing or M Shadowing while
 Eb  Nt  DL L Path  L  L L  L Indoor  M Shadowing
LT = ----------------------------------------------------------------------------------------------------------------------
- extra-cell interference is not.
TX Term
G G
 Eb  Nt 
For UL Eb/Nt calculation Therefore, M Shadowing or
DL
TX Term Body  Eb  Nt UL

 Eb  Nt  UL L Path  L  L L  L Indoor  M Shadowing P – CCPCH
M Shadowing are set to 1 in DL extra-
LT = ----------------------------------------------------------------------------------------------------------------------
-
TX Term
G G cell interference calculation.
9.1.2 P-CCPCH Eb/Nt and C/I Calculation

TX TX  ic 
i i
E b TX i  ic 
 ----
  RSCP P – CCPCH
- ------------------------------------------------  G Proc
P – CCPCH None P-CCPCH Eb/Nt for the cell TX i  ic 
 N t P – CCPCH TX i  ic 
N Tot – DL
TX i TX i  ic 
TX i  ic    RSCP P – CCPCH
C
--- ------------------------------------------------ None P-CCPCH C/I for the cell TX i  ic 
 I  P – CCPCH TX i  ic 
N Tot – DL
TX i  ic  TX i  ic  TX i  ic  Downlink total noise for the cell

Term W
N Tot – DL I Intra – DL + I Extra – DL + I IC – DL  ic jc  + N 0 TX i  ic 
TX  ic  TX TX  ic  TX
i i i i
RSCP P – CCPCH     + RSCP OCCH – TS0  
TX i TX i
TX i  ic  With  =    1 – F DL
Ortho
   1 – F JD
Term
 and Downlink intra-cell interference for
W
I Intra – DL the cell TX i  ic 

 =  0 Without Useful Signal
 1 Total Noise
TX j  ic  TX j  ic  Downlink extra-cell interference for

  RSCPP – CCPCH + RSCPOCCH – TS0
TX i  ic 
I Extra – DL W
the cell TX i  ic 
ji
I IC – DL  ic jc    RSCPP – CCPCH + RSCPOCCH – TS0 W Inter-carrier interference

TX
---------------------------------------------------------------------------------------
j -
F IRF  ic jc 
9.1.3 DwPCH C/I Calculation

TX i  ic    RSCP DwPCH
C
--- ------------------------------------------- None DwPCH C/I for the cell TX i  ic 
 I  DwPCH TX i  ic 
N Tot – DL

Term W
600

TX  ic  TX
i i
RSCP DwPCH    
TX TX
TX  ic  With 
i
= 
i
  1 – F DL
Ortho
   1 – F JD
Term
 and Intra-cell interference for the cell
i W
I Intra – DL TX i  ic 

 =  0 Without Useful Signal
 1 Total Noise
TX j  ic  Extra-cell interference for the cell

  RSCPDwPCH
TX i  ic 
W
I Extra – DL TX i  ic 
ji
TX j  jc 
I IC – DL  ic jc    RSCPDwPCH W Inter-carrier interference

TX
----------------------------------------
j -
9.1.4 DL TCH Eb/Nt and C/I Calculation

TX i  ic    RSCP TCH – DL Downlink TCH Eb/Nt for the cell
E b
 ----
- ----------------------------------------------  G Proc Div
DL  G DL None
 N t TCH – DL TX i  ic  TX i  ic 
N Tot – DL
TX i  ic    RSCP TCH – DL
C
--- ----------------------------------------------  G Div None Downlink TCH C/I for the cell TX i  ic 
 I  TCH – DL TX i  ic  DL
N Tot – DL

Term W
 TXi Ortho Term 

TX i 

    1 – F DL    1 – F JD  +  1 –    Downlink intra-cell interference for
TX i  ic    W
I Intra – DL the cell TX i  ic 
  RSCP TCH – DL + RSCP OCCH 
i i
 
TX j  ic  TX j  ic  Downlink extra-cell interference for

  RSCPTCH – DL + RSCPOCCH 
TX i  ic 
I Extra – DL W
the cell TX i  ic 
ji
I IC – DL  ic jc    RSCPTCH – DL + RSCPOCCH  W Inter-carrier interference

TX
---------------------------------------------------------------------------
j -
9.1.5 UL TCH Eb/Nt and C/I Calculation

Term TX i  ic 
TX i  ic    RSCP TCH – UL
E b
 ----
- -  G Proc
------------------------------------------------- Div
UL  G UL None Uplink TCH Eb/Nt for the cell TX i  ic 
 N t TCH – UL TX i  ic 
N Tot – UL
Term TX i  ic 
TX  ic    RSCP TCH – UL
i
C
--- -  G Div
------------------------------------------------- None Uplink TCH C/I for the cell TX i  ic 
 I  TCH – UL TX i  ic  UL
N Tot – UL
Req Req
Term Q TCH – UL Q TCH – UL
Term - or P Term
P Max  --------------------------- Max  -------------------------
- Uplink required power for the
P Req TX i  ic  TX i  ic  W
E
 -----
b  ---
C terminal
601
9.1.6 Interference Calculation

j j
I C2C  TX i TX j  TX j  ic  TX j  ic  W Cell to cell interference

  RSCPTCH – DL + RSCPOCCH  +---------------------------------------------------------------------------
TX
j -
TX
j
TX i  ic 
TX i  ic 
TX i X TS1 – UL
N0  ---------------------------------- W UpPCH interference
I TS1 – UL TX  ic 
1 – X i 
 TS1 – UL
9.1.7 HSDPA Dynamic Power Calculations

TX i  ic  TX  ic  TX  ic 
Ec
 ----  i
TX
i i 
 N t HS – SCCH   N Tot – DL –     RSCP HS – SCCH
TX i  ic  -
W HS-SCCH power
P HS – SCCH -----------------------------------------------------------------------------------------------------------------------------  L Model
T
TX
i

i i i i i W HS-PDSCH power
P HS – PDSCH P Max – DL – Eff – P R99 – DL – P HR – P HS – SCCH
TX i  ic  TX  ic 
Ec
 ----  i
M
i M
i 
 N t HS – SICH   N Tot – UL –     RSCP HS – SICH
Mi -
W HS-SICH power
P HS – SICH --------------------------------------------------------------------------------------------------------------------------  L Model
T
Mi

9.2 Signal Level Based Calculations

Two types of signal level based calculations are available in Atoll:
1. Point Analysis: Real-time calculations for profile and reception analysis using the mouse to move a probe mobile on
the map.
2. RSCP Based Coverage Predictions: Calculation of RSCP related parameters on each pixel and colouring according to
the selected display.

For the selected transmitted TXi and carrier (ic), you can study three parameters in point analysis Profile tab:

Signal level ( RSCP ) in dBm TX  ic  TX  ic 
i i Model
RSCP = EIRP – L Path – M Shadowing – L Indoor
TX i
Path loss ( L Path ) in dB L Path = L Model + L Ant
TX i TX i
Total losses ( L T ) in dB L T = L Path + L
Model
+ L Indoor + M Shadowing – G
Where,
RSCP is the received signal code power for the P-CCPCH.
EIRP is the effective isotropic radiated power of the transmitter. EIRP = P P – CCPCH + G –L .
ic is a carrier number
L Model is the loss on the transmitter-receiver path (path loss) calculated by the propagation model
602
TX
i
L Ant is the transmitter antenna attenuation (from antenna patterns)
Model
M Shadowing is the shadowing margin. This parameter is taken into account when the option “Shadowing taken into account”
is selected
L Indoor are the indoor losses, taken into account when the option "Indoor coverage" is selected
TX i
G is the transmitter antenna gain
TX i TX i
L is the transmitter loss ( L = L Total – DL )
It is possible to analyse the best carrier. In this case, Atoll takes the highest P-CCPCH
power of cells to calculate the signal level received from a transmitter.
9.2.1.1 Profile Tab

TX i  ic 
Atoll displays either the signal level received from the selected transmitter on a carrier ( RSCP P – CCPCH ), or the highest signal
level received from the selected transmitter on the best carrier.
For a selected transmitter, it is also possible to study the path loss, L Path , or the total
losses, L T . Path loss and total losses are the same on any carrier.

Analysis provided in the Reception tab is based on path loss matrices. You can study reception from TBC transmitters for which
path loss matrices have been calculated on their calculation areas.
TX i  ic 
For each transmitter, Atoll displays either the signal level received on a carrier, ( RSCP P – CCPCH ), or the highest signal level
received on the best carrier.
Received signal level bar graphs are displayed in a decreasing signal level order. The number of bars in the graph depends on
the signal level received from the best server. Only bars for transmitters whose signal level is within a 30 dB margin from the
best server signal are displayed.
You can use a value other than 30 dB for the margin from the best server signal level, for
example a smaller value for improving the calculation speed. For more information on
defining a different value for this margin, see the Administrator Manual.
9.2.2 RSCP Based Coverage Predictions

For each TBC transmitter, TXi, Atoll determines the value of the selected parameter on each studied pixel inside the TXi
calculation area. Each pixel within the TXi calculation area is considered a probe receiver.
• The study conditions to determine the service area of each TBC transmitter
• The display settings to for colouring the covered pixels
Atoll uses the parameters entered in the Condition tab of the coverage study properties dialogue to determine pixels covered
by the each transmitter. Coverage prediction display resolution is independent of the path loss matrix and geographic data
resolutions, and can be different for each coverage prediction. Coverage predictions are calculated using bilinear interpolation
of multi-resolution path loss matrices (similar to the evaluation of site altitudes).
9.2.2.1 Calculation Criteria

The RSCP from a transmitter TXi and a selected carrier (ic) is given by:
Model Term Term
RSCP = EIRP – L Path – M Shadowing – L Body – L Indoor + G –L
Where,
603
RSCP is the received signal code power. RSCP can be calculated for P-CCPCH, DwPCH, or the downlink TCH.
i i i i
EIRP is the effective isotropic radiated power of the transmitter. EIRP P – CCPCH = P P – CCPCH + G –L ,
TX  ic  TX  ic  TX TX TX  ic  TX  ic  TX TX
i i i i i i i i
EIRP DwPCH = P DwPCH +G –L , or EIRP DL – TCH = P DL – TCH +G –L .
ic is a carrier number
TX i
TX i
Model
is selected
Term
L is the terminal loss
L Body is the body loss defined in the service
Term
G is the receiver total gain
TX i
TX i TX i
9.2.2.2 P-CCPCH RSCP Coverage Prediction

9.2.2.2.1 Coverage Condition
This coverage prediction calculates and displays the Received Signal Code Power (RSCP) for the P-CCPCH.
The coverage prediction is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for TS0. The
best server for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest P-
CCPCH power, or from the master carrier in case of N-frequency mode compatible transmitters. Afterwards, the coverage
prediction is calculated for the selected carrier. If the selected carrier does not exist on a transmitter, there will not be any
pixels covered by this transmitter. If you perform this coverage prediction for the best carrier, Atoll calculates the RSCP
considering:
• the preferred carrier of the selected service, or
• the carrier with the highest P-CCPCH power, if no preferred carrier is defined for the service, or
• the master carrier in case of N-frequency mode compatible transmitters.
You can select the display colours according to the RSCP, or on any best server parameter.
9.2.2.2.2 Coverage Display

It is possible to colour the transmitter service areas using a unique colour per transmitter, or colour the pixels in the coverage
areas by any transmitter attribute or other criteria such as:
• Best Signal Level (dBm)
TX i  ic 
Atoll calculates the best RSCP P – CCPCH received from each transmitter TX i  ic  on each pixel. Where other service
areas overlap the studied one, Atoll chooses the highest RSCP. A pixel of a service area is coloured if the RSCP level is
greater than or equal to the defined thresholds. The pixel colour depends on the RSCP level. Coverage consists of
several independent layers whose visibility in the workspace can be managed. There are as many layers as defined
thresholds. Each layer corresponds to an area where the RSCP from the best server exceeds a defined minimum
threshold.
• RSCP Margin (dB)
Coverage consists of several layers with a layer per user-defined RSCP margin defined in the Display tab (Prediction
TX i  ic 
RSCP
properties). For each layer, area is covered if RSCP P – CCPCH – TAdd P – CCPCH  Mobility   M P – CCPCH . Each layer is
604
• Cell Edge Coverage Probability (%)

TX  ic 
i
On each pixel of each transmitter service area, the coverage corresponds to the pixels where the RSCP P – CCPCH from
the transmitter exceeds TAdd P – CCPCH defined in the mobility selected in the Conditions tab, with different cell edge
coverage probabilities. There is one coverage area per transmitter in the explorer.
9.2.2.3 Best Server P-CCPCH Coverage Prediction

This coverage prediction calculates and displays the best server RSCP for the P-CCPCH. The coverage prediction is calculated
for a given set of a terminal type, a mobility type, a service, a carrier, and for TS0. The best server for the coverage prediction
is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power, or from the master carrier in
case of N-frequency mode compatible transmitters. Afterwards, the coverage prediction is calculated for the selected carrier.
If the selected carrier does not exist on a transmitter, there will not be any pixels covered by this transmitter. If you perform
this coverage prediction for the best carrier, Atoll calculates the RSCP considering:
TX i  ic 
Min
The pixels in the TX i  ic  coverage area where RSCP P – CCPCH  Max (TAdd P – CCPCH,RSCP P – CCPCH) and where
TX i  ic 
RSCP P – CCPCH = Best  RSCP TXj  jc   will be covered and coloured according to the transmitter colour.
j = All 
P – CCPCH
9.2.2.4 P-CCPCH Pollution Analysis Coverage Prediction

This coverage prediction calculates and displays the number of P-CCPCH polluters. Atoll calculates the Received Signal Code
TX i  ic 
Power (RSCP) for the P-CCPCH for each pixel in the TX i  ic  coverage area where RSCP P – CCPCH  TAdd P – CCPCH  Mobility 
and determines the polluting transmitters according to:
RSCP P – CCPCH  Best  RSCP P – CCPCH – M
ji  
Where M is the specified pollution margin.

considering:
Atoll determines the number of transmitters covering each pixel and colours the pixel according to the number of polluting
transmitters. Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as
many layers as defined thresholds. Each layer corresponds to an area where the number of servers is greater than or equal to
a defined minimum threshold.
9.2.2.5 DwPCH RSCP Coverage Prediction

This coverage prediction calculates and displays the Received Signal Code Power (RSCP) for the DwPCH. The coverage
prediction is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for DwPTS. The best server
for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power,
or from the master carrier in case of N-frequency mode compatible transmitters. Afterwards, the coverage prediction is
calculated for the selected carrier. If the selected carrier does not exist on a transmitter, there will not be any pixels covered
by this transmitter. If you perform this coverage prediction for the best carrier, Atoll calculates the RSCP considering:
605
TX  ic 
i Min
The pixels in the TX i  ic  coverage area where RSCP P – CCPCH  Max (TAdd P – CCPCH,RSCP P – CCPCH) and where
TX  ic 
i Req
RSCP DwPCH  RSCP DwPCH  Mobility  are covered and coloured according to the selected display parameter.

• DwPCH RSCP (dBm)
TX i  ic 
Atoll calculates the best RSCP DwPCH received from each transmitter TX i  ic  on each pixel.. Where other service
areas overlap the studied one, Atoll chooses the highest RSCP. A pixel of a service area is coloured if
TX i  ic 
Req
RSCP DwPCH  RSCP DwPCH  Mobility  . The pixel colour depends on the RSCP level. Coverage consists of several
independent layers whose visibility in the workspace can be managed. There are as many layers as defined thresholds.
Each layer corresponds to an area where the RSCP from the best server exceeds a defined minimum threshold.
TX  ic 
i Req RSCP
properties). For each layer, area is covered if RSCP DwPCH – RSCP DwPCH  Mobility   M DwPCH . Each layer is assigned
• Cell edge coverage probability (%)
TX  ic 
i
On each pixel of each transmitter service area, the coverage corresponds to the pixels where the RSCP DwPCH from
TX i  ic 
the transmitter TX i  ic  exceeds RSCP DwPCH defined in the mobility selected in the Conditions tab, with different cell
edge coverage probabilities. There is one coverage area per transmitter in the explorer.
9.2.2.6 UpPCH RSCP Coverage Prediction

This coverage prediction calculates and displays the Received Signal Code Power (RSCP) for the UpPCH in the uplink. The
coverage prediction is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for UpPTS. The best
server for the coverage prediction is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH
power, or from the master carrier in case of N-frequency mode compatible transmitters. Afterwards, the coverage prediction
is calculated for the selected carrier. If the selected carrier does not exist on a transmitter, there will not be any pixels covered
Atoll uses the UpPCH power of the selected terminal to calculate the RSCP from each pixel of each transmitter’s best server
coverage area.
TX i  ic 
Min Term Req
The pixels where RSCP P – CCPCH  Max (TAdd P – CCPCH,RSCP P – CCPCH) and where RSCP UpPCH  RSCP UpPCH  Mobility  are
covered and coloured according to the selected display parameter.

• UpPCH RSCP (dBm)
Term
Atoll calculates the best RSCP UpPCH received from each pixel of each transmitter service area at the transmitter.
Where other service areas overlap the studied one, Atoll chooses the highest RSCP. A pixel of a service area is coloured
Term Req
if RSCP UpPCH  RSCP UpPCH  Mobility  . The pixel colour depends on the RSCP level. Coverage consists of several
independent layers whose visibility in the workspace can be managed. There are as many layers as defined thresholds.
Each layer corresponds to an area where the RSCP at the best server exceeds a defined minimum threshold.
606
Term Req RSCP
properties). For each layer, area is covered if RSCP UpPCH – RSCP UpPCH  Mobility   M UpPCH . Each layer is assigned a
Term
On each pixel of each transmitter service area, the coverage corresponds to the pixels from which the RSCP UpPCH at
Term
the transmitter exceeds RSCP UpPCH defined in the mobility selected in the Conditions tab, with different cell edge
coverage probabilities. There is one coverage area per transmitter in the explorer.
9.2.2.7 Baton Handover Coverage Prediction

This coverage prediction determines the pixels which receive RSCP from cells other than the best server high enough to
perform baton handovers. Received Signal Code Power (RSCP) is calculated for the P-CCPCH. The coverage prediction is
calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for TS0. The best server for the coverage
prediction is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power, or from the master
carrier in case of N-frequency mode compatible transmitters. Afterwards, the coverage prediction is calculated for the
selected carrier. If the selected carrier does not exist on a transmitter, there will not be any pixels covered by this transmitter.
If you perform this coverage prediction for the best carrier, Atoll calculates the RSCP considering:
The pixels are covered and coloured according to the selected display parameters, where
TX  ic  TX  jc  TX  jc 
i j j
RSCP P – CCPCH  TAdd P – CCPCH  Mobility  and RSCP P – CCPCH  TDrop P – CCPCH  Mobility  – TComp P – CCPCH .

It is possible to display the potential handover areas or the number of transmitters covering each pixel.
• Handover Areas
Atoll displays the pixels where there are transmitters other than the best server that satisfy the above criteria.
Coverage consists of a single layer with a defined colour whose visibility in the workspace can be managed.
• Number of Potential Servers
Atoll determines the number of transmitters covering each pixel and colours the pixel according to the number of
transmitters. Coverage consists of several independent layers whose visibility in the workspace can be managed.
There are as many layers as defined thresholds. Each layer corresponds to an area where the number of servers is
greater than or equal to a defined minimum threshold.
9.2.2.8 Scrambling Code Interference Analysis

This coverage prediction calculates and displays the pixels covered by two cells using the same scrambling code. Atoll
calculates the Received Signal Code Power (RSCP) for the P-CCPCH for each pixel in the TX i  ic  coverage area where
TX  ic 
i
RSCP P – CCPCH  TAdd P – CCPCH  Mobility  and determines the interfering transmitters according to:
RSCP P – CCPCH  Best  RSCP P – CCPCH – M
ji  
Where M is the specified pollution margin.

considering:
Atoll determines whether the cells of two transmitters covering a pixel have the same scrambling code. If the pixel is
interfered, Atoll colours it according to the colour assigned to the scrambling code in the display parameters. Coverage
607
consists of several independent layers whose visibility in the workspace can be managed. There are as many layers as
scrambling codes. Each layer corresponds to the area where the corresponding scrambling code has interference. A layer
corresponding to areas where more than one scrambling code interferes is also available.
9.3 Monte Carlo Simulations

• Generating a realistic user distribution as explained in "Generating a Realistic User Distribution" on page 608.
• Dynamic channel allocation and power control as explained under "Power Control Simulation" on page 613.

[Simulation]
RandomTotalUsers=0
Each user is randomly assigned a service, a terminal, and a mobility type. The activity status is determined based on the
calculations of activity probabilities using the traffic inputs.
The user activity status influences the next step of the simulation, i.e., the interference in the network. Both active and
inactive users use radio resources and generate interference.
indoor users per clutter class defined for the traffic maps.
Atoll also calculates the shadowing margin for each user based on the standard deviations defined for the clutter class of each
user.
In TD-SCDMA networks users accessing packet-switched services can transmit either on uplink or on downlink, but never on
both simultaneously. Users accessing circuit-switched services transmit on both uplink and downlink simultaneously. Circuit-
switched service users, mobiles connected in uplink and downlink both, are modelled in Atoll by two mobiles generated at
the same location with one connected on the uplink and the other on the downlink. If one of these two mobiles is rejected
for some reason, the other is also rejected due to the same reason.
9.3.1.1 Simulations Based on User Profile Traffic Maps

The user profile models the behaviour of the different user categories. Each user profile contains a list of services and their
associated parameters describing how these services are accessed by the user.
608
• In case of user profile traffic maps composed of lines, the number of users per user
profile is calculated from the line length (L) and the user profile density (DUP) (users
per km): N Users = L  D UP
• The number of users is an input when a user profile traffic map is composed of points.
At any given instant, Atoll calculates the probability for a user being active in the uplink and in the downlink according to the
duration of each voice call, or the volumes of the data exchanged in the uplink and the downlink in each data session.
9.3.1.1.1 Circuit Switched Service (i)

User profile parameters for circuit switched services are:
• The user terminal equipment used for the service (from the Terminals table),
• The average number of calls per hour N Call ,
N call  d
p o = -------------------
-
3600
• Calculation of the number of users trying to access the service i ( n i ):
n i = N Users  p 0
The activity status of each user depends on the activity periods during the connection, i.e., the uplink and downlink activity
UL DL
factors defined for the circuit switched service i, f Act and f Act .
• Calculation of activity probabilities:

UL DL
UL UL DL
Probability of being active on UL: p Active = f Act   1 – f Act 
DL DL UL
Probability of being active on DL: p Active = f Act   1 – f Act 
UL + DL UL DL
Probability of being active both on UL and DL: p Active = f Act  f Act

Number of inactive users: n i – Inactive = n i  p Inactive
UL UL
Number of users active in the uplink: n i – Active = n i  p Active
DL DL
Number of users active in the downlink: n i – Active = n i  p Active
UL + DL UL + DL
Number of users active in the uplink and downlink both: n i – Active = n i  p Active
Therefore, a connected user can be either active on both links, inactive on both links, active on UL only, or active on DL only.
9.3.1.1.2 Packet Switched Service (j)

User profile parameters for packet switched services are:
• The user terminal equipment used for the service (from the Terminals table),
• The average number of packet sessions per hour N Sess ,
DL UL
• The volume (in kBytes) which is transferred on the downlink V and the uplink V during a session.
A packet session consists of several packet calls separated by a reading time. Each packet call is defined by its size and may be
divided in packets of fixed size (1500 Bytes) separated by an inter-packet arrival time.
609
Figure 9.1: Description of a Packet Session
Calculation of the average packet call size (kBytes):

UL DL
UL V DL V
S PacketCall = --------------------------------------
- and S PacketCall = ---------------------------------------
UL UL DL DL
N PacketCall  f Eff N PacketCall  f Eff
UL DL
In case of HSDPA services, f Eff and f Eff are the uplink and downlink A-DPCH activity factors, respectively.
Calculation of the average number of packets per packet call:

UL DL
UL  S PacketCall   S PacketCall 
- + 1 and N DL
N Packet = Int  ------------------------------- Packet = Int  -------------------------------
- + 1
UL
 S Packet  1024  S DL
Packet  1024
1 kBytes = 1024 Bytes.
Calculation of the average duration of inactivity within a packet call (c):

UL UL DL DL
UL  N Packet – 1   T Packet DL  N Packet – 1   T Packet
 D Inactivity  PacketCall = ---------------------------------------------------------
- and  D Inactivity  PacketCall = ---------------------------------------------------------
-
1000 1000
Calculation of the average duration of inactivity in a session (s):

UL UL UL
 D Inactivity  Session = N PacketCall   D Inactivity  PacketCall and
DL DL DL
 D Inactivity  Session = N PacketCall   D Inactivity  PacketCall
Calculation of the average duration of activity in a session (s):

UL UL
UL UL N Packet  S Packet  8
 D Activity  Session = N PacketCall  ------------------------------------------------
UL
- and
R Nom  1000
DL DL
DL DL N Packet  S Packet  8
 D Activity  Session = N PacketCall  ------------------------------------------------
DL
-
R Nom  1000
Therefore, the average duration of a connection in the session s is:

UL UL UL DL DL DL
D Connection =  D Activity  Session +  D Inactivity  Session and D Connection =  D Activity  Session +  D Inactivity  Session
Calculation of the service usage duration per hour (probability of a connection):
UL N Sess UL DL N Sess DL
p Connection = ------------  D Connection and p Connection = ------------  D Connection
3600 3600
Calculation of the probability of being connected:
610
UL DL
p Connected = 1 –  1 – p Connection    1 – p Connection 
Therefore, the number of users trying to access the service j is:
n j = N Users  p Connected
As Figure 9.1 on page 610 shows, there can be three possible cases when a user is connected:
a. 1st case: At a given time, packets are downloaded and uploaded.

UL DL
UL + DL p Connection  p Connection
The probability of being connected is: p Connected = ---------------------------------------------------------
-
p Connected
b. 2nd case: At a given time, packet are uploaded only.

UL DL
UL p Connection   1 – p Connection 
The probability of being connected is: p Connected = -----------------------------------------------------------------------
-
p Connected
c. 3rd case: At a given time, packet are downloaded only.

DL UL
DL p Connection   1 – p Connection 
The probability of being connected is: p Connected = -----------------------------------------------------------------------
-
p Connected
Calculation of the probability of being active:

To determine the activity status of each user, the activity periods during the connection are taken into account.
UL DL
UL  D Activity  Session DL  D Activity  Session
f = ---------------------------------------------------------------------------------------
- and f = ---------------------------------------------------------------------------------------
-
UL UL DL DL
 D Inactivity  Session +  D Activity  Session  D Inactivity  Session +  D Activity  Session
Therefore, we have:
a. 1st case: At a given time, packets are downloaded and uploaded.

UL UL DL UL + DL
The probability of the user being active on UL and inactive on DL: p1 Active = f   1 – f   p Connected
DL DL UL UL + DL
The probability of the user being active on DL and inactive on UL: p1 Active = f   1 – f   p Connected
UL + DL UL DL UL + DL
The probability of the user being active on both UL and DL: p1 Active = f f  p Connected
UL DL UL + DL
The probability of the user being inactive on both UL and DL: p1 Inactive =  1 – f    1 – f   p Connected
b. 2nd case: At a given time, packet are uploaded only.

UL UL UL
The probability of the user being active on UL and inactive on DL: p2 Active = f  p Connected
UL UL
The probability of the user being inactive on both UL and DL: p2 Inactive =  1 – f   p Connected
c. 3rd case: At a given time, packet are downloaded only.

DL DL DL
The probability of the user being active on DL and inactive on UL: p1 Active = f  p Connected
DL DL
The probability of the user being inactive on both UL and DL: p3 Inactive =  1 – f   p Connected

Number of inactive users on UL and DL: n j – Inactive = n j   p1 Inactive + p2 Inactive + p3 Inactive 
UL UL UL
Number of users active on UL and inactive on DL: n j – Active = n j   p1 Active + p2 Active 
DL DL DL
Number of users active on DL and inactive on UL: n j – Active = n j   p1 Active + p3 Active 
UL + DL UL + DL
Number of users active on UL and DL: n j – Active = n j   p1 Active 
Therefore, a connected user can be active on both links, inactive on both links, active on UL only, or active on DL only.
611
The user distribution per service, and the activity status distribution between the users
are average distributions. The service and the activity status of each user are random in
each simulation. Therefore, if you compute several simulations at once, the average
number of users per service and average numbers of inactive, active on UL, active on DL,
and active on UL and DL users, will correspond to calculated distributions. But, if you
compare each simulation, you will observe that the user distribution between services as
well as the activity status distribution between users is different in each simulation.
9.3.1.2 Simulations Based on Sector Traffic Maps

Sector traffic maps are also referred to as live traffic maps. Live traffic data from the OMC is spread over the best server
coverage areas of the transmitters included in the traffic map. Throughput demands per service, the numbers of active users
per service, or Erlangs per service are assigned to the coverage areas of each transmitter.
9.3.1.2.1 Throughputs in Uplink and Downlink

When selecting Throughputs in Uplink and Downlink, you can input the throughput demands in the uplink and downlink for
each sector and for each listed service.
Atoll calculates the number of users active in uplink and in downlink in the Txi cell using the service (NUL and NDL) as follows:
UL DL
UL RS DL RS
N = -----------
- and N = -----------
- for R99 circuit and packet switched services
UL DL
R Nom R Nom
DL
DL RS
N = ---------- for HSDPA service
DL
R Avg
UL DL
R S and R S are the uplink and downlink throughputs for service S in the TXi cell from the traffic map.
NUL and NDL values include:

UL
• Users active in uplink and inactive in downlink ( n i – Active ),
DL
• Users active in downlink and inactive in uplink ( n i – Active ),
UL + DL
• And users active in both links ( n i – Active ).
UL DL
Probability of being inactive in UL and DL: p Inactive =  1 – f Act    1 – f Act 
UL UL DL
Probability of being active in UL only: p Active = f Act   1 – f Act 
DL DL UL
Probability of being active in DL only: p Active = f Act   1 – f Act 
UL + DL UL DL
Probability of being active both in UL and DL: p Active = f Act  f Act
UL DL
Where, f Act and f Act are respectively the UL and DL activity factors defined for the service i.

We have:
UL UL + DL UL DL UL + DL
 p Active + p Active    n i – Active + n i – Active + n i – Active  = N UL
DL UL + DL UL DL UL + DL
 p Active + p Active    n i – Active + n i – Active + n i – Active  = N DL
Therefore, we have:
UL + DL UL + DL
UL + DL  N UL  p Active N DL  p Active 
Number of users active in UL and DL both: n i – Active = min  ---------------------------------------
UL + DL
 ---------------------------------------
+ DL

UL DL
Active 
UL UL + DL
Number of users active in UL and inactive in DL: n i – Active = N UL – n i – Active
612
DL UL + DL
Number of users active in DL and inactive in UL: n i – Active = N DL – n i – Active
UL DL UL + DL
 n i – Active + n i – Active + n i – Active 
Number of inactive users in UL and DL: n i – Inactive = --------------------------------------------------------------------------------  p inactive
1 – p inactive
9.3.1.2.2 Total Number of Users (All Activity Statuses)

When selecting Total Number of Users (All Activity Statuses), you can input the number of connected users for each sector
and for each listed service ( n i ).
UL DL
Probability of being inactive in UL and DL: p Inactive =  1 – f Act    1 – f Act 
UL UL DL
Probability of being active in UL only: p Active = f Act   1 – f Act 
DL DL UL
Probability of being active in DL only: p Active = f Act   1 – f Act 
UL + DL UL DL
Probability of being active both in UL and DL: p Active = f Act  f Act
UL DL
Where, f Act and f Act are respectively the UL and DL activity factors defined for the service i.

Number of inactive users in UL and DL: n i – Inactive = n i  p Inactive
UL UL
Number of users active in UL and inactive in DL: n i – Active = n i  p Active
DL DL
Number of users active in DL and inactive in UL: n i – Active = n i  p Active
UL + DL UL + DL
Number of users active in UL and DL both: n i – Active = n i  p Active
9.3.1.2.3 Number of Users per Activity Status

When selecting Number of Users per Activity Status, you can directly input the number of inactive users ( n i – Inactive ), the
UL DL UL + DL
number of users active in the uplink ( n i – Active ), in the downlink ( n i – Active ) and in the uplink and downlink ( n i – Active ), for
each sector and for each service.
The activity status of users is based on an average distribution. The activity status of each
user is random in each simulation. Therefore, if you compute several simulations at once,
the average numbers of inactive, active on UL, active on DL, and active on UL and DL
users, will correspond to calculated distributions. But, if you compare each simulation,
you will observe that the activity status distribution between users is different in each
simulation.
9.3.2 Power Control Simulation

Based on CDMA air interface, a TD-SCDMA network automatically regulates itself by using uplink and downlink power control
in order to minimise interference and maximise capacity. For each user distribution, Atoll simulates these network regulation
mechanisms using an iterative algorithm and calculates network parameters such as traffic power per cell and per timeslot,
mobile terminal power, and handoff status for each terminal.
In each iteration, all the mobiles (R99 and HSDPA service users) selected during generation of the user distribution attempt
to connect to the network one by one. The process is repeated from iteration to iteration and ends when the network is
balanced, i.e., when the convergence criteria on uplink and downlink are satisfied.
The simulation algorithm also models the impact of smart antennas in the power control loop. The influence of smart
antennas is taken into account in signal quality calculations. Smart antennas improve the signal quality of each served mobile,
613
decrease the required powers and the loads of all the surrounding cells. Interference on the downlink and the uplink is
calculated on a per user. Power control is simulated over a sub-frame, i.e., 7 timeslots.
For HSDPA users, uplink and downlink power control is performed on the associated A-DCH bearer before fast link adaptation
on downlink. The steps of this algorithm are detailed below.
Figure 9.2: TD-SCDMA Power Control Algorithm
9.3.2.1 Algorithm Initialisation

At the start of each simulation, the system loads for each carrier and timeslot are reset to initial values:
• Downlink traffic powers of cells P TCH – DL are initialised to 0 Watts
• Uplink interference powers received on all the carriers and timeslots I Intra – UL and I Extra – UL are initialised to 0 Watts
(i.e., no connected mobiles)
Term
• Uplink required power for mobiles is set to P Min
9.3.2.2 R99 Part of the Algorithm

Req
The algorithm is described for an iteration k. Here, Xk is the value of the variable X at the iteration k. In the algorithm, all Q UL
Req
and Q DL thresholds depend on the user mobility, and are defined in the Service and Mobility parameter tables. All the
variables used in the description below are listed in "Definitions and Formulas" on page 595.
The following calculations are made for all R99 and HSDPA mobiles (Mi) using R99 bearers.
9.3.2.2.1 Determination of Mi’s Best Server (SBS(Mi))

This step is performed for TS0 for each station TXi containing Mi in its calculation area.
The best server for Mi is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power, or from
the master carrier in case of N-frequency mode compatible transmitters. Afterwards, the P-CCPCH RSCP is calculated for:
• the preferred carrier of the service used by Mi, or
614
The RSCP from a transmitter TXi and a selected carrier ic is given by:
TX  ic  TX  ic  TX TX M M M
i i i i Model i i i
RSCP P – CCPCH = P P – CCPCH + G –L – L Path – M Shadowing – L Body – L Indoor + G –L in dBm
Where,
TX i
TX i
Model
is selected
Mi
L is the los of the terminal used by Mi
Mi
L Body is the body loss defined in the service used by Mi
Mi
G is the receiver gain of the terminal user by Mi
TX i
TX i TX i
A cell TX i  ic  is considered the best server of a mobile Mi if it satisfies the following conditions:
TX i  ic 
Min
RSCP P – CCPCH  RSCP P – CCPCH ,
TX i  ic 
RSCP P – CCPCH  TAdd P – CCPCH  Mobility  ,
TX i  ic 
And RSCP P – CCPCH = Best  RSCP TXj  jc   .
j = All 
P – CCPCH
The best server is determined once for the whole simulation during the first iteration, i.e., k = 0, because the best server does
not change during the simulation and smart antennas do not influence this step.
Mi is considered unable to connect to the network if no best server has been selected. In this case, Mi is rejected for the reason
P-CCPCH RSCP < Min P-CCPCH RSCP. If Mi has no best server, it is not taken into account in the next steps.
9.3.2.2.2 Dynamic Channel Allocation

The dynamic channel allocation is performed once for the whole simulation during the first iteration, i.e., k = 0. The DCA
controls the mobile admission. Once a mobile has been admitted for a simulation, it remains admitted for the all the iterations
unless there are other reasons to reject it (following steps).
The aim of Dynamic Channel Allocation (DCA) is to reduce interference in order to maximise the usage of the radio resources.
In other words, the DCA tries to find the "best carrier" and the "best timeslots", which when allocated to the mobiles will
optimise the load balance between carriers.
If a preferred carrier is defined for the service requested by Mi and if it is available at TX i . BestCarrier  TX i M i  = the
carrier preferred for the service. In the case of N-frequency compatible transmitters, Mi can be allocated timeslots over more
than one slave carrier.
Mi is considered unable to connect to the network if no carrier or not enough timeslots have been selected. In this case, the
mobile Mi will be rejected for the reason "RU Saturation". If the carrier and timeslot(s) selected by the DCA do not satisfy the
control of radio resource limits for DL power or UL load, then the mobile will be rejected for the reason "DL Load Saturation"
or "Admission Rejection" respectively.
615
There are four strategies for the DCA available in Atoll. These strategies are described below one by one.
1. Load
Carrier Selection by Load: The DCA determines the least loaded carrier with enough timeslots to accomodate the
service being used by each mobile Mi. The best carrier for a mobile is the one that is least loaded:
BestCarrier  TX i M i  = Carrier DCA

Min  X 
DCA TX i  ic TS  M i  
DCA
Where, X = X DL = N Tot – DL if the mobile is connected in the downlink.
TX i  ic TS  M i  
DCA DCA N Tot – UL
And, X = X UL -  X DCA if the mobile is connected in the uplink.
= -----------------------------------------------
TX i  ic TS  M i   TX i
N Tot – UL + N0
DCA
X is the load increment given by:
Mi TX i
   1 – f UL    1 – f JD 
Ortho
DCA  
X = ----------------------------------------------------------------------
1
1 + -----------
Req
Q UL
E b
 ----
Req
-
Req  N t UL
Where Q UL =  ---
Req C Proc
= ------------------ is the uplink required signal quality. The uplink processing gain G UL calculated
 I  UL Proc
G UL
from the service parameters, if no smart antenna is used by the transmitter in the uplink.
If a smart antenna is used by the transmitter in the uplink, the smart antenna gain is taken into account in calculating
Req
Q UL .
TX i  ic TS  M i  
• N Tot – UL is described in "Uplink Power Control" on page 617.
TX i  ic TS  M i  
• N Tot – DL is described in "Downlink Power Control" on page 619.
• The carrier is the same in the uplink and in the downlink for mobiles accessing circuit-
switched services.
Timeslot selection by Load: From the selected carrier, Atoll selects the timeslots which are the least loaded and have
enough resource units for the service being accessed by Mi.
2. Available RUs
Carrier selection by Available RUs: The DCA determines the carrier which has the highest number of available
resource units with enough timeslots to accomodate the service being used by each mobile Mi. The best carrier for a
mobile is the one that has the highest number of resource units:
BestCarrier  TX i M i  = Carrier Max  RUs 
Timeslot selection by Available RUs: From the selected carrier, Atoll selects the timeslots which have the highest
numbers of available resource units.
3. Direction of Arrival
Carrier selection by Direction of Arrival: The DCA determines the direction of arrival of the signal from the served user
Mi and checks whether there is an interfering mobile in the same direction as Mi. Atoll searches for interfering mobiles
within the angle defined by the Angular Step. For example, if you enter an angular step of 15 degrees, Atoll searches
for interfering mobiles within 15 degrees to the right and to the left of the served user, and allocates a different carrier
than the ones used by any interfering mobiles found. The best carrier for a mobile is the one which is not interfered
by another mobile in the direction of the mobile Mi.
BestCarrier  TX i M i  = Carrier DoA  Mi   DoA  Mj 
In other words, the direction of arrival for the served user Mi should not be the direction of arrival of an interfering
mobile.
616
Timeslot selection by Direction of Arrival: From the selected carrier, Atoll selects the timeslots which are not being
used by any other mobile Mj located in the same direction as the served user Mi.
4. Sequential
Sequential carrier selection: The DCA allocates carriers to served users Mi in a sequential order.
Sequential timeslot selection: From the selected carrier, Atoll allocates timeslots to served users Mi in a sequential
order.
At the end of the DCA, each admitted mobile has an associated carrier and timeslots. In case of N-frequency mode compatible
transmitters, an admitted mobile can have associated timeslots over more than one slave carrier.
9.3.2.2.3 Uplink Power Control

For each mobile Mi, the uplink power control step calculates the uplink power required to satisfy the required quality level on
the traffic channel, which is defined for the service being accessed by Mi.
If the mobile Mi is connected (active or inactive) in the uplink and has a best server TX i  ic  assigned to it, Atoll calculates the
signal quality on the uplink timeslots allocated to Mi by the DCA:
Mi TX i  ic TS  M i   Mi TX i  ic TS  M i  
TX i  ic TS  M i   TX i  ic TS  M i  
E b
 ----   RSCP TCH – UL  C   RSCP TCH – UL
- -  G Proc
= -------------------------------------------------------
Div
UL  G UL or  --- -  G Div
= -------------------------------------------------------
 N t TCH – UL TX i  ic TS  M i   I TCH – UL TX i  ic TS  M i   UL
N Tot – UL N Tot – UL
Calculation of Uplink Total Noise ( N Tot – UL ):
The uplink total noise is calculated for the uplink connection between each mobile Mi and its best server TX i  ic  .
TX i  ic TS  M i   TX i  ic TS  M i   TX i
N Tot – UL = I Tot – UL + N0
Where
Mi Mi
TX i  ic TS  M i   RSCP TCH – UL  TX i  ic TS  M i        +
I Tot – UL = Mj Mi
 RSCP TCH – UL  TX i  ic TS  M i      +
M j  TX i  ic TS  M i  
Mj  Mi
M M
 1 –  j  RSCP j
   TCH – UL  TX i  ic TS  M i    +
M j  TX i  ic TS  M i  
Mj  Mi
M TX
 RSCP TCH – UL  TX i  ic TS  M i      1 – F MCJD
j i
 
M  TX  ic TS  M  
j i i
M M TX 
   1 – F JD  and  =  0 Without Useful Signal
i i Ortho i
 =    1 – F UL
 1 Total Noise
The above formula gives the value of I Tot – UL for the uplink connection between Mi and TX i  ic  , taking into account the
interference received from other mobiles, Mj, which are located in the Mi best server coverage area, as well as located in the
coverage areas of other cells. The mobile Mi is the focus, i.e., the mobile that is listened to by the transmitter TX i  ic  .
617
The four terms comprising I Tot – UL are:
• The useful signal for which the received mobile is the focus (Mi).
• The intra-cell interference for which the best-server is the same for the received mobile Mj and the focus Mi, TX i  ic  .
• The intra-cell interference due to distortion in the terminal transmission.

• The extra-cell interference for which the best-server for the received mobile Mj is not TX i  ic  .
Mi
P Req  TX i  ic TS  M i   
Mi
k–1
The uplink received signal code power is: RSCP TCH – UL  TX i  ic TS  M i    = ---------------------------------------------------------------
-
Model
LT
TX i Mi Mi
Model
Model L Path  L  L  L Body  L Indoor  M Shadowing Mi
LT - and P Req  TX i  ic TS  M i    is the uplink required mobile power
= ------------------------------------------------------------------------------------------------------------------
TX i Mi
G G
Mi Mi
calculated for the timeslot allocated to Mi. If Mi is an HSDPA user, P Req  TX i  ic TS  M i    = 0.1  PReq  TX i  ic TS  M i   
TX i TX i
Model SA SA
In L T , G = G UL and L = L UL are calculated according to the smart antenna modelling method used, for
Mi TX i TX i
P Req  TX i  ic TS  M i    , if a smart antenna is available in the uplink. Otherwise, G and L are read from the main antenna
model.
Interference is updated only for active mobiles on the uplink for circuit- and packet-
switched services. However, if these mobiles are rejected, they are considered in the
number of rejected mobiles.
Mi
Calculation of Uplink Required Power ( P Req ):
Then Atoll determines the required uplink power by:

E b Req
 -----
Mi Mi  N t TCH – UL
P Req  TX i  ic TS  M i    = P Req  TX i  ic TS  M i     --------------------------------------
-
k k–1 E b TXi  ic TS  M i  
 -----
 N t TCH – UL
Req
C---
Mi Mi  I  TCH – UL
or P Req  TX i  ic TS  M i    = P Req  TX i  ic TS  M i     ------------------------------------
TX i  ic TS  M i  
-
k k–1
 ---
C
 I  TCH – UL
M M M M
i i i i
And if P Req  TX i  ic TS  M i     P Min then P Req  TX i  ic TS  M i    = P Min
M M
i i
If P Req  TX i  ic TS  M i     P Max then the mobile Mi is rejected for the reason "Pmob > PmobMax", and
Mi
P Req  TX i  ic TS  M i    is set to 0.
Mi Mi
P Min and P Max are set in the properties of the terminal used by the mobile Mi.
Mi
TX i  ic TS  M i   P Req  TX i  ic TS  M i   
k–1
Where RSCP TCH – UL = ---------------------------------------------------------------
-
Model
LT
TX i Mi Mi
Model
Model L Path  L  L  L Body  L Indoor  M Shadowing Mi
LT - and P Req  TX i  ic TS  M i   
= ------------------------------------------------------------------------------------------------------------------ is the uplink required mobile
TX i Mi k–1
G G
power for iteration k - 1 transmitted on the timeslot allocated to Mi.
618
TX TX
Model i SA i SA
In L T , G = G UL and L = L UL are calculated according to the smart antenna modelling method used, for
M TX TX
i i i
P Req  TX i  ic TS  M i    , if a smart antenna is available in the uplink. Otherwise, G and L are read from the main antenna
model.
The uplink required powers for mobiles inactive in the uplink accessing circuit- or packet-
switched services are calculated for information only. However, if these mobiles are
rejected, they are considered in the number of rejected mobiles.
9.3.2.2.4 Downlink Power Control

For each mobile Mi, the downlink power control step calculates the downlink power for the best server TX i  ic  required to
satisfy the required quality level on the traffic channel, which is defined for the service being accessed by Mi.
If the mobile Mi is connected (active or inactive) in the downlink and has a best server TX i  ic  assigned to it, Atoll calculates
the signal quality on the uplink timeslots allocated to Mi by the DCA:
TX i TX i  ic TS  M i   TX i TX i  ic TS  M i  
E b
 ----   RSCP TCH – DL C    RSCP TCH – DL
- -  G Proc
= --------------------------------------------------------  G
Div
or --
- -  G Div
= --------------------------------------------------------
 N t TCH – DL TX i  ic TS  M i   DL DL  I  TCH – DL TX i  ic TS  M i   DL
N Tot – DL N Tot – DL
Calculation of Downlink Total Noise ( N Tot – DL ):
The downlink total noise is calculated for the downlink connection between each mobile Mi and its best server TX i  ic  .
TX i  ic TS  M i   TX i  ic TS  M i   Mi
N Tot – DL = I Tot – DL + I IC – DL  ic jc  + I MM  M i M j  + N 0
Where
TX i  ic TS  M i   RSCP Tot – DL  Mi      +
I Tot – DL =
 RSCP Tot – DL  Mj    +
M  TX  ic TS  M  
j i i
Mj  Mi
TX TX  ic TS  M i  
 1 –  i  RSCP i
   Tot – DL  Mj  +
M j  TX i  ic TS  M i  
Mj  Mi
TX j  ic TS  M i  
 RSCP Tot – DL  Mj 
M j  TX i  ic TS  M i  
The four terms comprising I Tot – DL are:
• The useful signal for which the received mobile is the focus (Mi).
• The intra-cell interference for which the best-server is the same for the received mobile Mj and the focus Mi, TX i  ic  .
• The intra-cell interference due to distortion in the transmitter.

• The extra-cell interference for which the best-server for the received mobile Mj is not TX i  ic  .
TX j  jc TS  M i  
 RSCPTot – DL  Mi 
All TX
I IC – DL  ic jc  = ----------------------------------------------------------------
j
619
TX TX M 
   1 – F JD  and  =  0 Without Useful Signal
i i Ortho i
 =    1 – F DL
 
 1 Total Noise
I IC – DL  ic jc  is the inter-carrier interference from a carrier jc to another carrier ic on the downlink, which is reduced by the
interference reduction factor F IRF  ic jc  defined for the pair (ic, jc).
TX i  ic TS  M i   TX i  ic TS  M i   TX i  ic TS  M i  

RSCP Tot – DL = RSCP TCH – DL + RSCP OCCH
TX i  ic TS  M i   TX  ic TS  M  
TX i  ic TS  M i   P TCH – DL TX i  ic TS  M i   P OCCH
i i
k–1
With RSCP TCH – DL = -----------------------------------------
- and RSCP OCCH = ------------------------------
-
Model Model
LT LT
TX i Mi Mi
Model
Model L Path  L  L  L Body  L Indoor  M Shadowing TX i  ic TS  M i  
LT = ------------------------------------------------------------------------------------------------------------------
- and P TCH – DL is the downlink traffic power transmitted
TX i Mi k–1
G G
TX  ic TS  M   TX  ic TS  M  
i i i i
on the timeslot allocated to Mi during the iteration k - 1. If Mi is an HSDPA user, P TCH – DL = 0.1  P TCH – DL
TX i TX i
Model SA SA
In L T , G = G DL and L = L DL are calculated according to the smart antenna modelling method used, for
TX i  ic TS  M i   TX i  ic TS  M i   TX i TX i
P TCH – DL only and not for P OCCH , if a smart antenna is available in the downlink. Otherwise, G and L are
read from the main antenna model.
Mj
 RSCPTCH – UL  Mi 
M
I MM  M i M j  = -------------------------------------------------
j is the interference from each mobile Mj transmitting in the uplink on the same
timeslots as those on which the mobile Mi is receiving in the downlink. Mj can interfere Mi directly if and only if:
Mi – Mj
• The distance between Mi and Mj ( d ) is less than the Max Distance between interfering mobiles defined by the
user when starting the simulation, and
• The downlink timeslot of Mi (TSMi) is the same as the uplink timeslot of Mj, (TSMj).
The interference received from the mobile Mj at the mobile Mi is calculated using either the free-space propagation model or
the Xia model.
Mj
Mj P TCH – UL
RSCP TCH – UL  M i  = -------------------
-
L MM
 Mi – Mj
 32.4 + 20  Log  F Avg  + 20  Log  d  If d 3m
L MM =  with F Avg being the average frequency in MHz of the
 Mi – Mj
 49 + 30  Log  F Avg  + 40  Log  d  If d 3m
frequency band used by the best server of the mobile Mi, and d is the distance between the mobiles Mi and Mj in km.
TX i  ic TS  M i  
Calculation of Downlink Required Power ( P Req ):
Then Atoll determines the required downlink power by:

E b Req
 -----
TX i  ic TS  M i   TX i  ic TS  M i    N t TCH – DL
P Req = P Req  --------------------------------------
-
k k–1 E b TXi  ic TS  Mi  
 -----
 N t TCH – DL
Req
C---
TX i  ic TS  M i   TX i  ic TS  M i    I  TCH – DL
or P Req = P Req  ------------------------------------
TX i  ic TS  M i  
-
k k–1
C ---
 I  TCH – DL
Min Min
And if P Req  P TCH – DL  Service  then P Req = P TCH – DL  Service 
620
TX  ic TS  M   TX  ic TS  M  
i i Max i i
If P Req  P TCH – DL  Service  then the mobile Mi is rejected for the reason "Ptch > PtchMax", and P Req is set
to 0.
Min Max
P TCH – DL  Service  and P TCH – DL  Service  are set in the properties of the R99 bearer associated with the service used by the
mobile Mi.
Otherwise, the downlink traffic power is incremented P TCH – DL = P TCH – DL + P Req
For each mobile, Atoll also calculates the downlink traffic power for the different values of the Angular Step  Step .
SA
TX i  ic TS  M i   TX i  ic TS  M i   G DL
RSCP TCH – DL = RSCP TCH – DL  ------------------------
 Step SA
L DL   Step 
TX i  ic TS  M i  
TX i  ic TS  M i   P Req
k–1
Where RSCP TCH – DL = -----------------------------------------
-
Model
LT
TX i Mi Mi
Model
Model L Path  L  L  L Body  L Indoor  M Shadowing TX i  ic TS  M i  
LT = ------------------------------------------------------------------------------------------------------------------
- and P Req is the downlink traffic power for iteration k
TX i Mi k–1
G G
- 1 transmitted on the timeslot allocated to Mi.
TX i TX i
Model SA SA
TX i  ic TS  M i   TX i TX i
P Req , if a smart antenna is available in the downlink. Otherwise, G and L are read from the main antenna
model.
The downlink power for mobiles inactive in the downlink accessing circuit- or packet-
switched services are calculated for information only.
9.3.2.2.5 Uplink Signals Update

This step uses the uplink terminal powers calculated for each timeslot allocated to the mobiles. The Dynamic Channel
Allocation allocates timeslots and carriers to all the connected and active mobiles. The Dynamic Channel Allocation is
performed once only, during the first iteration, and the timeslot and carrier allocation remains the same for all the following
iterations of a simulation.
This step updates the received signals for all the mobiles Mi interfered in the uplink by the uplink connection between
interfering mobiles Mj and their best servers TX j  ic  , only if TX j  ic  contain Mi in their coverage areas. TX i  ic  is the
interfered receiver and Mi is the focus, i.e., the mobile that is listened to by the transmitter TX i  ic  .
TX i  ic TS  M i  
For each mobile Mi interfered by Mj in the uplink by the connection between Mj and TX j  ic  , Atoll updates RSCP TCH – UL .
9.3.2.2.6 Downlink Signals Update

For the first iteration, i.e., k = 0, the downlink traffic powers for all the downlink timeslots are set to 0 Watts. Therefore, for
the first iteration, this step is performed for any downlink timeslot for each mobile Mi that is connected and active.
However, for the following iterations, the downlink signals update step uses the actual downlink traffic powers calculated for
each timeslot and the actual timeslots allocated to the mobiles. The Dynamic Channel Allocation allocates timeslots and
carriers to all the connected and active mobiles. The Dynamic Channel Allocation is performed once only during the first
iteration and the timeslot and carrier allocation remains the same for all the following iterations of a simulation.
Therefore, this step is performed for any downlink timeslot for each mobile Mi that is connected and active for the first
iteration, and this step is performed for all the downlink timeslots allocated to the mobile Mi on which it is connected and
active, for the following iterations since the DCA has been performed.
This step updates the received signals for all the mobiles in the TX i  ic  coverage area which are interfered in the downlink
by the connection between TX i  ic  and Mi.
TX i  ic TS  M i  
For each mobile interfered by Mi, Atoll updates RSCP TCH – DL
621
Where TX i  ic  is the transmitter considered and Mi is the focus, i.e., the mobile that is the target for TX i  ic  .
9.3.2.2.7 Control of Radio Resource Limits (Downlink Traffic Power and Uplink Load)
This step checks whether the downlink traffic powers of the downlink timeslots and the uplink loads of the uplink timeslots
of all the cells satisfy the conditions defined globally or per cell and timeslot.
Downlink Power Control:
Atoll verifies that the total R99 power transmitted by any cell on any timeslot does not exceed the effective maximum cell
power per timeslot. The effective maximum cell traffic power per timeslot is calculated as:
P Max – DL – Eff = P Max – DL  %P Max – DL
TX i  ic TS  M i  
Where P Max – DL is the maximum cell power per timeslot defined per cell, and %P Max – DL is the maximum allowed
downlink load either taken from the properties of each cell or from the simulation properties if a global value is defined.
For each transmitter TXi, carrier ic, and downlink timeslot TS M ,
i

P R99 – DL = P TCH – DL + P OCCH
If P R99 – DL  P Max – DL – Eff the mobile with the lowest service priority is rejected for the reason "DL Load Saturation".
Uplink Load Control:
Atoll verifies that the uplink load of any cell on any timeslot does not exceed the maximum uplink cell load allowed per
timeslot.
TX i  ic TS  M i  
The maximum allowed uplink cell load, X Max – UL , is either taken from the properties of each cell or from the simulation
properties if a global value is defined.
For each transmitter TXi, carrier ic, and uplink timeslot TS M ,
i
If X UL  X Max – UL the mobile with the lowest service priority is rejected for the reason "UL Load Saturation".
The uplink load is given by:

TX i  ic TS  M i  
TX i  ic TS  M i   N Tot – UL
X UL = -----------------------------------------------
- if no smart antenna is used by the transmitter in the uplink.
N Tot – UL + N0
If a smart antenna is used by the transmitter in the uplink, the smart antenna gain is taken into account in the calculation of
uplink load.
9.3.2.3 HSDPA Part of the Algorithm

The following calculations are made for all HSDPA mobiles (Mi).
9.3.2.3.1 HSDPA Power Allocation

TX i  ic 
The total transmitted power of the cell ( P Tot – DL ) is the sum of the R99 transmitted power and the HSDPA powers.
P Tot – DL = P R99 – DL + P HR + P HS – SCCH + P HS – PDSCH
The HSDPA powers, i.e., the HS-SCCH and HS-PDSCH powers are calculated as follows:
• HS-SCCH Power:
HS-SCCH channels are transmitted on DL traffic timeslots. The maximum number of supported HS-SCCH channels is
defined per cell. Power can be allocated to HS-SCCH statically or dynamically:
• Static Allocation
The static HS-SCCH power is defined in the properties of the HSDPA cell.
• Dynamic Allocation
622
E TXi  ic  TX  ic  TX  ic 
HS-SCCH power is calculated for  ----c-
Req i i
= Q HS – SCCH  Mobility  so that P HS – SCCH  P Available – HS – SCCH .
 N t HS – SCCH
i i i i
Where P Available – HS – SCCH = P Max – DL – Eff – P R99 – DL – P HR is the power available for HS-SCCH in the cell
i i i
TX i  ic  , and P R99 – DL = P TCH – DL + P OCCH .
The effective maximum cell traffic power per timeslot is calculated as: P Max – DL – Eff = P Max – DL  %P Max – DL .
TX  ic 
i
P Max – DL is the maximum power defined per cell, and %P Max – DL is the maximum allowed downlink load either
taken from the properties of each cell or from the simulation properties if a global value is defined.
Ec
 ----
i TX
 N –     RSCP HS – SCCH
i i i
-
TX i  ic   N t HS – SCCH  Tot – DL  Model
P HS – SCCH = -----------------------------------------------------------------------------------------------------------------------------
TX i
 LT

TX i  ic 
Where N Tot – DL is the downlink total noise calculated in "Downlink Power Control" on page 619,
TX i TX i Mi 
Ortho
 =    1 – F DL
 1 Total Noise
TX i Mi Mi
Model
Model L Path  L  L  L Body  L Indoor  M Shadowing TX i  ic 
LT = ------------------------------------------------------------------------------------------------------------------
- and P HS – SCCH is the HS-SCCH power calculated for the
TX i Mi
G G
timeslots allocated to Mi.
TX TX
Model i SA i SA
In L T ,G = G DL and L = L DL are calculated according to the smart antenna modelling method used, for
P HS – SCCH , if a smart antenna is available in the downlink. Otherwise, G and L are read from the main
antenna model.
• HS-PDSCH Power:
HS-PDSCH channels are transmitted on DL traffic timeslots. Power can be allocated to HS-PDSCH statically or
dynamically:
The static HS-PDSCH power is defined in the properties of the HSDPA cell.
HS-PDSCH power is calculated as follows:
i i i i i
P HS – PDSCH = P Max – DL – Eff – P R99 – DL – P HR – P HS – SCCH
Where P R99 – DL = P TCH – DL + P OCCH . The effective maximum cell traffic power per timeslot is calculated as:
P Max – DL – Eff = P Max – DL  %P Max – DL . P Max – DL is the maximum power defined per cell, and %P Max – DL is the
maximum allowed downlink load either taken from the properties of each cell or from the simulation properties
if a global value is defined.
The HS-SICH power is calculated as follows:
• HS-SICH Power:
HS-SICH channels can be transmitted on any UL traffic timeslot. The maximum number of supported HS-SICH channels
is defined per cell. Power can be allocated to HS-SICH statically or dynamically:
The static HS-SICH power is defined in the properties of the terminal used by the HSDPA mobile Mi.
623
E TXi  ic  M TX  ic 
HS-SICH power is calculated for  ----c-
Req i i
= Q HS – SICH  Mobility  so that P HS – SICH  P Max – HS – SICH and
 N t HS – SICH
M M
i i
P HS – SICH  P Max – HS – SICH .
Ec 
 ----
i M M
 N –     RSCP HS – SICH
i i i
-
Mi  N t HS – SICH  Tot – UL 
P HS – SICH -  L Model
= ------------------------------------------------------------------------------------------------------------------------- T
Mi

TX  ic 
i
Where N Tot – UL is the uplink total noise calculated in "Uplink Power Control" on page 617,
M M TX 
i i Ortho i
 =    1 – F UL
 1 Total Noise
TX i Mi Mi
Model
Model L Path  L  L  L Body  L Indoor  M Shadowing M
i
LT = ------------------------------------------------------------------------------------------------------------------
- and P HS – SICH is the HS-SICH power calculated for the
TX i Mi
G G
timeslots allocated to Mi.
TX i TX i
Model SA SA
In L T ,G = G UL and L = L UL are calculated according to the smart antenna modelling method used,
Mi TX i TX i
for P HS – SICH , if a smart antenna is available in the uplink. Otherwise, G and L are read from the main
antenna model.
TX i  ic 
The total transmitted power of the cell ( P Tot – DL ) is the sum of the R99 transmitted power and the HSDPA powers.
P Tot – DL = P R99 – DL + P HR + P HS – SCCH + P HS – PDSCH
9.3.2.3.2 Connection Status and Number of HSDPA Users

HSDPA users cannot receive HS-SCCH and HS-PDSCH powers simultaneously. HS-PDSCH arrives 3 timeslots after the HS-SCCH.
HS-SICH is 9 timeslots after the HS-PDSCH. Atoll assumes that an active HSDPA user has the same probability of receiving HS-
SCCH and HS-PDSCH, and transmitting HS-SICH because their occurrence is equally likely. Therefore, each HSDPA user is
assigned a sub-connection status randomly. The sub-connection status can be:
• HS-SCCH: HSDPA mobile that is receiving HS-SCCH power
• HS-PDSCH: HSDPA mobile that is receiving traffic power
• HS-SICH: HSDPA mobile that is transmitting HS-SICH power
The number of active HSDPA users belonging to each sub-connection status is 1/3rd of the total number of active HSDPA
users.
n HS – SCCH is the maximum number of HS-SCCH channels and n HS – SICH is the maximum number of HS-SICH channels that the
cell can manage. Each HSDPA user consumes one HS-SCCH and HS-SICH channels. Therefore, at a given instance, the number
of connected HSDPA users cannot exceed the number of HS-SCCH and HS-SICH channels per cell. The maximum number of
HSDPA users ( n Max ) corresponds to the maximum number of HSDPA users that the cell can support.
9.3.2.3.3 HSDPA Admission Control

HS-SCCH
HS-SCCH admission control is performed for active HSDPA users connected to A-DCH bearers on the downlink and having an
HS-SCCH sub-connection status. Each cell is able to manage a maximum number of HS-SCCH channels, n HS – SCCH . During the
R99 part, the DCA provides a DL timeslot with one SF16 resource unit that has the downlink Ec/Nt higher than the required
quality. If no cell with such a resource unit is available, the user is rejected.
HS-SICH
HS-SICH admission control is performed for active HSDPA users connected to A-DCH bearers on the uplink and having an HS-
SICH sub-connection status. Each cell is able to manage a maximum number of HS-SICH channels, n HS – SICH . During the R99
part, the DCA provides an UL timeslot with one SF16 resource unit that has the uplink Ec/Nt higher than the required quality.
If no cell with such a resource unit is available, the user is rejected.
624
HS-PDSCH
Scheduling is performed for active HSDPA users connected to A-DCH bearers on the downlink and having an HS-PDSCH sub-
connection status. The scheduling is performed as follows:
1. Each HS-PDSCH user is considered as the only served user. The scheduler allocates the best available HSDPA bearer to
each user. The best available HSDPA bearer is selected depending on the user’s Ec/Nt. If no bearer can be allocated
due to low Ec/Nt, the user is rejected for the reason "HSDPA Scheduler Saturation".
The required HS-PDSCH Ec/Nt value is read from receiver equipment properties. For each bearer, Atoll checks that the
Ec/Nt reaches the quality target. HS-PDSCH Ec/Nt is calculated by taking into account all intra and extra cells
interferences.
2. The scheduler sorts the HS-PDSCH users to whom bearers have been assigned in the order of decreasing RLC peak
throughputs. If two users have the same bearer, the user with the higher Ec/Nt has the higher rank.
3. The scheduler considers the group of HS-PDSCH users to whom bearers, HS-SCCH, and HS-SICH have been assigned.
The number of HS-PDSCH users cannot exceed the maximum number of HSDPA users ( n Max ) supported by the cell. If
there are enough HSDPA power and resource units available in order to obtain a HSDPA bearer, the users will be
connected. Otherwise, they will be delayed and their connection status will be “HSDPA Delayed”.
4. Other HS-PDSCH users will be rejected for the reason "HSDPA Scheduler Saturation".
For N-frequency mode compatible transmitters, the resource units available in the master and slave carriers can be shared,
i.e., a mobile can be connected to timeslots belonging more than one carrier.
9.3.2.3.4 HSDPA Dynamic Channel Allocation

For each mobile connected to the A-DPCH bearer:
1. Atoll selects the HSDPA bearers that match to the mobile terminal and UE category parameters.
2. For each bearer supported by a mobile:
a. The scheduler searches for the best collection of "n" ordered timeslots that can provide enough resource units to
support the service, and whose Ec/Nt is better than the minimum required and enough to reach the bearer’s
resource unit requirements. The best is determined by applying the R99 Dynamic Channel Allocation algorithm.
b. The scheduler calculates the HS-PDSCH Ec/Nt for each timeslot of the best collection. The Ec/Nt value associated
with the mobile-bearer pair is the worst one of all selected timeslots.
c. If the scheduler is unable to find a satisfactory timeslot collection, the bearer is removed from the list of supported
bearers.
3. The mobile is connected to the supported bearer having the highest RLC peak throughput. If two bearers have the
same RLC peak throughput, the best one is the one with the highest Ec/Nt.
9.3.2.3.5 Ressource Unit Saturation

For each time slot, a minimum and maximum number of resource units for HSDPA users are defined in the cell properties.
Atoll dynamically allocates the required number of codes respecting these limitations. The minimum number of HSDPA codes
is excluded from the set of codes available for R99 users. The scheduler checks if enough codes are available for the selected
HSDPA bearer (taking into account the maximum number of HSDPA codes). If not, the scheduler allocates a lower HSDPA
bearer which needs fewer codes. If there are no more resource units available for the lowest HSDPA bearer, the user will be
delayed or rejected.
9.3.2.4 Convergence Criteria

The convergence criteria are evaluated for each iteration and can be written as follows:
 Max  TXi  ic TS  M i   

 DL = Int  P Err  100
 All TX i   
 TX i  ic TS  M i   TX i  ic TS  M i   
 Max N Tot – UL – N Tot – UL 
 UL = Int  -------------------------------------------------------------------------------------  100
k k – 1
 All TX i TX i  ic  TS  M i   
 N Tot – UL
k

TX  ic TS  M  
i i
Where, P Err is given by:
625
TX  ic TS  M   TX  ic TS  M  
i i i i
TX  ic TS  M   P Rec – P Rec
i i Max  k  k–1
P Err = Step Step
-------------------------------------------------------------------------------------------------------------
- with smart antennas.
0   Step  360 TX  ic TS  M  
i i
P Rec
 k
Step
TX  ic TS  M   TX  ic TS  M  
i i i i
TX  ic TS  M  
i i
P Rec – P Rec
k – 1 without smart antennas.
= ----------------------------------------------------------------------------------
k
P Err -
TX i  ic TS  M i  
P Rec
k
Atoll stops the simulations in the following cases:

• Convergence: Between two successive iterations,  DL and  UL are less than or equal to their respective thresholds
(defined when creating a simulation).
Example: Let us assume that the maximum number of iterations is 100, and the UL and DL convergence thresholds
are set to 5 %. If  DL  5 and  UL  5 between the 4th and the 5th iteration, Atoll stops the algorithm after the 5th
iteration. The simulation has converged.
• Divergence: After 30 iterations,  DL and/or  UL are still higher than their respective thresholds and from the 30th
iteration,  DL and/or  UL do not decrease during the next 15 successive iterations.
Examples: Let us assume that the maximum number of iterations is 100, and the UL and DL convergence thresholds
are set to 5 %.
a. After the 30th iteration,  DL and/or  DL equal 100 and do not decrease during the next 15 successive iterations.
Atoll stops the algorithm at the 46th iteration. The simulation has not converged.
b. After the 30th iteration,  DL and/or  UL equal 80, they start decreasing slowly until the 40th iteration (without
going under the thresholds) and then, do not change during 15 successive iterations. Atoll stops the algorithm at
the 56th iteration without converging.
• Last Iteration: If  DL and/or  UL are still much higher than their respective thresholds after the last iteration, the
simulation has not converged. If  DL and  UL are lower than their respective thresholds, the simulation has reached
convergence.
9.4 TD-SCDMA Prediction Studies

For each TBC transmitter, TXi, Atoll determines the value of the selected parameter on each studied pixel inside the TXi
calculation area. Each pixel within the TXi calculation area is considered a probe receiver.
• The study conditions to determine the service area of each TBC transmitter
• The display settings to for colouring the covered pixels
Atoll uses the parameters entered in the Condition tab of the coverage study properties dialogue to determine pixels covered
by the each transmitter. Coverage prediction display resolution is independent of the path loss matrix and geographic data
resolutions, and can be different for each coverage prediction. Coverage predictions are calculated using bilinear interpolation
of multi-resolution path loss matrices (similar to the evaluation of site altitudes).
9.4.1 P-CCPCH Reception Analysis (Eb/Nt) or (C/I)

E
These coverage predictions calculate and display the Eb/Nt or C/I on the P-CCPCH,  ----b- or  ---
C
. The
 N t P – CCPCH  I  P – CCPCH
coverage predictions are calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for TS0. The best
servers for the coverage predictions are determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH
power, or from the master carrier in case of N-frequency mode compatible transmitters. Afterwards, the coverage predictions
are calculated for the selected carrier. If the selected carrier does not exist on a transmitter, there will not be any pixels
covered by this transmitter. If you perform these coverage predictions for the best carrier, Atoll calculates the Eb/Nt or C/I
considering:
626
TX  ic 
i Min
The pixels in the TX i  ic  coverage area where RSCP P – CCPCH  Max (TAdd P – CCPCH,RSCP P – CCPCH) , and
E b
 ----
i
C
i
Req
--- Req
 N t P – CCPCH  Q P – CCPCH or  I  P – CCPCH  Q P – CCPCH are covered and coloured according to the selected display option.
-
TX TX  ic  TX TX  ic 
E TX i  ic  TX i  ic 
i i i i
  RSCP P – CCPCH   RSCP P – CCPCH
Where  ----b- C
---
Proc
= ------------------------------------------------
TX i  ic 
 G P – CCPCH and  I  P – CCPCH = ------------------------------------------------
N t P – CCPCH TX i  ic 
TX i  ic 
TX i  ic  P P – CCPCH
RSCP P – CCPCH = ----------------------
LT
The downlink total noise is calculated as follows:

Term
N Tot – DL = I Intra – DL + I Extra – DL + I IC – DL  ic jc  + N 0
Where
TX i  ic  TX i  ic  TX i TX i  ic  TX i
I Intra – DL = RSCP P – CCPCH     + RSCP OCCH – TS0  
TX i TX i 
 and  =  0 Without Useful Signal
Ortho Term
With  =    1 – F DL    1 – F JD
 1 Total Noise
i j j
I Extra – DL =
ji
TX
I IC – DL  ic jc  = ---------------------------------------------------------------------------------------
j -
TX i  ic 
TX i  ic  P OCCH – TS0
RSCP OCCH – TS0 = ------------------------
-
LT
TX i Term Eb  Nt
L Path  L  L  L Body  L Indoor  M Shadowing
L T = -----------------------------------------------------------------------------------------------------------------------
-
TX i
Term
G G
TX i
Term
 and N 0 are defined in "Definitions and Formulas" on page 595.
Coverage Display
• Eb/Nt or C/I (dB)
Atoll calculates the Eb/Nt or C/I on each pixel of the TX i  ic  best server coverage area. The pixel colour depends on
the Eb/Nt or C/I level. Coverage consists of several independent layers whose visibility in the workspace can be
managed. There are as many layers as thresholds defined in the Display tab (Prediction properties). For each layer, a
Eb i C i
pixel is covered if  -----  Threshold or  ---  Threshold . Each layer is assigned a colour and
• Eb/Nt Margin or C/I Margin (dB)
Atoll calculates the Eb/Nt or C/I margin on each pixel of the TX i  ic  best server coverage area. The pixel colour
depends on the Eb/Nt or C/I margin value. Coverage consists of several independent layers whose visibility in the
workspace can be managed. There are as many layers as thresholds defined in the Display tab (Prediction properties).
627
E TXi  ic  C i
TX  ic 
For each layer, a pixel is covered if  ----b- – Q P – CCPCH  M P – CCPCH or  ---
Req Eb  Nt Req CI
– Q P – CCPCH  M P – CCPCH .
Atoll calculates the cell edge coverage probability on each pixel of the TX i  ic  best server coverage area. The pixel
colour depends on the cell edge coverage probability value. Coverage consists of several independent layers whose
visibility in the workspace can be managed. There are as many layers as thresholds defined in the Display tab
TX i  ic 
E b
 ---- Req
(Prediction properties). For each layer, a pixel is covered if -  Q P – CCPCH or
 N t P – CCPCH
CECP
TX i  ic 
C
--- Req
 Q P – CCPCH . Each layer is assigned a colour and displayed with intersections between layers.
 I  P – CCPCH
CECP
9.4.2 DwPCH Reception Analysis (C/I)

This coverage prediction calculates and displays the C/I on the DwPCH,  ---
C
. The coverage prediction is calculated for a
 I  DwPCH
given set of a terminal type, a mobility type, a service, a carrier, and for DwPTS. The best server for the coverage prediction is
determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power, or from the master carrier in
case of N-frequency mode compatible transmitters. Afterwards, the coverage prediction is calculated for the selected carrier.
If the selected carrier does not exist on a transmitter, there will not be any pixels covered by this transmitter. If you perform
this coverage prediction for the best carrier, Atoll calculates the C/I considering:
The pixels in the TX i  ic  coverage area where RSCP P – CCPCH  Max (TAdd P – CCPCH,RSCP P – CCPCH) and  ---
i Min C Req
 Q DwPCH
I DwPCH
are covered and coloured according to the selected display option.
TX TX  ic 
TX  ic  i i
C i   RSCP DwPCH
Where  --- = ------------------------------------------
-
 I  DwPCH TX  ic 
i
N Tot – DL
TX  ic 
TX i  ic  i
P DwPCH
RSCP DwPCH = ----------------
-
LT
The downlink total noise is calculated as follows:

Term
Where
TX i  ic  TX i  ic  TX i
I Intra – DL = RSCP DwPCH    
TX i TX i 
 and  =  0 Without Useful Signal
Ortho Term
With  =    1 – F DL    1 – F JD
 1 Total Noise
TX i  ic  TX j  ic 
I Extra – DL =   RSCPDwPCH
ji
TX  jc 
  RSCPDwPCH
j
TX
I IC – DL  ic jc  = ----------------------------------------
j -
628
TX
i Term Eb  Nt
L Path  L  L  L Body  L Indoor  M Shadowing
L T = -----------------------------------------------------------------------------------------------------------------------
-
TX
i Term
G G
TX
i Term
 and N 0 are defined in "Definitions and Formulas" on page 595.
Coverage Display
• C/I (dB)
Atoll calculates the C/I on each pixel of the TX i  ic  best server coverage area. The pixel colour depends on the C/I
level. Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as
many layers as thresholds defined in the Display tab (Prediction properties). For each layer, a pixel is covered if
TX i  ic 
C
---  Threshold . Each layer is assigned a colour and displayed with intersections between layers.
 I  DwPCH
• C/I Margin (dB)

Atoll calculates the C/I margin on each pixel of the TX i  ic  best server coverage area. The pixel colour depends on the
C/I margin value. Coverage consists of several independent layers whose visibility in the workspace can be managed.
There are as many layers as thresholds defined in the Display tab (Prediction properties). For each layer, a pixel is
TX i  ic 
covered if  ---
C Req CI
– Q DwPCH  M DwPCH . Each layer is assigned a colour and displayed with intersections between
 I  DwPCH
layers.
TX  ic 
C i
(Prediction properties). For each layer, a pixel is covered if  ---
Req
 Q DwPCH . Each layer is assigned a colour
I DwPCH
CECP
and displayed with intersections between layers.
9.4.3 Downlink TCH RSCP Coverage

This coverage prediction calculates and displays the RSCP for the downlink traffic channel, RSCP TCH – DL . The coverage
prediction is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for a downlink timeslot. The
considering:
TX i  ic 
Min
The pixels in the TX i  ic  coverage area where RSCP P – CCPCH  Max (TAdd P – CCPCH,RSCP P – CCPCH) and
TX  ic 
i Req
RSCP TCH – DL  RSCP TCH – DL  Service Mobility  are covered and coloured according to the selected display option.
TX  ic 
i
Where RSCP TCH – DL is given by:
Max
TX  ic 
i P TCH – DL  Service 
RSCP TCH – DL = -------------------------------------------
-
Model
LT
629
TX
i Term Model
Model L Path  L  L  L Body  L Indoor  M Shadowing Max
LT - and P TCH – DL  Service  is the maximum downlink traffic power
= -----------------------------------------------------------------------------------------------------------------------
TX
i Term
G G
defined for the selected service.
TX TX
Model i SA i SA
TX i TX i
Max
P TCH – DL  Service  , if a smart antenna is available in the downlink. Otherwise, G and L are read from the main antenna
model.
Coverage Display
• DL TCH RSCP (dBm)
Atoll calculates the DL TCH RSCP on each pixel of the TX i  ic  best server coverage area. The pixel colour depends on
the RSCP level. Coverage consists of several independent layers whose visibility in the workspace can be managed.
TX i  ic 
covered if RSCP TCH – DL  Threshold . Each layer is assigned a colour and displayed with intersections between layers.

Atoll calculates the RSCP margin on each pixel of the TX i  ic  best server coverage area. The pixel colour depends on
the RSCP margin value. Coverage consists of several independent layers whose visibility in the workspace can be
TX i  ic 
Req RSCP
pixel is covered if RSCP TCH – DL – RSCP TCH – DL  Service Mobility   M TCH – DL . Each layer is assigned a colour and
TX i  ic 
Req
(Prediction properties). For each layer, a pixel is covered if RSCP TCH – DL  RSCP TCH – DL  Service Mobility  .
CECP
9.4.4 Uplink TCH RSCP Coverage

This coverage prediction calculates and displays the RSCP for the uplink traffic channel, RSCP TCH – UL . The coverage prediction
is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for an uplink timeslot. The best server
TX i  ic  Min
The pixels in the TX i  ic  coverage area where RSCP P – CCPCH  Max (TAdd P – CCPCH,RSCP P – CCPCH) and
TX i  ic 
Req
RSCP TCH – UL  RSCP TCH – UL  Service Mobility  are covered and coloured according to the selected display option.
TX i  ic 
Where RSCP TCH – UL is given by:
TX i  ic  Term
P Max
RSCP TCH – UL = --------------
-
Model
LT
630
TX
i Term Model
Model L Path  L  L  L Body  L Indoor  M Shadowing Term
LT = -----------------------------------------------------------------------------------------------------------------------
- and P Max is the maximum uplink traffic power defined for the
TX
i Term
G G
selected terminal.
TX TX
Model i SA i SA Term
In L T ,G = G UL and L = L UL are calculated according to the smart antenna modelling method used, for P Max , if
TX i TX i
a smart antenna is available in the uplink. Otherwise, G and L are read from the main antenna model.
Coverage Display
• UL TCH RSCP (dBm)
Atoll calculates the UL TCH RSCP on each pixel of the TX i  ic  best server coverage area. The pixel colour depends on
the RSCP level. Coverage consists of several independent layers whose visibility in the workspace can be managed.
TX i  ic 
covered if RSCP TCH – UL  Threshold . Each layer is assigned a colour and displayed with intersections between layers.

Atoll calculates the RSCP margin on each pixel of the TX i  ic  best server coverage area. The pixel colour depends on
the RSCP margin value. Coverage consists of several independent layers whose visibility in the workspace can be
TX  ic 
i Req RSCP
pixel is covered if RSCP TCH – UL – RSCP TCH – UL  Service Mobility   M TCH – UL . Each layer is assigned a colour and
TX i  ic 
Req
(Prediction properties). For each layer, a pixel is covered if RSCP TCH – UL  RSCP TCH – UL  Service Mobility  .
CECP
9.4.5 Downlink Total Noise

This coverage prediction calculates and displays the total noise on the downlink, N Tot – DL . The coverage prediction is
calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for a downlink timeslot. The best server
by this transmitter. If you perform this coverage prediction for the best carrier, Atoll calculates the downlink noise for all the
carriers but keeps the worst case value, i.e., the most interfered carrier. You can choose to display the minimum, the
maximum, or the average total noise values from among the values calculated for all the carriers. Pixels are covered and
coloured according to the total downlink noise thresholds defined in the display options.

Term
Total downlink noise is given by: N Tot – DL =  RSCP TCH – DL + RSCP OCCH  + N 0
All TX, c, and TS
P TCH – DL P OCCH
With RSCP TCH – DL = -------------------
- and RSCP OCCH = --------------
-
Model Model
LT LT
TX
i Term Model
Model L Path  L  L  L Body  L Indoor  M Shadowing
LT = -----------------------------------------------------------------------------------------------------------------------
- and P TCH – DL and P TCH – DL are respectively the downlink traffic
TX
i Term
G G
power and the other common control channel power for the selected timeslot.
631
TX TX
Model i SA i SA
TX TX
Max i i
model.
9.4.6 Downlink Service Area Analysis (Eb/Nt) or (C/I)

E
These coverage predictions calculate and display the Eb/Nt or C/I on the downlink traffic channel,  ----b- or  ---
C
.
N t TCH – DL I TCH – DL
The coverage predictions are calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for a
downlink timeslot. The best servers for the coverage predictions are determined according to the P-CCPCH RSCP from the
carrier with the highest P-CCPCH power, or from the master carrier in case of N-frequency mode compatible transmitters.
Afterwards, the coverage predictions are calculated for the selected carrier. If the selected carrier does not exist on a
transmitter, there will not be any pixels covered by this transmitter. If you perform these coverage predictions for the best
carrier, Atoll calculates the Eb/Nt or C/I considering:
TX i  ic 
Min
The pixels in the TX i  ic  coverage area where RSCP P – CCPCH  Max (TAdd P – CCPCH,RSCP P – CCPCH) ,
TX  ic  TX i  ic  TX i  ic 
Eb
RSCP TCH – DL  RSCP TCH – DL  Service Mobility  , and  -----  Q TCH – DL or  ---
i Req Req C Req
 Q TCH – DL are covered and
coloured according to the selected display option.
TX i TX i  ic  TX i TX i  ic 
E b TXi  ic    RSCP TCH – DL TX  ic 
i   RSCP TCH – DL
Where  ----- -  G Proc  C
Div
= --------------------------------------------- DL  G DL and  --- -  G Div
= ---------------------------------------------
N t TCH – DL TX  ic  I TCH – DL TX i  ic  DL
i
Max
TX i  ic  P TCH – DL  Service 
With RSCP TCH – DL = -------------------------------------------
-
 Eb  Nt  DL
LT
TX  Eb  Nt 
i Term DL
L Path  L  L
 Eb  Nt  DL  L Body  L Indoor  M Shadowing Max
LT - and P TCH – DL  Service  is the maximum downlink traffic
= -----------------------------------------------------------------------------------------------------------------------
TX
i Term
G G
power defined for the selected service.
 Eb  Nt  TX TX
DL i SA i SA
Max TX i TX i
model.
i i i Term
Where
TX i  ic   TXi TX i  TX i  ic  TX i  ic 
I Intra – DL =     1 – F DL    1 – F JD  +  1 –      RSCP TCH – DL + RSCP OCCH 
Ortho Term
   
 
TX  ic 
TX i  ic  i
P OCCH
With RSCP OCCH = ----------------------
-
 Eb  Nt 
DL
LT
TX i  ic  TX j  ic  TX j  ic 
I Extra – DL =   RSCPTCH – DL + RSCPOCCH 
ji
TX
I IC – DL  ic jc  = ---------------------------------------------------------------------------
j -
632
Coverage Display
• Max Eb/Nt or Max C/I (dB)
E b TXi  ic  C i
TX  ic 
pixel is covered if  -----  Threshold or  ---  Threshold . Each layer is assigned a colour and displayed
• Effective Eb/Nt or Effective C/I (dB)
Atoll calculates the effective Eb/Nt or C/I on each pixel of the TX i  ic  best server coverage area. The pixel colour
depends on the effective Eb/Nt or C/I level. Coverage consists of several independent layers whose visibility in the
 E TXi  ic  
Min   ----b-
Req
For each layer, a pixel is covered if  Q TCH – DL  Threshold or
 N t TCH – DL 
 C TXi  ic  
Min   ---
Req
 Q TCH – DL  Threshold . Each layer is assigned a colour and displayed with intersections between
 I TCH – DL 
layers.
E b TX i  ic  C i
TX  ic 
For each layer, a pixel is covered if  ----- – Q TCH – DL  M TCH – DL or  ---
– Q TCH – DL  M TCH – DL . Each
• Required Power (dBm)
Atoll calculates the downlink required power on each pixel of the TX i  ic  best server coverage area. The pixel colour
depends on the required power level. Coverage consists of several independent layers whose visibility in the
Req
Req Req Q TCH – DL
-  P Max
For each layer, a pixel is covered if P TCH – DL  Threshold , where P TCH – DL = --------------------------- TCH – DL  Service  or
TX  ic 
E b
 ----
i
-
Req
Req Q TCH – DL Max
P TCH – DL = -------------------------
TX i  ic 
 P TCH – DL  Service  . Each layer is assigned a colour and displayed with intersections
C ---
between layers.
• Required Power Margin (dB)
Atoll calculates the downlink required power margin on each pixel of the TX i  ic  best server coverage area. The pixel
colour depends on the required power margin value. Coverage consists of several independent layers whose visibility
in the workspace can be managed. There are as many layers as thresholds defined in the Display tab (Prediction
Req Max
properties). For each layer, a pixel is covered if P TCH – DL – P TCH – DL  Service   M arg in , where
Req Req
Req Q TCH – DL Max Req Q TCH – DL Max
P TCH – DL = ---------------------------
TX i  ic 
-  P TCH – DL  Service  or P TCH – DL = -------------------------
TX i  ic 
 P TCH – DL  Service  . Each layer is assigned
 E b
----- C
---
633

TX i  ic 
E b
 ---- Req
(Prediction properties). For each layer, a pixel is covered if -  Q TCH – DL or
CECP
TX i  ic 
C
--- Req
 Q TCH – DL . Each layer is assigned a colour and displayed with intersections between layers.
CECP
9.4.7 Uplink Service Area Analysis (Eb/Nt) or (C/I)

E
These coverage predictions calculate and display the Eb/Nt or C/I on the uplink traffic channel,  ----b- or  ---
C
.
N t TCH – UL I TCH – UL
The coverage predictions are calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for an uplink
timeslot. The best servers for the coverage predictions are determined according to the P-CCPCH RSCP from the carrier with
the highest P-CCPCH power, or from the master carrier in case of N-frequency mode compatible transmitters. Afterwards, the
coverage predictions are calculated for the selected carrier. If the selected carrier does not exist on a transmitter, there will
not be any pixels covered by this transmitter. If you perform these coverage predictions for the best carrier, Atoll calculates
the Eb/Nt or C/I considering:
TX i  ic 
Min
The pixels in the TX i  ic  coverage area where RSCP P – CCPCH  Max (TAdd P – CCPCH,RSCP P – CCPCH) ,
TX i  ic  E b TXi  ic  C i
TX  ic 
RSCP TCH – UL  RSCP TCH – UL  Service Mobility  , and  -----  Q TCH – UL or  ---
Req Req Req
 Q TCH – UL are covered and
N t TCH – UL I TCH – UL
coloured according to the selected display option.
E b TXi  ic  TX  ic 
Term Term
  RSCP TCH – UL i   RSCP TCH – UL
Where  ----- -  G Proc C
---
Div
= -------------------------------------------------  -  G Div
-------------------------------------------------
N t TCH – UL TX i  ic  UL G UL and  I  TCH – UL = TX i  ic  UL
N Tot – UL N Tot – UL
Term Req Req

TX i  ic  P Max Q TCH – UL Q TCH – UL
Term Term
With RSCP TCH – UL = ---------------------- - or P Term
- and P Req = P Max  ---------------------------
Term
Req = P Max  -------------------------
-
 Eb  Nt UL TX i  ic  TX i  ic 
LT E
 -----
b  ---
C
TX i Term  Eb  Nt  UL
 Eb  Nt  UL L Path  L  L  L Body  L Indoor  M Shadowing Term
LT = -----------------------------------------------------------------------------------------------------------------------
- and P Max is the maximum power defined for the selected
TX i
Term
G G
terminal.
 Eb  Nt  UL TX i TX i
SA SA Term
In L T ,G = G UL and L = L UL are calculated according to the smart antenna modelling method used, for P Max ,
TX i TX i
if a smart antenna is available in the uplink. Otherwise, G and L are read from the main antenna model.
Coverage Display
• Max Eb/Nt or Max C/I (dB)
Eb
pixel is covered if  -----  Threshold or  ---
C
 Threshold . Each layer is assigned a colour and displayed
• Effective Eb/Nt or Effective C/I (dB)
634
Atoll calculates the effective Eb/Nt or C/I on each pixel of the TX i  ic  best server coverage area. The pixel colour
depends on the effective Eb/Nt or C/I level. Coverage consists of several independent layers whose visibility in the
 E b TXi  ic  
Min   -----
Req
For each layer, a pixel is covered if  Q TCH – UL  Threshold or
 N t TCH – UL 
 C TXi  ic  
Min   ---
Req
 Q TCH – UL  Threshold . Each layer is assigned a colour and displayed with intersections between
 I TCH – UL 
layers.
E TXi  ic  C i
TX  ic 
For each layer, a pixel is covered if  ----b- – Q TCH – UL  M TCH – UL or  ---
– Q TCH – UL  M TCH – UL . Each
• Required Power (dBm)
Atoll calculates the uplink required power on each pixel of the TX i  ic  best server coverage area. The pixel colour
depends on the required power level. Coverage consists of several independent layers whose visibility in the
Req
Term Term Q TCH – UL
For each layer, a pixel is covered if P Req  Threshold , where -  P Term
P Req = --------------------------- or
TX i  ic  Max
E
 ----b-
Req
Term Q TCH – UL
-  P Term
P Req = ------------------------- Max . Each layer is assigned a colour and displayed with intersections between layers.
TX  ic 
i
C ---
• Required Power Margin (dB)

Atoll calculates the uplink required power margin on each pixel of the TX i  ic  best server coverage area. The pixel
colour depends on the required power margin value. Coverage consists of several independent layers whose visibility
in the workspace can be managed. There are as many layers as thresholds defined in the Display tab (Prediction
Req
Term Term Term Q TCH – UL
-  P Term
properties). For each layer, a pixel is covered if P Req – P Max  M arg in , where P Req = --------------------------- Max or
TX i  ic 
E b
 -----
Req
Term Q TCH – UL
-  P Term
P Req = ------------------------- Max . Each layer is assigned a colour and displayed with intersections between layers.
TX  ic 
i
C ---

TX i  ic 
E b
 ---- Req
(Prediction properties). For each layer, a pixel is covered if -  Q TCH – UL or
CECP
TX i  ic 
C
--- Req
 Q TCH – UL . Each layer is assigned a colour and displayed with intersections between layers.
CECP
9.4.8 Effective Service Area Analysis (Eb/Nt) or (C/I)

These coverage predictions consist of pixels covered by the both the uplink and the downlink service areas. These coverage
E
predictions calculate the Eb/Nt or C/I on the downlink and uplink traffic channels,  ----b- or  ---
C
and
635
E b
 ---- or  ---
C
- , and display the pixels where both downlink and uplink Eb/Nt or C/I are above the required quality
thresholds.
The coverage predictions are calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for all the
6 timeslots. The best servers for the coverage predictions are determined according to the P-CCPCH RSCP from the carrier
with the highest P-CCPCH power, or from the master carrier in case of N-frequency mode compatible transmitters.
Afterwards, the coverage predictions are calculated for the selected carrier. If the selected carrier does not exist on a
transmitter, there will not be any pixels covered by this transmitter. If you perform these coverage predictions for the best
carrier, Atoll calculates the Eb/Nt or C/I considering:
The pixels in the TX i  ic  coverage area are covered and coloured according to the selected display option if all the following
conditions are satisfied:
TX i  ic 
Min
• RSCP P – CCPCH  Max (TAdd P – CCPCH,RSCP P – CCPCH)
TX i  ic 
Req
• RSCP TCH – DL  RSCP TCH – DL  Service Mobility 
TX i  ic 
Req
• RSCP TCH – UL  RSCP TCH – UL  Service Mobility 
E b i
 ---- C i
 Q TCH – DL or  ---
Req Req
• -  Q TCH – DL for any of the 6 timeslots
E b
 ----  Q TCH – UL or  ---
Req C Req
• -  Q TCH – UL for any of the 6 timeslots
9.4.9 Cell to Cell Interference

This coverage prediction calculates and displays the interference received by cells receiving in uplink from other cells which
are transmitting in downlink. The timeslot configuration of each cell defines the direction of the link at any given instance.
During each subframe, the direction of the link changes twice (downlink to uplink, and then uplink to downlink). These
transitions are referred to as switching points.
The coverage prediction is calculated for a given set of a terminal type, a mobility type, a service, a carrier, and a timeslot. The
considering:
The mobility, service, and terminal are used to calculate the best server coverage of the interfered cell.
Assuming that a transmitter TX j is interfering a studied transmitter TX i on a timeslot, on the same carrier ic or on another
carrier jc, the cell to cell interference is given by:
j j
  RSCPTCH – DL + RSCPOCCH  + ---------------------------------------------------------------------------
j j TX
I C2C  TX i TX j  = j -
TX j
TX j  ic  TX j  jc 
TX  ic 
j P TCH – DL    TX  jc 
j P TCH – DL   
Where RSCP TCH – DL = ---------------------------
- and RSCP TCH – DL = ---------------------------
- using a smart antenna, and
LT LT
TX j  ic  TX j TX j  jc  TX j
TX  ic  P TCH – DL G Ant TX  jc  P TCH – DL G Ant
j
RSCP TCH – DL -  ---------- and RSCP TCHj – DL = -------------------
= ------------------- -  ---------- otherwise.
LT TX
j LT TX j
L Ant L Ant
636
TX  ic  TX TX  jc  TX
j j j j
TX  ic  P OCCH G Ant TX  jc  P OCCH G Ant
j j
RSCP OCCH = ----------------  ----------
TX
-  ---------- otherwise.
and RSCP OCCH = --------------- TX
LT j LT j
L Ant L Ant
TX TX
ITU526 – 5 j i
L T = L Path  L TX  L RX
ITU526 – 5
L Path is the path loss calculated using the ITU526-5 propagation model without antenna loss.
 is the angle for the smart antenna pattern.

TX j
L Ant is the main antenna attenuation.
TX j
G Ant is the main antenna gain.
Atoll calculates the cell to cell interference on each pixel of the TX i  ic  best server coverage area. The pixel colour depends
on the cell to cell interference level. Coverage consists of several independent layers whose visibility in the workspace can be
managed. There are as many layers as thresholds defined in the Display tab (Prediction properties). For each layer, a pixel is
covered if I C2C  TX i TX j   Threshold . Each layer is assigned a colour and displayed with intersections between layers.
9.4.10 UpPCH Interference

UpPCH is usually carried by the UpPTS timeslot. However, if the interference on UpPTS is high, from unsynchronised DwPTS
or TS0 timeslots of other cells, it is possible to shift the UpPCH to TS1. This is called UpPCH shifting. If some cells in a network
use UpPCH shifting, you can use this coverage prediction to study the interference on the shifted UpPCH of these cells from
other cells. The interference from other cells is in this case generated by the traffic on the TS1 of interfering cells.
This coverage prediction calculates and displays the uplink interference on the TS1, I TS1 – UL . The coverage prediction is
calculated for a given set of a terminal type, a mobility type, a service, a carrier, and for TS1. The best server for the coverage
prediction is determined according to the P-CCPCH RSCP from the carrier with the highest P-CCPCH power, or from the master
carrier in case of N-frequency mode compatible transmitters. Afterwards, the coverage prediction is calculated for the
selected carrier. If the selected carrier does not exist on a transmitter, there will not be any pixels covered by this transmitter.
If you perform this coverage prediction for the best carrier, Atoll calculates the interference for all the carriers but keeps the
worst case value, i.e., the most interfered carrier. You can choose to display the minimum, the maximum, or the average total
noise. The coverage prediction is calculated using the main antenna.
i i
Pixels in the TX i  ic  coverage area where RSCP P – CCPCH  TAdd P – CCPCH  Mobility  and I TS1 – UL  Threshold are covered
and coloured according to the selected display option.
The uplink interference on TS1 is calculated from the uplink load calculated in the simulations or manually defiend for the TS1.
TX i  ic 
TX i  ic  TX i X TS1 – UL
The uplink interference on TS1 is given by: I TS1 – UL = N 0  ----------------------------------
TX  ic 
1 – X i 
 TS1 – UL
9.4.11 HSDPA Predictions

This coverage prediction calculates and displays the peak RLC throughput or the Peak MAC throughput per pixel covered by
HSDPA cells. The coverage prediction is calculated for a given set of an HSDPA terminal type, a mobility type, an HSDPA
service, a carrier, and for all downlink timeslots. The best server for the coverage prediction is determined according to the P-
CCPCH RSCP from the carrier with the highest P-CCPCH power, or from the master carrier in case of N-frequency mode
compatible transmitters. Afterwards, the coverage predictions are calculated for the selected carrier. If the selected carrier
does not exist on a transmitter or if it does not support HSDPA, there will not be any pixels covered by this transmitter. If you
perform these coverage predictions for the best carrier, Atoll calculates the RLC or Peak MAC throughput considering:
The pixels in the TX i  ic  coverage area are covered and coloured if:
TX  ic 
i
• RSCP P – CCPCH  TAdd P – CCPCH  Mobility  ,
637
TX  ic 
E C
 ----
i Req
• -  Q HS – PDSCH , and
 N t HS – PDSCH
TX i  ic 
E C
 ----
• - is enough to select a bearer for the pixels.
For more information on HSDPA bearer selection, see "HSDPA Part of the Algorithm" on page 622.
Coverage Display
It is possible to colour the pixels in the coverage areas by criteria such as:
TX  ic 
i
• Min HS-PDSCH RSCP: On each pixel, Atoll calculates RSCP HS – SCCH for all timeslots and selects the lowest value.
TX i  ic 
• Average HS-PDSCH RSCP: On each pixel, Atoll calculates RSCP HS – SCCH for all timeslots and calculates the average of
these values.
TX i  ic 
• Max HS-PDSCH RSCP: On each pixel, Atoll calculates RSCP HS – SCCH for all timeslots and selects the highest value.
TX i  ic 
E
• Min HS-PDSCH Ec/Nt: On each pixel, Atoll calculates  ----C- for all timeslots and selects the lowest value.
TX i  ic 
E
• Average HS-PDSCH Ec/Nt: On each pixel, Atoll calculates  ----C- for all timeslots and calculates the average
N t HS – PDSCH
of these values.
E TXi  ic 
• Max HS-PDSCH Ec/Nt: On each pixel, Atoll calculates  ----C- for all timeslots and selects the highest value.
• Peak RLC Throughput: After selecting the bearer, Atoll reads the corresponding RLC peak throughput. This is the
highest throughput that the bearer can provide on each pixel. The pixel colour depends on the peak RLC throughput.
Coverage consists of several independent layers whose visibility in the workspace can be managed. There are as many
layers as thresholds defined in the Display tab (Prediction properties). Each layer is assigned a colour and displayed
MAC
• Peak MAC Throughput: Atoll displays the Peak MAC throughput ( R DL ) provided on each pixel. The Peak MAC
throughput is calculated as follows:
MAC
R DL = S Block  500
Where, S Block is the transport block size (in kbits) of the selected HSDPA bearer; it is defined for each HSDPA bearer
in the related table. The value 500 corresponds to the number of blocks per second (there are 4 blocks per TTI and
2000
2000 TTI in one second, i.e ------------ blocks per second).
4
The pixel colour depends on the Peak MAC throughput. Coverage consists of several independent layers whose
(Prediction properties). Each layer is assigned a colour and displayed with intersections between layers.
E b TXi  ic TS 
• Max DL A-DPCH Eb/Nt: Atoll displays the A-DPCH Eb/Nt at the receiver (  ----- ) for the best server and the
 N t TCH – DL – Max
selected timeslot. No power control is performed as in simulations. Here, Atoll determines downlink traffic channel
quality at the receiver for a maximum traffic channel power allowed for the best server.
E b TX i  ic TS 
• Max UL A-DPCH Eb/Nt: Atoll displays the A-DPCH Eb/Nt at the best server (  ----- ) and the selected
 N t TCH – UL – Max
timeslot. No power control is performed as in simulations. Here, Atoll determines uplink traffic channel quality for the
maximum terminal power allowed.
TX i  ic 
• HS-SCCH Power: On each pixel, Atoll calculates P HS – SCCH for the selected timeslot.
TX i  ic 
• HS-SCCH RSCP: On each pixel, Atoll calculates RSCP HS – SCCH for the selected timeslot.
E TX i  ic 
• HS-SCCH Ec/Nt: On each pixel, Atoll calculates  ----c- for the selected timeslot.
N t HS – SCCH
Mi
• HS-SICH Power: On each pixel, Atoll calculates P HS – SICH for the selected timeslot.
638
M
i
• HS-SICH RSCP: On each pixel, Atoll calculates RSCP HS – SICH for the selected timeslot.
M
E i
• HS-SICH Ec/Nt: On each pixel, Atoll calculates  ----c- for the selected timeslot.
N t HS – SICH
TX  ic 
i
• HS-PDSCH RSCP: On each pixel, Atoll calculates RSCP HS – PDSCH for the selected timeslot.
E c TXi  ic 
• HS-PDSCH Ec/Nt: On each pixel, Atoll calculates  ----- for the selected timeslot.
N t HS – PDSCH
9.5 Smart Antenna Modelling

Atoll calculates the smart antenna gains and losses in the direction of a user during the simulations, and in the direction of
each pixel in coverage predictions. During simulations, Atoll determines the gains and losses using the smart antenna models.
In coverage predictions, Atoll determines the gains and losses from the angular distributions calculated during the simulations
for each timeslot and stored in the Cell Parameters per Timeslot table.
If a smart antenna model is only downlink or only uplink, the other direction uses the main antenna gain and losses for
calculations. Therefore,
• If a smart antenna is available on the downlink and uplink:
TX SA TX SA TX SA TX SA
G UL = G UL , L UL = L UL and G DL = G DL , L DL = L DL
• If a smart antenna is available on the downlink only:

TX SA TX SA TX TX TX TX
G DL = G DL , L DL = L DL and G UL = G Ant , L UL = L = L Total – UL
• If a smart antenna is available on the uplink only:

TX SA TX SA TX TX TX TX
G UL = G UL , L UL = L UL and G DL = G Ant , L DL = L = L Total – DL
• If no smart antenna equipment is defined:

TX TX TX TX TX TX TX
G DL = G UL = G Ant , L UL = L = L Total – UL , and L DL = L = L Total – DL
9.5.1 Modelling in Simulations

9.5.1.1 Grid of Beams Modelling
A grid-of-beams smart antenna, called GOB, consists of more than one directional antenna pattern (beam) in different
directions. Each beam of a GOB has a different azimuth so that the GOB as a whole covers an entire sector. During the
simulations, Atoll determines the most suitable beam from the GOB for each user served by the smart antenna. The most
suitable beam (best beam) is the one which provides the highest gain towards the served user:
BeamBest = Beam
H V
Max  G Beam – L Beam – L Beam
H V
Where G Beam , L Beam , and L Beam are the gains, horizontal, and vertical attenuations of the beams of the GOB. In words, the
best beam is the one among all the beams of a GOB that has the highest difference between gain, and horizontal and vertical
SA SA SA SA
attenuations. The gains and losses of the GOB ( G DL , G UL , L DL , and L UL ) are determined from the selected best beam.
The following example shows how Atoll calculates the GOB gains and losses.
Example:
Let us assume a GOB with 5 beams that have the same vertical patterns, and whose horizontal patterns are pointed towards
different directions as shown in the figure below:
639
Figure 9.3: Grid Of Beams Modelling
Let us assume that all the beams and the main antenna have the same 18 dBi gain, and the vertical attenuation at the user
location is 15 dB, which is also the same for all the beams because we assume that the vertical patterns are the same.
If the user is located at  = 70 azimuth, as shown in the figure below, Atoll determines the best beam, which has the
highest gain towards  , as follows:
Horizontal Vertical H V Total Gain

Beam Gain (dBi) G Beam – L Beam – L Beam
Attenuation (dB) Attenuation (dB) (dB)
0° 18 60 15 18 - 60 - 15 -57
30° 18 60 15 18 - 60 - 15 -57
60° 18 2.21 15 18 - 2.21- 15 0.79
-30° 18 60 15 18 - 60 - 15 -57
-60° 18 60 15 18 - 60 - 15 -57
Transmitter

located

azimuth
Figure 9.4: GOB Modelling - Determination of the Best Beam
In our example, the total gain of the beam at 60° is the highest. Therefore this beam is selected as the best beam.
If this beam has been selected in the downlink,
SA SA H V
G DL = 18 dB and L DL = L Beam + L Beam = 17.21 dB
If this beam has been selected in the uplink,

SA SA H V
G UL = 18 dB and L UL = L Beam + L Beam = 17.21 dB
640
9.5.1.2 Adaptive Beam Modelling

An adaptive beam smart antenna is capable of steering a given antenna pattern towards the direction of the served signal. In
Atoll, this is modelled using a single antenna pattern, called a beam because of its highly directional shape. During the
simulations, this adaptive beam is oriented in the direction of each served user in order to model the effect of the smart
antenna.
SA SA SA
The adaptive beam gains ( G DL and G UL ) are the antenna gains defined for the beam, and the adaptive beam losses ( L DL and
SA H V
L UL ) are the horizontal and vertical pattern attenuations L Beam + L Beam towards the user direction.
The following example shows how Atoll calculates the adaptive beam gains and losses.
Example:
Let us assume an adaptive beam smart antenna selected for a transmitter along with a main antenna. Let us assume that the
adaptive beam and the main antenna have the same 18 dBi gain, and the vertical attenuation at the user location is 15 dB.
If the user is located at  = 60 azimuth, as shown in the figure below:
Transmitter

located

azimuth
Figure 9.5: Adaptive Beam Modelling - Determination of the Best Beam
If the adaptive beam smart antenna is selected in the downlink, the gain and losses of the adaptive beam at  are:
SA SA H V
G DL = 18 dB and L DL = L Beam + L Beam = 15 dB
If the adaptive beam smart antenna is selected in the uplink, the gain and losses of the adaptive beam at  are:
SA SA H V
G UL = 18 dB and L UL = L Beam + L Beam = 15 dB
H
In fact, as the ideal beam steering algorithm steers the beam towards the served user, L Beam = 0 . These values are used in
interference calculation to determine the downlink interfering signal due to transmission towards the served user, as well as
for calculating the uplink interfering signals received at transmitter when decoding signal received from the served user.
9.5.1.3 Statistical Modelling

A statistical modelling approach is also available in Atoll which can be used to model the effect of smart antennas through C/
I gains. You can create smart antenna equipment in Atoll based on the statistical approach by providing C/I gains and their
cumulative probabilities for different spreading angles,  Spread .
You can assign a spreading angle to each clutter class in your document. Atoll reads the clutter class in which the served user
is located to determine the spreading angle. Different clutter types have different spreading effects on the propagation of
radio waves. Urban and dense urban clutter types introduce more multipath and spread the signal at a wider angle than an
open or rural clutte type.
Once you have assigned the spreading angles to clutter classes, you can enter the C/I gains and their cumulative probabilities
for each spreading angle, in the smart antenna equipment based on the statistical model. For each smart antenna equipment
SA
based on statistical modelling, you can set a probability threshold, TProb .
To find the smart antenna gain, Atoll determines the clutter class of the served user, it reads the spreading angle from the
clutter class properties, it reads the probability threshold from the smart antenna properties, and reads the smart antenna C/
SA
I gain defined for the Probability = 1 – TProb corresponding to the spreading angle.
The following example shows how Atoll calculates the statistical C/I gains and losses.
641
Example:
Let us assume that the served user is located at a an urban clutter class with  Spread = 10 . The smart antenna equipment
SA SA
has TProb = 80 % . Atoll will read the smart antenna C/I gain G for Prob = 20 % . If a gain for the exact probability
value of 20% is not defined, Atoll linearly interpolates the gain value from the two surrounding values.
SA SA SA
If G Prob = 19%
= 4.6298 dB and G Prob = 20.4%
= 4.7196 dB , then G Prob = 20%
= 4.6941 dB
The smart antenna gains are the same for uplink and downlink. Their are no losses for this type of smart antenna equipment.
Negative values of C/I gains are considered as losses.
9.5.1.4 Beamforming Smart Antenna Models

See "Beamforming Smart Antenna Models" on page 43.
9.5.1.5 3rd Party Smart Antenna Modelling

3rd party smart antenna models can be used in Atoll to determine the gains and losses during the simulations for a given user
distribution generated. The smart antenna gains and losses are used during the simulations and the results are stored in the
Cell Parameters per Timeslot table, which can be used in coverage predictions.
9.5.2 Construction of the Geographic Distributions

During simulations, Atoll uses the smart antenna model selected for each transmitter to calculate the smart antenna gains
and losses. These values are calculated and stored for each user generated for the simulations. Therefore, these values are
calculated and are available for the given locations of the users, i.e., points, only. Atoll uses the Angular Step value that you
set when creating and running simulations to construct the geographic distributions of these results.
Once Atoll has calculated the downlink traffic power and the uplink load using the smart antenna gains and losses determined
as explained in the previous section, at the location of a given user, it calculates the same for points located at the angle equal
to that of the Angular Step of the simulations.
At the end of the simulations, Atoll has a number of points, Angular Step apart, available with the values of these results. The
geographic distribution of these results, i.e., downlink traffic power and uplink loads, is constructed by connecting the
resulting value points.
The following example explains how the geographic distribution of downlink traffic power is created. The geographic
distribution of uplink loads is constructed in the same manner.
Example:
Let us assume a smart antenna equipment using adaptive beam modelling. The angular step defined for the simulations is
 Step = 30 . Therefore, the results are calculated for each point located at regular steps of 30 , i.e., 12 points. The downlink
traffic power at the served user (W) with the adaptive beam pointing in the user’s direction is P W . The downlink traffic
powers, using the same adaptive beam pointed towards the served user, at the 12 other points are also determined.
642
Figure 9.6: Construction of the Geographic Distribution of Downlink Traffic Power
The resulting geographic distribution is formed by linearly joining the obtained results.
Figure 9.7: Geographic Distribution of Downlink Traffic Power
The accuracy of the geographic distribution depends upon the value of the angular step. A radiation pattern created at a 1
step will be much more accurate than one created at 45 , for example. But, the latter will be computed 45 times faster than
the first. The value of the Angular Step should be the best possible compromise between calculation speed and accuracy.
9.5.3 Modelling in Coverage Predictions

The results of Monte Carlo simulations, including the smart antenna results, can be stored in the Cells and in the Cell
Parameters per Timeslot tables, and can be used to carry out coverage predictions. The main results of Monte Carlo
simulations used in coverage predictions are:
• If a smart antenna is used in both uplink and downlink:
UL –  DL – 
Geographic distribution of UL load X and DL traffic power P Traffic
• If a smart antenna is used in downlink only:

DL – 
Geographic distribution of DL traffic power P Traffic
• Without smart antenna:

UL DL
UL load X and DL traffic power P Traffic
643
The uplink load and the downlink traffic power at a given pixel are determined by calculating the angle  of that pixel with
respect to the transmitter azimuth, and reading the uplink load and downlink traffic power from the geographic distribution
results. If an exact value for the angle is not available, the load and power are determined using linear interpolation for the
given angle between two available values.
For example, the figure below shows the distribution of downlink traffic power and uplink traffic load results from a
DL – 315 UL – 315
simulation. For a pixel located at  = 315 , the downlink traffic power P Traffic and the uplink load X are read
DL – 315 UL – 315
from these results. In this example, P Traffic  30 dBm , and X = 2.75 % .
For each pixel, Atoll determines the downlink traffic powers and the uplink loads from all the transmitters.
Figure 9.8: Geographic Distribution of downlink traffic power and uplink load
9.5.4 HSDPA Quality and Throughput Analysis

Fast link adaptation (or Adaptive Modulation and Coding) is used in HSDPA. The power on the HS-DSCH channel is transmitted
at a constant power while the modulation, the coding, and the number of codes are changed to adapt to the radio conditions
variations. Based on the reported channel quality indicator (CQI), the Node-B may change the modulation (QPSK and
optionally 16QAM), the coding, and the number of codes every 2 ms during a communication.
Fast link adaptation is modelled in a dedicated HSDPA coverage prediction. Let us assume each bin on the map corresponds
to a probe receiver with HSDPA capable terminal, mobility, and HSDPA service. The probe receiver on each bin is allocated the
cell’s HSDPA. This receiver may be using a specific carrier or all of them. The probe receiver does not create any interference.
Atoll calculates on each bin either the best pilot quality (P-CCPCH Ec/Nt) or the best HS-PDSCH quality (HS-PDSCH Ec/Nt); this
depends on the option selected in Global parameters (HSDPA part): CQI based on P-CCPCH quality or CQI based on HS-PDSCH
quality (CQI means channel quality indicator). Then, it determines the HS-PDSCH CQI, deduces the best HSDPA bearer that can
be used and selects the suitable bearer so as to comply with cell and terminal user equipment capabilities. Once the bearer
selected, Atoll finds the highest downlink throughput that can be carried at each bin and may deduce the application
throughput. Coverage area is limited by the RSCP P-CCPCH threshold.
The coverage prediction can be calculated for an HSDPA compatible terminal, an HSDPA service, a mobility, a carrier, and a
downlink timeslot. Smart antenna results are taken into account in the computation of this study.
9.5.4.1 Fast Link Adaptation Modelling

As explained above, the way of calculating the dedicated HSDPA study depends on if CQI is based on the P-CCPCH quality or
on the HS-PDSCH quality.
9.5.4.1.1 CQI Based on P-CCPCH Quality

When the option “CQI based on CPICH quality” is selected, Atoll proceeds as follows.
P-CCPCH Quality Calculation

Ec
corresponds to the P-CCPCH quality.
 Nt  P – CCPCH
Two options, available in Global Parameters, may be used to calculate Nt: option Without useful signal or option Total noise.
Therefore, we have:
644
TXi
 -----
Ec-   BTS    RSCP P – CCPCH  ic 
 Nt  ic  P – CCPCH = --------------------------------------------------------------------
DL
N Tot  ic 
And
TXi
 -----  BTS    RSCP P – CCPCH  ic 
 ic 
Ec-
= ------------------------------------------------------------------------------------------------------------
 Nt  P – CCPCH DL TXi
N Tot  ic  –  1 –     BTS  RSCP P – CCPCH  ic 
With
DL DL DL DL Term
N Tot  ic  = I Intra  ic  + I Extra  ic  + I Inter – Carrier  ic jc  + N 0
DL
I Inter – Carrier  ic jc  is the inter-carrier interference from a carrier jc to another carrier ic on the downlink, which is reduced
by the interference reduction factor F IRF  ic jc  defined for the pair (ic, jc).
TXi P P – CCPCH  ic 
RSCP P – CCPCH  ic  = -------------------------------
-
LT
CI
L Path  L TX  L Term  L Body  L Indoor  M Shadowing
L T = ----------------------------------------------------------------------------------------------------------------------
-
G TX  G Term
Term
 BTS ,  and N 0 are defined in "Definitions and Formulas" on page 595.
P-CCPCH CQI Determination
Let us assume the following notation: CQI P – CCPCH corresponds to the P-CCPCH CQI. CQIP – CCPCH is deduced from the table
CQIP – CCPCH = f   ------  ic   . This table is defined for the terminal reception equipment and the specified mobility.
Ec
  Nt  P – CCPCH
HS-PDSCH Quality Calculation
1st step: Atoll calculates the HS-PDSCH power ( P HS – PDSCH ).
P HSDPA  ic  is the power available for HSDPA on the carrier ic. This parameter is a user-defined cell input.
Therefore, we have:
n HS – SCCH is the number of HS-SCCH channels and P HS – SCCH  ic  is the HS-SCCH power on carrier ic. It is either fixed by the
Req
user. P HS – SCCH  ic  is controlled so as to reach the required HS-SCCH Ec/Nt (  ------  ic 
Ec
). It is specified in mobility
 Nt  HS – SCCH
properties.
We have:
TXi
 -----  BTS  RSCP HS – SCCH  ic 
 ic 
Ec-
= ----------------------------------------------------------
N Tot  ic 
And
TXi
 -----  BTS  RSCP HS – SCCH  ic 
 ic 
Ec-
= --------------------------------------------------------------------------------------------------------------------------------------------------------
 Nt  HS – SCCH DL DL Term TXi
N Tot  ic  –  1 – F Ortho    1 – F JD    BTS  RSCP HS – SCCH  ic 
With
DL DL DL DL Term
DL
645
TXi P HS – SCCH  ic 
RSCP HS – SCCH  ic  = -------------------------------
-
LT
and
CI
L T = ----------------------------------------------------------------------------------------------------------------------
-
G TX  G Term
Term Term
 BTS , F Ortho , F JD and N 0 are defined in "Definitions and Formulas" on page 595.
Therefore,
Req
  -----
Ec-
 ic   N Tot  ic 
DL
TXi   Nt  HS – SCCH 
RSCP HS – SCCH  ic  =  -------------------------------------------------------------------  L T for the total noise option,
  BTS 
 
And
Req
  -----
Ec-
 ic 
DL
 N Tot  ic  
TXi   Nt  HS – SCCH 
RSCP HS – SCCH  ic  =  --------------------------------------------------------------------------------------------------------------------------------------------
Req
  L T for the without useful signal option.
    1 +  1 – F DL    1 – F Term    ----- Ec-
 ic   
 BTS  Ortho JD  Nt  HS – SCCH 
2nd step: Then, Atoll calculates the HS-PDSCH quality

Ec
Therefore, we have:
TXi
 -----
Ec-   BTS  RSCP HS – PDSCH  ic 
 Nt  ic  HS – PDSCH
= -------------------------------------------------------------
DL
N Tot  ic 
And
TXi
 -----  BTS  RSCP HS – PDSCH  ic 
 ic 
Ec-
= -----------------------------------------------------------------------------------------------------------------------------------------------------------
 Nt  HS – PDSCH TXi
DL DL Term RSCP HS – PDSCH  ic 
N Tot  ic  –  1 – F Ortho    1 – F JD    BTS  ---------------------------------------------
n
Here, Atoll works on the assumption that five HS-PDSCH channels are used (n=5).
With
DL DL DL DL Term
DL
TXi P HS – PDSCH  ic 
RSCP HS – PDSCH  ic  = ----------------------------------
-
LT
And
CI
L T = ----------------------------------------------------------------------------------------------------------------------
-
G TX  G Term
Term Term
HS-PDSCH CQI Determination
The best bearer that can be used depends on the HS-PDSCH CQI. Let us assume the following notation: CQI HS – PDSCH
corresponds to the HS-PDSCH CQI. Atoll deduces CQI HS – PDSCH as follows:
CQI HS – PDSCH = CQI P – CCPCH – P P – CCPCH + P HS – PDSCH
646
Bearer Selection
Knowing the HS-PDSCH CQI, Atoll finds the best bearer that can be used in the table Best Bearer=f(HS-PDSCH CQI). This table
is defined for the terminal reception equipment and the specified mobility.
Then, Atoll checks if best bearer characteristics are compliant with cell and user equipment category capabilities. Atoll selects
the bearer which is the best bearer compliant with the cell and UE category capabilities.
Bearer characteristics are provided in the HSDPA Bearer table. Assuming the best bearer = 23. Characteristics of this bearer
are:
• Transport block size: 9719 Bytes
• 16QAM modulation used: Yes
• Peak Throughput: 4.48 Mb/s
Figure 9.9: Radio Bearers Table
Assuming user equipment category = 3. Its capabilities are:

• Maximum transport block size: 7298 Bytes
• Maximum number of HS-PDSCH channels used: 5
• 16QAM modulation used: Yes
• Minimum number of TTI between two TTI used: 2
Figure 9.10: UE Categories Table
HSDPA cell capabilities are:

• Maximum number of HS-PDSCH channels: 15.
647
The bearer 23 cannot be selected because:

• The number of HS-PDSCH channels (7) exceeds the maximum number of HS-PDSCH channels the terminal can use (5),
• And the transport block size (9719 Bytes) exceeds the maximum transport block size (7298 Bytes) the terminal can
carried.
In the Bearer table, Atoll searches a suitable bearer and selects the bearer index 22.
• The number of HS-PDSCH channels (5) does not exceed the maximum number of HS-PDSCH channels the terminal can
use (5) and the maximum number of HS-PDSCH channels available at the cell level (15),
• The transport block size (7168 Bytes) does not exceed the maximum transport block size (7298 Bytes) the terminal can
carried.
• 16QAM modulation is supported by the terminal.
HS-PDSCH Quality Update
Once the bearer selected, Atoll knows the number of HS-PDSCH channels. Therefore, when the method “Without useful
signal” is used, Atoll can recalculate the HS-PDSCH quality with the real number of HS-PDSCH channels (A default value of 5
was taken into account in the first HS-PDSCH quality calculation).
9.5.4.1.2 CQI Based on HS-PDSCH Quality

When the option “CQI based on HS-PDSCH quality” is selected, Atoll proceeds as follows.
HS-PDSCH Quality Calculation
1st step: Atoll calculates the HS-PDSCH power ( P HS – PDSCH ).
P HSDPA  ic  is the power available for HSDPA on the carrier ic. This parameter is a user-defined cell input.
Therefore, we have:
n HS – SCCH is the number of HS-SCCH channels and P HS – SCCH  ic  is the HS-SCCH power on carrier ic fixed by the user. The HS-
Req
SCCH power is controlled so as to reach the required HS-SCCH Ec/Nt (  ------  ic 
Ec
) specified in mobility properties.
Nt HS – SCCH
We have:
TXi
 -----
Ec-   BTS  RSCP HS – SCCH  ic 
 Nt  ic  HS – SCCH = ----------------------------------------------------------
DL
N Tot  ic 
And
TXi
 -----
Ec-   BTS  RSCP HS – SCCH  ic 
 Nt  ic  HS – SCCH
= --------------------------------------------------------------------------------------------------------------------------------------------------------
DL DL Term TXi
N Tot  ic  –  1 – F Ortho    1 – F JD    BTS  RSCP HS – SCCH  ic 
With
DL DL DL DL Term
DL
TXi P HS – SCCH  ic 
RSCP HS – SCCH  ic  = -------------------------------
-
LT
i
And
CI
L T = ----------------------------------------------------------------------------------------------------------------------
-
G TX  G Term
Term Term
Therefore,
648
Req
  -----
Ec-
  DL 
 HS – SCCH  N Tot  ic 
ic
  Nt
P HS – SCCH  ic  =  -------------------------------------------------------------------  L T for the total noise option,
  BTS 
 
And
Req
  -----
Ec-  DL 
  Nt  ic  HS – SCCH  N Tot  ic  
P HS – SCCH  ic  =  --------------------------------------------------------------------------------------------------------------------------------------------
Req
  L T for the without useful signal option.
    1 +  1 – F DL    1 – F Term    ----- Ec-
 ic   
 BTS  Ortho JD  Nt  HS – SCCH 
2nd step: Then, Atoll evaluates the HS-PDSCH quality

Ec
We have:
TXi
 -----
Ec-   BTS  RSCP HS – PDSCH  ic 
 Nt  ic  HS – PDSCH
= -------------------------------------------------------------
DL
N Tot  ic 
And
TXi
 -----  BTS  RSCP HS – PDSCH  ic 
 ic 
Ec-
= -----------------------------------------------------------------------------------------------------------------------------------------------------------
 Nt  HS – PDSCH TXi
DL DL Term RSCP HS – PDSCH  ic 
N Tot  ic  –  1 – F Ortho    1 – F JD    BTS  ---------------------------------------------
n
Here, Atoll works on the assumption that five HS-PDSCH channels are used (n=5). Then, it deduces the HS-PDSCH CQI and the
bearer to be used. Once the bearer selected, Atoll exactly knows the number of HS-PDSCH channels and recalculates the HS-
PDSCH quality with the real number of HS-PDSCH channels.
With
DL DL DL DL Term
DL
TXi P HS – PDSCH  ic 
RSCP HS – PDSCH  ic  = ----------------------------------
-
LT
And
CI
L T = ----------------------------------------------------------------------------------------------------------------------
-
G TX  G Term
Term Term
HS-PDSCH CQI Determination
Let us assume the following notation: CQIHS – PDSCH corresponds to the HS-PDSCH CQI. CQI HS – PDSCH is deduced from the
table CQI HS – PDSCH = f   ------  ic   . This table is defined for the terminal reception equipment and the specified
Ec
  Nt  HS – PDSCH
mobility.
Bearer Selection
The bearer is selected as described in "Bearer Selection" on page 646.
9.5.4.2 Coverage Prediction Display Options

Three display options are available in the study property dialogue.
649
9.5.4.2.1 Colour per CQI

Atoll displays either the P-CCPCH CQI when the selected option in Global Parameters (HSDPA part) is CQI based on P-CCPCH
quality, or the HS-PDSCH CQI when considering the CQI based on HS-PDSCH quality option.
Coverage consists of several layers with a layer per CQI threshold ( CQI Threshold ). For each layer, area is covered if
CQI  CQI Threshold . Each layer is assigned a colour and displayed with intersections between layers.
9.5.4.2.2 Colour per Peak Throughput

After selecting the bearer, Atoll reads the corresponding RLC peak throughput. This is the highest throughput that the bearer
can provide on each bin.
DL
Coverage consists of several layers with a layer per possible peak throughput ( R Peak ). For each layer, area is covered if the
peak throughput can be provided. Each layer is assigned a colour and displayed with intersections between layers.
9.5.4.2.3 Colour per HS-PDSCH Ec/Nt

Atoll displays on each bin the HS-PDSCH quality. Coverage consists of several layers with a layer per threshold. For each layer,
area is covered if  ------  ic 

Ec
 Threshold . Each layer is assigned a colour and displayed with intersections between
Nt HS – PDSCH
layers.
9.6 N-Frequency Mode and Carrier Allocation

Transmitters that support N-frequency mode are multi carrier transmitters with a master and one or more slave carrier. You
can assign master and slave carriers to transmitters manually, or use the automatic frequency allocation in Atoll to assign
carrier types automatically.
9.6.1 Automatic Carrier Allocation

For each transmitter, Atoll determines a list of "near" transmitters. For any transmitter TXi, its "near" transmitters are
geographically located close to the transmitter, and are sorted according to their distance from it. The calculation of distance
between TXi and any other transmitter TXj is performed using the equation below:
TX i – TX j TX i – TX j
D = d   1 + x   cos    – cos    – 2  
TX i – TX j TX i – TX j
Where D is the weighted distance between TXi and TXj, d is the real distance between between TXi and TXj
TX i – TX j
considering any offsets with respect to the site locations, x is set to 15 % so that the maximum variation in D due to
the azimuths does not exceed 60 %.  and  are calculated from the azimuths of the two cells as shown in Figure 9.11 on
page 650.
Figure 9.11: Weighted Distance Between Transmitters
The above formula implies that two transmitters facing each other will have a shorter weighted distance between them than
the real distance, and two transmitters pointing in opposite directions will have a greater weighted distance.
650
Allocation of All Carriers
Co-N-Frequency Allocation Diff-N-Frequency Allocation
Atoll assigns the same carriers to cells of each co-site Atoll assigns different carriers to cells of each co-site
transmitter. transmitter.
Allocation of Master Carriers
Atoll assigns one master carrier to each transmitter TXi, such that the master carrier of TXi is different from the master carrier
of TXj, where TXj belongs to the list of "near" transmitters. The master carrier is one of the cells defined in the transmitter. All
the other cells of the transmitter are assigned the carrier-type "slave".
For transmitters that support the N-frequency mode and have master carriers properly assigned, Atoll performs the
neighbour and scrambling code allocation for the master carrier only.
9.7 Neighbour Allocation

Atoll permits the automatic allocation of intra-technology neighbours in a TD-SCDMA document. The intra-technology
neighbour allocation algorithms take into account all the cells of TBC transmitters. It means that all the cells of TBC
transmitters of your .atl document are potential neighbours.
The cells to be allocated will be called TBA cells. They must fulfill the following conditions:
• They are active
• Their transmitters support the N-frequency mode, and the cells are master carriers of their transmitters (neighbours
are not allocated to standalone carriers)
• They satisfy the filter criteria applied to the Transmitters folder
• They are located inside the focus zone
zone.

• "Neighbour Allocation for All Transmitters" on page 651.

We assume that we have a reference cell A and a candidate neighbour, cell B. When automatic allocation starts, Atoll checks
1. The distance between both cells must be less than the user-defined maximum inter-site distance. If the distance
between the reference cell and the candidate neighbour is greater than this value, the candidate neighbour is
discarded.
651
• Force co-site cells as neighbours: This option enables you to force cells located on the same site as reference cell
in the candidate neighbour list. This constraint can be weighted among the others and ranks the neighbours
through the importance field.
• Force adjacent cells as neighbours: This option enables you to force cells geographically adjacent to the reference
cell in the candidate neighbour list. This constraint can be weighted among the others and ranks the neighbours
through the importance field.
• Adjacency criterion: Geographically adjacent cells are determined on the basis of

their best server coverages in TD-SCDMA projects. Let CellA be a candidate neighbour
cell of CellB. CellA is considered adjacent to CellB if there exists at least one pixel in
the CellB best server coverage area (and P-CCPCH RSCP of CellB > P-CCPCH RSCP
T_Add) where CellA is best server (of several cells have the same best server value) or
CellA is the second best server that enters the handover set (i.e., P-CCPCH RSCP of
CellA > P-CCPCH RSCP T_Drop and P-CCPCH RSCP of CellA > P-CCPCH RSCP of CellB -
T_Comp.)
• When this option is selected, adjacent cells are sorted and listed from the most
adjacent to the least, depending on the above criterion. Adjacency is relative to the
number of pixels satisfying the criterion.
• Force symmetry: This option enables you to force the reciprocity of a neighbourhood link. Therefore, if the
reference cell is a candidate neighbour of another cell, the later will be considered as candidate neighbour of the
reference cell.
If the neighbours list of a transmitter is full, the reference transmitter will not be added
as a neighbour of that transmitter and that transmitter will be removed from the
reference transmitter’s neighbours list. You can force Atoll to keep that transmitter in
the reference transmitter’s neighbours list by adding the following option in the Atoll.ini
file:
[Neighbours]
3. If the Use Coverage Conditions check box is selected, there must be an overlapping zone ( S A  S B ) with a given cell
N-frequency handover is a baton handover. Assuming that the reference cell A and the candidate cell B are located
inside a continuous layer of cells:
SA is the area where the cell A is the best serving cell.
• The P-CCPCH RSCP from the cell A is greater than the P-CCPCH RSCP T_Add.
• The P-CCPCH RSCP from the cell A is greater than the P-CCPCH RSCP from all other cells.
652
SB is the area where the cell B can enter the handover set.
• The P-CCPCH RSCP from the cell B is greater than the P-CCPCH RSCP T_Drop.
• The P-CCPCH RSCP from the cell B is greater than the P-CCPCH RSCP from the cell A minus the P-CCPCH RSCP
T_Comp.
Figure 9.12: N-frequency Neighbour Allocation
SA  SB
-  100 ), which it compares with the % minimum covered
SA
area. If this percentage is not exceeded, the candidate neighbour B is discarded.
The coverage condition can be weighted among the others and ranks the neighbours through the importance field.
: Overlapping Coverages
4. The importance of neighbours.

653
Importance values are used by the allocation algorithm to rank the neighbours according to the allocation reason.
Atoll lists all neighbours and sorts them by importance value so as to eliminate some of them from the neighbour list
if the maximum number of neighbours to be allocated to each transmitter is exceeded.
If we consider the case for which there are 15 candidate neighbours and the maximum number of neighbours to be
allocated to the reference cell is 8. Among these 15 candidate neighbours, only 8 (having the highest importance
values) will be allocated to the reference cell. Note that maximum numbers of neighbours can be defined at the cell
level (properties dialogue or Cells table). If defined there, this value is taken into account instead of the default one
available in the dialogue.
In the Results part, Atoll provides the list of neighbours, the number of neighbours, and the maximum number of neighbours
allowed for each cell. In addition, it indicates the importance (in %) of each neighbour and the allocation reason, i.e., a
neighbour may be marked as exceptional pair, co-site, adjacent, coverage, or symmetric. For neighbours accepted for co-site,
adjacency, and coverage reasons, Atoll displays the percentage of area that satisfies the coverage conditions and the
corresponding surface area (km2), the percentage of area that satisfies the adjacency conditions and the corresponding
surface area (km2). Finally, if cells have previous allocations in the list, neighbours are marked as existing.
site distance with the effective inter-transmitter distance. As a consequence,
there can be cases where the real distance between assigned neighbours is
higher than the Max inter-site distance, because the effective distance is
smaller. You can force Atoll to compare the Max inter-site distance with the real
inter-site distance by adding the following lines in Atoll.ini:
[Neighbours]
Atoll to prioritise the individual distances between reference transmitters and
their respective neighbour candidates by adding the following lines in Atoll.ini:
[Neighbours]
654
• You do not require simulations or coverage predictions for an automatic neighbour

allocation. For automatic neighbour allocation, Atoll automatically calculates the
missing path loss matrices.
• Although no specific terminal, mobility, or service is selected for automatic neighbour
allocation, the algorithm tries to find the maximum number of neighbours by
selecting:
• The service with the lowest body loss
• The terminal with the highest difference between Gain and Losses. If this is the
same for all terminals, Atoll uses the terminal with the lowest noise figure.
• Mobility does not impact the allocation
relationship already exists and the Delete existing neighbours option is not selected.
In this case, Atoll displays a warning message in the Event Viewer indicating that the
constraint on the forbidden neighbour will be ignored by the algorithm because the
neighbour already exists.
• Symmetric neighbour relations are only added to the neighbour lists if the neighbour
lists are not already full. Thus, if the cell B is a neighbour of the cell A, but cell A is not
a neighbour of the cell B, there can be two possibilities:
1. There is space in the cell B neighbour list: cell A will be added to the list. It will be
the last one.
2. The cell B neighbour list is full: Atoll will not include cell A in the list and will remove
the symmetric relation by deleting cell B from the cell A neighbour list.
• If you select Force exceptional pairs and Force symmetry options, Atoll considers the
constraints between exceptional pairs in both directions so as to respect the
symmetric relation. On the other hand, if a neighbour relation is forced in one
direction and forbidden in the other, symmetry cannot be respected. In this case,
Atoll displays a warning message in the Event Viewer.
• In the results, Atoll displays only the cells for which it finds new neighbours.

Transmitter
• TBA cells
• Neighbours of TBA cells marked as exceptional pair, adjacent, or symmetric
• Neighbours of TBA cells that satisfy coverage conditions

Importance values are used by the allocation algorithm to rank the neighbours according to the allocation reason and the
distance, and to quantify the neighbour importance. The neighbour importance depends on the distance from the reference
transmitter and on the neighbourhood cause (cf. table below); this value varies between 0 and 100%.
selected
selected
relationship
655
transmitter.
d-
 Di  = 1 – ----------
d max

Neighbourhood cause Resulting IF using the default values

Importance Function
Co-site Adjacent from the table above
No No 10%+20%{10%(Di)+90%(O)}+1%+9%(Di)
No Yes
Yes Yes
Where:
overlapping.
causes.
9.7.4 Appendix: Calculation of the Inter-Transmitter Distance

d = D   1 + x  cos  – x  cos  
656
9.8 Scrambling Code Allocation

Downlink scrambling codes enable mobile to distinguish one cell from another. In TD-SCDMA, there are 128 scrambling codes
(or P-CCPCH midamble codes) distributed in 32 clusters of 4 codes each. A different DL synchronisation code, or SYNC_DL
code, is assigned to each cluster. Scrambling codes are numbered from 0 to 127, and SYNC_DL codes from 0 to 31.
Depending on the options you select for automatic allocation of scrambling and SYNC_DL codes, Atoll takes into account
either all the cells of TBC transmitters, or only cells of active and filtered transmitters located inside the computation zone.
Atoll calculates a scrambling code and a SYNC_DL code to all these cells. But, it allocates scrambling codes and SYNC_DL codes
only to TBA cells (cells to be allocated). TBA cells are the cells that fulfill the following conditions:
• They are active
• They satisfy the filter criteria applied to the Transmitters folder
Furthermore, if there are transmitters that support the N-frequency mode among the TBC transmitters of your network, the
scrambling code allocation also considers the master and slave carrier allocations.
zone.

9.8.1.1 Allocation Constraints and Options
The scrambling code and SYNC_DL code allocation algorithm can take into account following constraints:
1. Neighbour relations between cells
You may consider:
• Atoll can take into account inter-technology neighbour relations as constraints to

allocate different scrambling codes to the TD-SCDMA neighbours of a GSM
transmitter. In order to consider inter-technology neighbour relations in the
scrambling code allocation, you must make the Transmitters folder of the GSM.atl
document accessible in the TD-SCDMA.atl document. For information on making links
between GSM and TD-SCDMA .atl documents, see the User Manual.
2. The scrambling code reuse distance
657
Reuse Distance: It is a constraint on the allocation of scrambling codes. The same scarmbling code or SYNC_DL code
cannot be allocated to two sites that are not farther apart than the reuse distance. Scrambling code reuse distance
can be defined for each cell in the cell properties. If this value is not defined, Atoll uses the default reuse distance
defined in the Automatic Scrambling Code and SYNC_DL code Allocation dialogue. The reuse distance constraint is
used for clustered and distributed per cell allocation strategies.
3. The carrier for which you want to perform the automatic allocation
Carrier: You can select "All" or a specific carrier. If you select "All", Atoll allocates the same scrambling code to each
carrier of a transmitter.
4. The number of scrambling codes per SYNC_DL code
Each SYNC_DL code corresponds to a group of scrambling codes as defined in 3GPP specifications. 3GPP specifications
define 32 SYNC_DL codes with 4 corresponding scrambling codes each (SYNC_DL codes are numbered from 0 to 31).
However, it is possible to define a different value (e.g. if you set the number of scrambling codes per SYNC_DL codes
to 2, scrambling codes will be distributed among 64 SYNC_DL codes).
When the allocation is based on a Distributed strategy (Distributed per Cell or Distributed per Site), this parameter can
also be used to define the interval between the scrambling codes assigned to cells on a same site. The defined interval
is applied by adding the following lines in the Atoll.ini file:
[PSC]
ConstantStep=1
For more information about setting options in the atoll.ini file, see the Administrator Manual.
5. Atoll can use a maximum of codes
Use a Maximum of Codes: If you choose to use a maximum of codes, Atoll will try to spread the allocated spectrum of
scrambling codes as much as possible.
6. Existing allocation
Delete All Codes: If you select this option, Atoll will delete any existing scrambling code allocation and perform a fresh
allocation. Otherwise, Atoll keeps the existing allocation.
9.8.1.2 Allocation Strategies

You can choose from the following four allocation strategies:
• Clustered: The purpose of this strategy is to choose for a group of mutually constrained cells, scrambling codes among
a minimum number of clusters. In this case, Atoll will preferentially allocate all the codes within the same cluster.
• Distributed per Cell: This strategy consists in using as many clusters as possible. Atoll will preferentially allocate codes
from different clusters.
• One SYNC_DL code per site: This strategy allocates one cluster, i.e., one SYNC_DL code, per site, then one scrambling
code from the cluster to each cell of the site. When all the clusters have been allocated but there are still sites
remaining, Atoll reuses the clusters as far as possible at another site.
Same carriers must be assigned different scrambling codes. Different carriers of the same
site can be assigned the same scrambling code. Therefore, cells of a transmitter (i.e.
different carriers) are assigned the same scrambling code, if the scrambling code
domains associated with the carriers have a common cluster or enough codes in one
cluster.
• Distributed per site: This strategy allocates a group of adjacent clusters, i.e., consecutive SYNC_DL codes, to each site,
then one cluster, or SYNC_DL code, to each transmitter on the site according to its azimuth, and finally one scrambling
code from each cluster to each cell of each transmitter. The number of adjacent clusters, or consecutive SYNC_DL
codes, depends on the number of transmitters per site. When all the sites have been allocated adjacent clusters, and
there are still sites remaining to be allocated, Atoll reuses the adjacent clusters as far as possible at another site.
In the Results table, Atoll only displays scrambling codes and SYNC_DL codes allocated to TBA cells.

TBA cell may be:
"First Order"),
658
Additional constraints are considered when:
• The cell and its near cells are neighbours of a same GSM transmitter (only if the Transmitters folder of the GSM.atl
document is accessible in the TD-SCDMA.atl document),
• The neighbour cells cannot share the same cluster (for the "Distributed per site" allocation strategy only).
and the cost of the scrambling code plan. During the allocation, Atoll tries to assign different scrambling codes to the TBA cell
and its near cells. If it respects all the constraints, the cost of the scrambling code plan is 0. When a cell has too many
constraints and there are not anymore scrambling codes available, Atoll breaks the constraint with the lowest cost so as to
generate the scrambling code plan with the lowest cost. For information on the cost generated by each constraint, see "Cell

Strategies: Clustered and Distributed per Cell
Atoll processes TBA cells according to their priority. It allocates scrambling codes starting with the highest priority cell and its
near cells, and continuing with the lowest priority cells not allocated yet and their near cells. For information on calculating
cell priority, see "Cell Priority" on page 660.
Strategy: One SYNC_DL Code per Site
Atoll assigns a cluster, i.e., a SYNC_DL code, to each site, starting with the highest priority site and its near sites, and continuing
with the lowest priority sites not allocated yet and their near sites. When all the clusters have been allocated but there are
still sites remaining, Atoll reuses the clusters at the other sites. When the Reuse Distance option is selected, the algorithm
reuses the clusters as soon as the reuse distance is exceeded. Otherwise, when the option is not selected, the algorithm tries
to assign reused clusters as spaced out as possible.
Then, Atoll allocates a scrambling code from the cluster to each cell located on the sites (codes belong to the assigned
clusters). It starts with the highest priority cell and its near cells and goes on with the lowest priority cells not allocated yet
and their near cells.
Strategy: Distributed per Site
Atoll assigns a group of adjacent clusters, i.e., SYNC_DL codes, to each site, starting with the highest priority site and its near
sites, and continuing with the lowest priority sites not allocated yet and their near sites. When all the sites have been allocated
adjacent clusters, and there are still sites remaining to be allocated, Atoll reuses the adjacent clusters at other sites. When
the Reuse Distance option is selected, the algorithm reuses the clusters as soon as the reuse distance is exceeded. Otherwise,
when the option is not selected, the algorithm tries to assign reused clusters as spaced out as possible.
Then, Atoll assigns each cluster of the group to each transmitter of the site according to the transmitter azimuth and selected
neighbourhood constraints (options "Neighbours in Other Clusters" and "Secondary Neighbours in Other Clusters"). Then,
Atoll allocates a scrambling code to each cell located on the transmitters (codes belong to the assigned clusters). It starts with
the highest priority cell and its near cells and goes on with the lowest priority cells not allocated yet and their near cells.
For information on calculating site priority, see "Site Priority" on page 663. or information on calculating cell priority, see "Cell
Determination of Groups of Adjacent Clusters
In order to determine the groups of adjacent clusters to be used, Atoll:

• Defines theoretical groups of adjacent clusters, independent of the defined domain, considering the 128 scrambling
codes available and 4 codes per cluster.
• Starts the distribution of clusters to groups from the cluster 0
• Takes into account the maximum number of transmitters per site in order to determine the number of clusters in each
group
• Determines the total number of groups
659
If the number of scrambling codes per cluster is set to 4 and the maximum number of transmitters per site in the network is
3, the theoretical groups of adjacent clusters will be:
Group 1 Group 2 Group 3 Group 4 ... Group 11
Cluster 1 Cluster 4 Cluster 7 Cluster 10 ... Cluster 31
Cluster 2 Cluster 5 Cluster 8 Cluster 11
If no domain is assigned to cells, Atoll can use all these groups for the allocation. On the other hand, if a domain is used, Atoll
compares adjacent clusters actually available in the assigned domain with the theoretical groups and only keeps adjacent
clusters common with the theoretical groups.
If we have a domain comprising 12 clusters: clusters 1 to 8 and clusters 12 to 15. In this case, Atoll will use the following groups
of adjacent clusters:
• Group 2 with cluster 3, 4 and 5
The clusters 1, 2 and 15 will not be used.
If a domain does not contain any adjacent clusters, Atoll displays a warning message in the Event Viewer.

In case you have a multi-carrier network and you run the scrambling code allocation on all the carriers, the allocation order
changes. It is no longer based on the cell priority but depends on the transmitter priority. All transmitters which have
constraints with the studied transmitter will be referred to as near transmitters.
In case of a "Per cell" strategy (Clustered and Distributed per cell), Atoll starts scrambling code allocation with the highest
priority transmitter and its near transmitters and continues with the lowest priority transmitters not allocated yet and their
near transmitters. The same scrambling code is assigned to each cell of the transmitter.
In case of the "One SYNC_DL code per site" strategy, Atoll assigns a cluster, i.e., a SYNC_DL code, to each site and then,
allocates a scrambling code to each transmitter. It starts with the highest priority transmitter and its near transmitters and
continues with the lowest priority transmitters not allocated yet and their near transmitters. The same scrambling code is
assigned to each cell of the transmitter.
In case of the "Distributed per site" strategy, Atoll assigns a group of adjacent clusters, i.e., SYNC_DL codes, to each site, then
a cluster to each transmitter and finally, allocates a scrambling code to each transmitter. It starts with the highest priority
transmitter and its near transmitters and continues with the lowest priority transmitters not allocated yet and their near
transmitters. The same scrambling code is assigned to each cell of the transmitter.
For information on calculating transmitter priority, see "Transmitter Priority" on page 662.
alphanumeric order.

Scrambling code allocation algorithm in Atoll allots priorities to cells before performing the actual allocation. Priorities
assigned to cells depend upon how much constrained each cell is and the cost defined for each constraint. A cell without any
constraint has a default cost, C , equal to 0. The higher the cost on a cell, the higher the priority it has for the scrambling code
allocation process.
There are seven criteria employed to determine the cell priority. The total cost due to constraints on any cell i is defined as:
C i = C i  Dom  + C i  U 
With
C i  U  = C i  Dist  + C i  EP  + C i  N  + C i  N 2G  + C i  Cluster  + C i  CN 
All the cost components are described below:

• Scrambling Code Domain Criterion
The cost due to the domain constraint, C i  Dom  , depends on the number of scrambling codes available for the allocation.
The domain constraint is mandatory and cannot be broken.
660
When no domain is assigned to cells, 128 scrambling codes are available and we have:
C i  Dom  = 0
When domains of scrambling codes are assigned to cells, each unavailable scrambling code generates a cost. The higher the
number of codes available in the domain, the less will be the cost due to this criterion. The cost is given as:
C i  Dom  = 128 – Number of scrambling codes in the domain
• Distance Criterion
C i  Dist  =  Cj  Dist  i  
j
Where

C i  EP  =  cEP  i – j 
j
Where
• Neighbourhood Criterion

     
Ci  N  = 

C j  N1  i   = I j  c N1
Where
661
Because two first order neighbours must not have the same scrambling code, Atoll considers the cost created by two first
C j  N1  i   + C j  N1  i  
C j – j  N1  i   = -----------------------------------------------------
-
2
Where
Because two second order neighbours must not have the same scrambling code, Atoll considers the cost created by two
second order neighbours to be each other.
C k  N2  i   + C k  N2  i  
C k – k  N2  i   = -------------------------------------------------------
-
2
C l  N3  i   = Max  j   c N3
Where
Because two third order neighbours must not have the same scrambling code, Atoll considers the cost created by two third
C l  N3  i   + C l  N3  i  
C l – l  N3  i   = -----------------------------------------------------
-
2

And
• Close Neighbour Criterion

The constraint level of any cell i depends on the number of its close neighbour cells j. The close neighbour cost ( C i  CN  )
depends on two components: the importance of the neighbour relation ( I i – j ) and the distance ( d i – j ) relative to maximum
Max
close neighbour distance ( d CN ).
 I +  1 – ----------- di – j 
 
 i–j  Max
 
 d CN 
C i  CN  =   -----------------------------------------  c CN
 2 
j  
 
Where
c CN is the cost of the close neighbour constraint. This value can be defined in the Constraint Cost dialogue.
• GSM Neighbour Criterion

This criterion is considered when the co-planning mode is activated (i.e. the Transmitters folder of the GSM .atl document is
made accessible in the TD-SCDMA.atl document) and inter-technology neighbours have been allocated. If the cell i is
neighbour of a GSM transmitter, the cell constraint level depends on how many cells j are neighbours of the same GSM
transmitter. The total cost due to GSM neighbour constraint is given as:
662
C i  N 2G  =  cN2G  j – Tx2G 
j
Where
cN is the cost of the GSM neighbour constraint. This value can be defined in the Constraint Cost dialogue.
2G
• Cluster Criterion
When the "Distributed per Site" allocation strategy is used, you can consider additional constraints on allocated clusters (one
cell, its first order neighbours and its second order neighbours must be assigned scrambling codes from different clusters). In
this case, the constraint level of any cell i depends on the number of first and second order neighbours, j and k. The total cost
due to the cluster constraint is given as:
C i  Cluster  =  Cj  N1  i    cCluster +  Ck  N2  i    cCluster

j k
Where
c Cluster is the cost of the cluster constraint. This value can be defined in the Constraint Cost dialogue.

In case you have a multi-carrier network and you run scrambling code allocation on "all" the carriers, Atoll allots priorities to
transmitters. Priorities assigned to transmitters depend on how much constrained each transmitter is and the cost defined
for each constraint. The higher the cost on a transmitter, the higher the priority it has for the scrambling code allocation
process.
With C Tx  U  = Max  C  U   and C  Dom  = 128 – Number of scrambling codes in the domain
i  Tx i Tx

In case of "Per Site" allocation strategies (One SYNC_DL code per Site and Distributed per Site), Atoll allots priorities to sites.
Priorities assigned to sites depend on how much constrained each site is and the cost defined for each constraint. The higher
the cost on a site, the higher the priority it has for the scrambling code allocation process.
Let us consider a site S with three transmitters; each of them has two cells using carriers 0 and 1. The cost due to constraints
on the site is given as:
C S = C S  U  + C S  Dom 
With C S  U  = Max  C  U   and C  Dom  = 128 – Number of scrambling codes in the domain
Tx  S Tx S
Here, the domain considered for the site is the intersection of domains available for transmitters of the site. The domain
663
9.8.2 Scrambling Code Allocation Example

9.8.2.1 Single Carrier Network
In order to understand the differences between the different allocation strategies and the behaviour of algorithm when using
a maximum of codes or not, let us consider the following sample scenario:
Figure 9.15: Scrambling Code Allocation Example
Let Site0, Site1, Site2, and Site3 be four sites, with 3 transmitters each using carrier 0, to whom scrambling codes have to be
allocated out of 6 clusters of 4 scrambling codes. This implies that the domain of scrambling codes for the four sites is from 0
to 23 (cluster 0 to cluster 5). The reuse distance is supposed to be less than the inter-site distance. Only co-site neighbours
exist.
The following section shows the results of each combination of options with explanations where necessary.
9.8.2.1.1 Strategy: Clustered

Since the restrictions of neighbourhood only apply to co-sites and, in our case, the distances between sites are greater than
the reuse distance, every cell has the same priority. Allocation is performed in an alphanumeric order.
Without "Use a Maximum of Code" With "Use a Maximum of Code"
As it is possible to use a maximum of codes, Atoll starts

allocation at the start of a different cluster at each site.
Atoll starts allocating the codes from the start of cluster 0 at
When a cluster is reused, and there are non allocated codes
each site.
left in the cluster, Atoll first allocates those codes before
reusing the already used ones.
664
9.8.2.1.2 Strategy: Distributed per Cell

Since the restrictions of neighbourhood only apply to co-sites and, in our case, the distances between sites are greater than
the reuse distance, every cell has the same priority. Allocation is performed in an alphanumeric order.
Atoll allocates codes from different clusters to each cell of

Atoll allocates codes from different clusters to each site’s
the same site. Under given constraints of neighbourhood
cells. As it is possible to use a maximum of codes, Atoll
and reuse distance, same codes can be allocated to each
allocates the codes so that there is least repetition of codes.
site’s cells.
9.8.2.1.3 Strategy: One SYNC_DL Code per Site

Since the restrictions of neighbourhood only apply to co-sites, therefore, every site has the same priority. Cluster allocation
to sites is performed in an alphanumeric order.
In this strategy, a cluster of codes is limited to be used at just

one site at a time unless all codes and clusters have been When it is possible to use a maximum of codes, Atoll can
allocated and there are still sites remaining to be allocated. allocate different codes from a reused cluster at another
In this case, Atoll reuses the cluster as far as possible at site.
another site.
665
9.8.2.1.4 Strategy: Distributed per Site

Since the restrictions of neighbourhood only apply to co-sites, therefore, every site has the same priority. Cluster allocation
to sites is performed in an alphanumeric order.
A group of adjacent clusters is allocated to one site at a time,

unless all the codes and groups of adjacent clusters have
When it is possible to use a maximum of codes, Atoll can
been allocated but there are still sites remaining to be
allocate different codes from a reused groups of adjacent
allocated. In this case (here only one group of adjacent
clusters at another site.
clusters 0, 1, and 2 is available), Atoll reuses the group as far
as possible at another site.
9.8.2.2 Multi Carrier Network

If you have a multi carrier network, i.e., transmitters with more than one cells using different carriers, and you run scrambling
code allocation on "all" the carriers, Atoll allocates the same scrambling code to each carrier of a transmitter.
Let Site0, Site1, Site2, and Site3 be four sites with 3 cells using carrier 0 and 3 cells using carrier 1. Scrambling codes have to
be allocated out of 6 clusters consisted of 4 scrambling codes. This implies that the domain of scrambling codes for the four
sites is from 0 to 23 (cluster 0 to cluster 5). The reuse distance is supposed to be less than the inter-site distance. Only co-site
neighbours exist. Every site has the same priority and the cluster allocation to sites is performed in an alphanumeric order.
Atoll allocates one cluster at each site and then, one code to each transmitter. Then, the same code is given to each cell of
the transmitter.
Figure 9.16: Scrambling Code Allocation to All Carriers
666
9.9 Automatic GSM/TD-SCDMA Neighbour Allocation

It is possible to automatically calculate and allocate neighbours between GSM and TD-SCDMA networks. In Atoll, it is called
inter-technology neighbour allocation.
• When the TD-SCDMA coverage is not continuous. In this case, the TD-SCDMA coverage is extended by TD-SCDMA to
GSM handovers.
• In order to balance traffic and service distribution between both networks.
Atoll’s automatic inter-technology neighbour allocation algorithm takes into account both cases.
• An .atl document containing the GSM network, GSM.atl, and another one containing the TD-SCDMA network, TD-
SCDMA.atl,
• An existing link on the Transmitters folder of GSM.atl into TD-SCDMA.atl.
transmitters of GSM.atl are potential neighbours. The TD-SCDMA cells, in TD-SCDMA.atl, to be allocated neighbours are called
TBA cells which fulfill following conditions:
• They are active
• They satisfy the filter criteria applied to Transmitters folder
one of its subfolders.
Only TD-SCDMA TBA cells can be assigned neighbours.

• The inter-transmitter distance
• The maximum number of neighbours
• Allocation options
• The selected allocation strategy
We assume we have a TD-SCDMA reference cell, A, and a GSM candidate neighbour transmitter, B.

1. The distance between the TD-SCDMA reference cell and the GSM neighbour must be less than the user-defined
maximum inter-site distance. If the distance between the TD-SCDMA reference cell and the GSM neighbour is greater
than this value, then the candidate neighbour is discarded.
Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. You may choose one
or more carriers. Atoll will allocate neighbours to cells using the selected carriers.
Force co-site cells as neighbours: It enables you to automatically include GSM transmitters located on the same site
than the reference TD-SCDMA cell in the candidate neighbour list. This option is automatically selected.
Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. Therefore, you
may force/forbid a GSM transmitter to be candidate neighbour of the reference TD-SCDMA cell.
Delete existing neighbours: When selecting the Delete existing neighbours option, Atoll deletes all the current
neighbours and carries out a new neighbour allocation. If not selected, existing neighbours are kept.
Next, Atoll calculates the importance of the automatically allocated neighbours. Atoll sorts the neighbours by
decreasing importance in order to keep the ones with high importance. If the maximum number of neighbours to be
allocated to each cell is exceeded, Atoll keeps the ones with high importance.
667
As indicated in the table below, the neighbour importance depends on the distance and on the neighbourhood cause;
this value varies between 0 to 100%.

Where d is the effective distance between the TD-SCDMA reference cell and the GSM neighbour and d max is the
maximum inter-site distance.
as existing.

1. The distance between the TD-SCDMA reference cell and the GSM neighbour must be less than the user-defined
maximum inter-site distance. If the distance between the TD-SCDMA reference cell and the GSM neighbour is greater
than this value, then the candidate neighbour is discarded.
Carriers: This option enables you to select the carrier(s) on which you want to run the allocation. You may choose one
or more carriers. Atoll will allocate neighbours to cells using the selected carriers.
Force co-site cells as neighbours: It enables you to automatically include GSM transmitters located on the same site
than the reference TD-SCDMA cell in the candidate neighbour list. This option is automatically selected.
Force exceptional pairs: This option enables you to force/forbid some neighbourhood relationships. Therefore, you
may force/forbid a GSM transmitter to be candidate neighbour of the reference TD-SCDMA cell.
Delete existing neighbours: When selecting the Delete existing neighbours option, Atoll deletes all the current
neighbours and carries out a new neighbour allocation. If not selected, existing neighbours are kept.
3. There must be an overlapping zone ( S A  S B ) with a given cell edge coverage probability.
Two different cases may be considered for SA:
• 1st case: SA is the area where the cell A is the best serving cell of the TD-SCDMA network.
• The pilot signal received from A is greater than the minimum pilot signal level and is the highest one.
• The margin is set to 0 dB.
• 2nd case: The margin is different from 0 dB and SA is the area where:
• The pilot signal level received from A exceeds the user-defined minimum pilot signal level and is within a
margin from the highest signal level.
• 1st case: SB is the area where the cell B is the best serving transmitter of the GSM network.
In this case, the margin must be set to 0 dB.
highest one.
• 2nd case: The margin is different from 0 dB and SB is the area where:
668
SA  SB
-  100 ) and compares this value to the % minimum covered
SA
area. If this percentage is less than the minimum, the candidate neighbour B is discarded.
Candidate neighbours fulfilling coverage conditions are sorted in descending order with respect to percentage of
covered area.
Next, Atoll calculates the importance of the automatically allocated neighbours. Atoll sorts the neighbours by
decreasing importance in order to keep the ones with high importance. If the maximum number of neighbours to be
allocated to each cell is exceeded, Atoll keeps the ones with high importance.
As indicated in the table below, the neighbour importance depends on the distance and on the neighbourhood cause;
this value varies between 0 to 100%.

Co-site transmitter If the Force co-site cells as neighbours option is selected IF
transmitter.
d-
 Di  = 1 – ----------
d max


IF
Yes 60%+40%{1/7%(Di)+6/7%(O)}
Where
669
causes.

9.9.1.3 Appendices
9.9.1.3.1 Delete Existing Neighbours Option
selected. If a new TBA cell i is created in TD-SCDMA.atl, you can run the automatic allocation with the Delete existing
neighbours option not selected, in order to allocate neighbours to the new cell i only.
If you change some allocation criteria (e.g., increase the maximum number of neighbours or create a new GSM TBC
transmitter) and start a new allocation without selecting the Delete existing neighbours option, Atoll examines the neighbour
list of the TBA cells and checks allocation criteria only if there is still space left in their neighbour lists. A new GSM TBC
transmitter can enter the TBA cell neighbour list if allocation criteria are satisfied. It will be the first one in the neighbour list.
670
Chapter 10
WiMAX BWA Networks

• "Calculation Quick Reference" on page 678
• "Available Calculations" on page 690
• "Calculation Details" on page 702
• "Automatic Planning Algorithms" on page 758
© Forsk 2015
672
AT330_TRR_E1 Chapter 10: WiMAX BWA Networks
10 WiMAX BWA Networks

This chapter describes all the calculations performed in Atoll WiMAX documents. The first part of this chapter lists all the input
parameters in the WiMAX BWA documents, their significance, location in the Atoll GUI, and their usage. It also contains the
lists of the formulas used for the calculations.
The second part describes all the calculation processes, i.e., signal level coverage predictions, point analysis calculations,
signal quality coverage predictions, calculations on subscriber lists, and Monte Carlo simulations. The calculation algorithms
used by these calculation processes are available in the next part.
The third part describes all the calculation algorithms used in all the calculations. These algorithms include the calculation of
signal levels, noise, and interference for downlink and uplink considering the effects of smart antennas, power control,
subchannelisation, MIMO etc., and the radio resource management algorithms used by the different available schedulers.
If you are new to WiMAX, you can also see the Glossary of WiMAX Terms in the User Manual for information on WiMAX terms
and concepts, especially in the context of their user in Atoll.
• A cell refers to a transmitter-carrier (TX-c) pair. The cell being studied during a
calculation is referred to as TXi(ic) in this chapter.
• All the calculation algorithms in this section are described for two types of cells.
• A studied cell (represented by the subscript "i") comprising the studied transmitter
TXi and its carrier ic. It is the cell which is currently the focus of the calculation. For
example, a victim cell when calculating the interference it is receiving from other
cells.
• Other cells (represented by the subscript "j") comprising the other transmitter TXj
and its carrier jc. The other cells in the network can be interfering cells (downlink)
or the serving cells of interfering mobiles (uplink).
• All the calculation algorithms in this section are described for two types of receivers.
• Mi: A pixel (coverage predictions), subscriber (calculations on subscriber lists), or
mobile (Monte Carlo simulations) covered/served by the studied cell TXi(ic).
• Mj: A mobile (Monte Carlo simulations) covered/served by any other cell TXj(jc).
10.1 Definitions
This table lists the input to calculations, coverage predictions, and simulations.
K 1.38 x 10-23 J/K Boltzmann’s constant

Power spectral density of thermal
n0 Calculation result ( 10  Log  K  T  1000  = – 174 dBm/Hz ) dBm/Hz
noise
Frame Duration
D Frame Global parameter ms
Choice List: 2, 2.5, 4, 5, 8, 10, 12.5, 20
Cyclic Prefix Ratio
r CP Frame configuration or, otherwise, global parameter None
Choice List: 1/4, 1/8, 1/16, 1/32
DL
O Fixed Global parameter SD Fixed time-domain overhead (DL)
UL
O Fixed Global parameter SD Fixed time-domain overhead (UL)
DL
O Variable Global parameter % Variable time-domain overhead (DL)
UL
O Variable Global parameter % Variable time-domain overhead (UL)
TDD Ratio of the DL subframe to the

r DL-Frame Global parameter %
entire frame (TDD only)
673
Chapter 10: WiMAX BWA Networks ©Forsk 2015

Number of symbol durations per
TDD
N SD – DL Global parameter None frame that corresponds to the DL
subframe (TDD only)
Number of symbol durations per
TDD
N SD – UL Global parameter None frame that corresponds to the UL
subframe (TDD only)
TDD
D TTG Global parameter ms Transmit Time Guard (TDD only)
TDD
D RTG Global parameter ms Receive Time Guard (TDD only)
M PC Global parameter dB Uplink power control margin

Minimum signal to thermal noise
CNR Min Global parametera dB
threshold (interferer cutoff)
PZ Number of subchannels per channel
N SC – UL Permutation zone parameter None
in UL subframe
PZ Number of subchannels per channel
N SC – DL Permutation zone parameter None
in DL subframe
Total number of subcarriers per
N SCa – Total Frame configuration parameter None
channel (FFT size)
Number of subcarriers used by the
N SCa – Preamble Frame configuration parameter None
preamble
PZ Number of used subcarriers per
N SCa – Used Permutation zone parameter None
channel
PZ Number of subcarriers per channel
N SCa – Data Permutation zone parameter None
used for data transfer
Number of DC subcarriers per
N SCa – DC Hard-coded parameter ( N SCa – DC = 1 ) None
channel
PZ PZ PZ PZ Number of pilot subcarriers per
N SCa – Pilot Calculation result ( N SCa – Pilot = N SCa – Used – N SCa – Data ) None
channel
Calculation result Number of guard subcarriers per
PZ
N SCa – Guard PZ PZ None
( N SCa – Guard = N SCa – Total – N SCa – Used – N SCa – DC ) channel
PZ UL Permutation zone parameter None Uplink permutation zone
PZ DL Permutation zone parameter None Downlink permutation zone

Quality threshold: Required
QT PZ Permutation zone parameter dB preamble C/N or C/(I+N) for
accessing a zone
Speed limit for mobiles trying to
Speed Max – PZ Permutation zone parameter Km/hr
access a permutation zone
Maximum distance from the
d Max – PZ Permutation zone parameter m
transmitter covered by a zone
p PZ Permutation zone parameter None Permutation zone priority
W Channel Frequency band parameter MHz Channel bandwidth
First First channel number of the

frequency band
Last Last channel number of the
frequency band
Start frequency of the TDD frequency
F Start – FB – TDD Frequency band parameter MHz
band
DL Start frequency of the FDD
F Start – FB – FDD – DL Frequency band parameter MHz
frequency band
UL Start frequency of the FDD
F Start – FB – FDD – UL Frequency band parameter MHz
frequency band
674
f Sampling Frequency band parameter None Sampling factor
f ACS – FB Frequency band parameter dB Adjacent Channel Suppression Factor
ICS FB Frequency band parameter MHz Inter-channel spacing
CN FB Frequency band parameter None Channel number step
Inter – Tech Inter-technology interference

f IRF Network parameter dB
reduction factor
B Bearer parameter None Bearer index
Mod B Bearer parameter None Modulation used by the bearer
CR B Bearer parameter None Coding rate of the bearer

bits/
B Bearer parameter Bearer Efficiency
symbol
TB Bearer parameter dB Bearer selection threshold
Site Maximum backhaul site downlink

TP BH – DL Site parameter kbps
throughput
Site Maximum backhaul site uplink
TP BH – UL Site parameter kbps
throughput
nf dB Transmitter noise figure
Number of antennas used for MIMO
N Ant – TX Transmitter parameter None
in transmission
N Ant – RX Transmitter parameter None
in reception
TX Antenna parameter dB Transmitter antenna gain
G

L dB Transmitter loss
TX
E SA Smart antenna parameter None Number of smart antenna elements
Array
G SA Smart antenna parameter dB Array gain offset
Combining
G SA Smart antenna parameter dB Power combining gain offset
Div
G SA Smart antenna parameter dB Diversity gain (cross-polarisation)
N Channel Cell parameter None Cell’s channel number
P Preamble Cell parameter dBm Preamble power
Cell parameter P Traffic = P Preamble – P Traffic in dB

P Traffic P Traffic dB Traffic power reduction
-------------------------
Ratio 10
P Traffic = 10 in %
Cell parameter P Pilot = P Preamble – P Pilot in dB

P Pilot P Pilot dB Pilot power reduction
--------------------
Ratio 10
P Pilot = 10 in %
Cell parameter P Idle – Pilot = P Preamble – P Idle – Pilot in dB

P Idle – Pilot P Idle – Pilot dB Idle pilot power reduction
------------------------------------
Ratio 10
P Idle – Pilot = 10 in %
TL DL Cell parameter % Downlink traffic load
TL UL Cell parameter % Uplink traffic load
675
TL DL – Max Cell parameter % Maximum downlink traffic load
TL UL – Max Cell parameter % Maximum uplink traffic load
NR UL Cell parameter dB Uplink noise rise
NRUL – Seg Cell parameter dB Segmented zone uplink noise rise
N Users – Max Cell parameter None Maximum number of users per cell
N Users – DL Cell parameter None
cell in downlink
N Users – UL Cell parameter None
cell in uplink
SU DL Cell parameter % Downlink segmentation usage ratio
AU DL Cell parameter % Downlink AAS usage ratio
T AMS Cell parameter dB Adaptive MIMO switch threshold
T MU – MIMO Cell parameter dB Multi-user MIMO threshold
PI Cell parameter None Preamble index
T Preamble Cell parameter dB Preamble C/N threshold

Channel and preamble index reuse
D Reuse Cell parameter m
distance
G MU – MIMO Cell parameter None Uplink MU-MIMO gain
Inter – Tech
NR DL Cell parameter dB Inter-technology downlink noise rise
Inter – Tech
NR UL Cell parameter dB Inter-technology uplink noise rise
ZPBDL Cell parameter None Downlink zone permbase
ZPB UL Cell parameter None Uplink zone permbase

TX i  ic  Downlink multi-user diversity gain
G MUG – DL Proportional Fair scheduler parameter None
(MUG)
TX i  ic  Uplink multi-user diversity gain
G MUG – UL Proportional Fair scheduler parameter None
(MUG)
Max Maximum C/(I+N) above which no
CINR MUG Proportional Fair scheduler parameter dB
MUG gain is applied
Max
G SU – MIMO Cell WiMAX equipment parameter None Maximum SU-MIMO gain
Uplink STTD/MRC, SU-MIMO or MU-

G Div – UL Cell WiMAX equipment parameter dB
MIMO diversity gain
QoS
f Bias Scheduler parameter % QoS class bias factor
QoS Service parameter None QoS class of the service
p Service parameter None Service priority

B DL – Highest Service parameter None
the downlink
B UL – Highest Service parameter None
the uplink
B DL – Lowest Service parameter None
the downlink
B UL – Lowest Service parameter None
the uplink
UL
f Act Service parameter % Uplink activity factor
676

DL
f Act Service parameter % Downlink activity factor

TPD Min – UL Service parameter kbps
uplink
TPD Min – DL Service parameter kbps
downlink
TPD Max – UL Service parameter kbps
uplink
TPD Max – DL Service parameter kbps
downlink
UL Average requested throughput in the
uplink
DL Average requested throughput in the
downlink
TP Offset Service parameter kbps Throughput offset
L Body Service parameter dB Body loss

Min
N SC – UL Service parameter None Minimum number of subchannels
P Min Terminal parameter dBm Minimum terminal power allowed
P Max Terminal parameter dBm Maximum terminal power allowed
nf Terminal parameter dB Terminal noise figure
G Terminal parameter dB Terminal antenna gain
L Terminal parameter dB Terminal loss

N Ant – TX Terminal parameter None
in transmission
N Ant – RX Terminal parameter None
in reception
Max
G SU – MIMO Terminal WiMAX equipment parameter None Maximum SU-MIMO gain
Downlink STTD/MRC or SU-MIMO

G Div – DL Terminal WiMAX equipment parameter dB
diversity gain
G Div – Preamble Terminal WiMAX equipment parameter dB Preamble diversity gain
UL
G Div Clutter parameter dB Additional uplink diversity gain
DL
G Div Clutter parameter dB Additional downlink diversity gain
f SU – MIMO Clutter parameter None SU-MIMO gain factor
L Indoor Clutter parameter dB Indoor loss
L Path Propagation model result dB Path loss

Inter-technology downlink channel
F protection ratio for a frequency
ICP DL Network parameter None
offset F between the interfered
and interfering frequency channels
Monte Carlo simulations: Random result calculated from model
standard deviation
M Shadowing – Model dB Model shadowing margin
probability and model standard deviation
M Shadowing – C  I dB C/I shadowing margin
probability and C/I standard deviation
677
a. Any interfering cell whose signal to thermal noise ratio is less than CNR Min will be discarded.
10.2 Calculation Quick Reference

The following tables list the formulas used in calculations.

TX  ic  TX  ic  TX  ic   N TXi  ic  – N First – TXi  ic 

F Start – FB +  W Channel + ICS FB    -------------------------------------------------------
-
TX i  ic  i i i Channel Channel Start frequency for the channel
F Start  TX i  ic   MHz
 CN FB 
TX i  ic  TX i  ic  TX i  ic  End frequency for the channel

F End F Start + W Channel MHz
Min  F End  F End  – Max  F Start  F Start 
j i j i MHz Co-channel overlap bandwidth
W CCO
TX i  ic  – TX j  jc  W CCO
r CCO ----------------------------------- None Co-channel overlap ratio
TX i  ic 
W Channel
TX i  ic  – TX j  jc  TX  jc  TX  ic  TX  jc  TX  ic  TX  ic  Bandwidth of the lower-frequency

Min  F End  F Start  – Max  F Start  F Start – W Channel
j i j i i MHz
W ACO
L adjacent channel overlap
TX i  ic  – TX j  jc  W ACO Lower-frequency adjacent channel
L
r ACO ----------------------------------
- None
L TX i  ic  overlap ratio
W Channel
TX i  ic  – TX j  jc  TX j  jc  TX i  ic  TX  ic  TX  jc  TX  ic  Bandwidth of the higher-frequency

Min  F End  F End + W Channel – Max  F Start  F End 
i j i MHz
W ACO
H adjacent channel overlap
TX i  ic  – TX j  jc  W ACO Higher-frequency adjacent channel
H
r ACO ----------------------------------
- None
H TX  ic 
i
overlap ratio
W Channel
r ACO r ACO + r ACO None Adjacent channel overlap ratio
L H
TDD
r DL – Frame
----------------------- if interferer uses a TDD frequency band and victim uses None FDD – TDD overlap ratio
r FDD – TDD 100
an FDD frequency band, 1 otherwise
TX  ic 
 – f ACS – FB
i
 TX  ic  – TX  jc  TX  ic  – TX  jc  ---------------------------- TX  ic  – TX  jc 
r i j i
+ r ACO
j
 10
10  r i j
 CCO  FDD – TDD
 
 
TX i  ic  – TX j  jc  if W Channel  W Channel
rO None Total overlap ratio
TX  ic 
 – f ACS – FB
i
TX  ic 
 TX  ic  – TX  jc  TX  ic  – TX  jc  ---------------------------
- TX  ic  – TX  jc  W i
r i j i
+ r ACO
j
10
10  
i j Channel
 ---------------------
 CCO  r FDD – TDD TX j  jc 
  W Channel
 
if W Channel  W Channel
678
10.2.2 Preamble Signal Level Calculation

TX  ic  M
i i
TX  ic 
i EIRP Preamble – L Path – M Shadowing – Model – L Indoor + G
C Preamble Mi Mi Mi
dBm Received preamble signal level
–L – L Ant – L Body
Without smart antennas: P Preamble + G –L
EIRP Preamble dBm Preamble EIRP of a cell
+ 10  Log  E SA  +
Combining Div
P Preamble + G –L G SA + G SA
TX
L Path L Model + L Ant
i dB Path loss
TX i TX i Mi Mi
L Path + L + L Indoor + M Shadowing – Model – G +L –G
L Total Mi Mi
dB Total losses
+ L Ant + L Body
10.2.3 Preamble Noise Calculation

TX i  ic 
 TX  ic  N SCa – Preamble Preamble
TX i  ic   i
n 0 + 10  Log  F Sampling  ---------------------------------
-  f Segment dBm Preamble thermal noise for a cell
n 0 – Preamble TX i  ic  
 N SCa – Total 
Preamble 1
f Segment --- None Preamble segmenting factor
3
n Preamble n 0 – Preamble + nf dBm Preamble noise for a cell
10.2.4 Preamble Interference Calculation

TX j  jc  TX j  jc  TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  Total interference generated by an
Inter – Tech dBm
I Preamble C Preamble + f O + f Seg – Preamble + I DL interfering cell
TX i  ic  – TX j  jc  TX  ic  – TX  jc  Interference reduction factor due to
10  Log  r O 
i j dB
fO  the co- and adjacent channel overlap
TX  ic  – TX  jc  TX i  ic  – TX j  jc  Interference reduction factor due to

10  Log  p Collision 
i j dB
f Seg – Preamble   segmentation
TX i  ic  – TX j  jc  TX i  ic  TX j  jc  TX i  ic  TX j  jc  Preamble subcarrier collision
p Collision 1 if N Seg = N Seg and 0 if N Seg  N Seg None
probability
TX k
 --------------------------------------
-

Inter – Tech
I DL  W
k
F  TX  ic   TX  interference
i
TX k  ICP DL 
10.2.5 Preamble C/N Calculation

TX i  ic  TX i  ic  TX i  ic  Mi
DL dB Preamble C/N for a cell
CNR Preamble C Preamble – n Preamble + G Div – Preamble + G Div
679
10.2.6 Preamble C/(I+N) Calculation

TX  ic 
   TXj  jc   n
i  
TX i  ic     IPreamble -----------------------------
Preamble 
   ---------------------------- Inter – Tech  Inter – Tech

10
C Preamble – 10  Log 10 I + 10 + NR DL
TX i  ic     10  DL  
CINR Preamble   All TXj  jc     dB  Preamble C/(I+N) for a cell
     
Mi
DL
+ G Div – Preamble + G Div
TX  ic 
  TXj  jc   n
i 
TX  ic    IPreamble -----------------------------
Preamble
10  Log     +I  + NR Inter – Tech
i ---------------------------
- Inter – Tech 10 dBm Preamble Total Noise (I+N) for a cell
 I + N  Preamble   10
10
 DL
+ 10
 DL
 All TXj  jc    
   
10.2.7 Traffic and Pilot Signal Level Calculation (DL)

TX i  ic  Mi
TX i  ic  EIRP Traffic – L Path – M Shadowing – Model – L Indoor + G
C Traffic Mi Mi Mi
dBm Received traffic signal level
TX i  ic  Mi
TX i  ic  EIRP Pilot – L Path – M Shadowing – Model – L Indoor + G
C Pilot Mi Mi Mi
dBm Received pilot signal level
Array Combining Div dBm Traffic EIRP of a cell
EIRP Traffic P Traffic + G + G SA + G SA + G SA – L
Array Combining Div dBm Pilot EIRP of a cell
EIRP Pilot P Pilot + G + G SA + G SA + G SA – L
P Traffic P Preamble – P Traffic dBm Traffic transmission power of a cell
P Pilot P Preamble – P Pilot dBm Pilot transmission power of a cell
10.2.8 Traffic and Pilot Noise Calculation (DL)

Mi
 PZ DL 
 TXi  ic  N SCa – Used 
n 0 + 10  Log  F Sampling  -------------------------
TX i  ic 

 N SCa – Total

TX i  ic 
n 0 – DL With Segmentation: dBm Thermal noise for a cell
Mi
 PZ DL 
 TXi  ic  N SCa – Used 
n 0 + 10  Log  F Sampling  -------------------------
TX i  ic 
 f Segment – DL
 N SCa – Total 
 
3  PSG + 2  SSG
f Segment – DL 1 without and --------------------------------------------- with downlink segmentation None Downlink segmenting factor
15
n DL n 0 – DL + nf dBm Downlink noise for a cell
680
10.2.9 Traffic and Pilot Interference Calculation (DL)

 I j I
j 
 ----------------------------
Non – AAS
- -------------------
Idle
Monte Carlo Simulations: 10  Log  10 
10 10
+ 10
 
 
 
TX  jc 
 I j 
 ------------------
AAS 
-
without smart antennas, or 10  Log  10
TX j  jc  10  Total interference generated by an
with smart
I Total   dBm
interfering cell
 
 
antennas
 I j I
j
I
j 
 ----------------------------
Non – AAS
-
Idle
------------------- -------------------
AAS
Coverage Predictions: 10  Log  10 
10 10 10
+ 10 + 10
 
 
 
Monte Carlo Simulations:
TX j  jc  Mi Mi Mi Mi
EIRP Traffic – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body
TX j  jc  Coverage Predictions: Traffic interference power of an
I Traffic dBm
TX j  jc  interfering cell
EIRP Traffic – L Path – M Shadowing – Model + M Shadowing – C  I – L Indoor
Mi Mi Mi Mi
+G –L – L Ant – L Body
TX  jc  M M M M
j i i i i
EIRP Pilot – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body
TX j  jc  Coverage Predictions: Pilot interfering power of an
I Pilot dBm
TX  jc  interfering cell
j
EIRP Pilot – L Path – M Shadowing – Model + M Shadowing – C  I – L Indoor
Mi Mi Mi Mi
TX j  jc  TX j  jc  TX j TX j
EIRP Traffic P Traffic + G –L dBm Traffic EIRP of an interfering cell
EIRP Pilot P Pilot + G –L dBm Pilot EIRP of an interfering cell
TX  jc 
  I j TX j  jc 
 TX  jc  TX  jc   -------------------
Traffic
- N SCa – Data

10  Log TL DL
j 
 1 – AU DL
j  
 10
10
 ------------------------
-+
    TX j  jc 
Interference from the loaded part of
  N SCa – Used
TX j  jc    dBm
the frame transmitted using the
I Non – AAS TX j  jc  transmitter antenna of an interfering
I Pilot TX j  jc 

-------------------  N  cell
  1 – -------------------------  
10 SCa – Data
10
 TX j  jc  
 N SCa – Used  

TX j  jc  Mi Mi Mi Mi
EIRP AAS – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body
TX j  jc  Coverage Predictions: Interference power of an interfering
I AAS dBm
TX j  jc  cell transmitted using smart antenna
EIRP AAS – L Path – M Shadowing – Model + M Shadowing – C  I – L Indoor
Mi Mi Mi Mi
TX  jc  TX  jc  TX TX Traffic EIRP of an interfering cell
j j j j dBm
EIRP AAS P Traffic + G –L using smart antenna
681

Interference from empty part of the
TX  jc  TX  jc  M M M M frame transmitted using the
j j i i i i dBm
I Idle – Pilot EIRP Idle – Pilot – L Path – L Indoor + G –L – L Ant – L Body transmitter antenna of an interfering
cell
EIRP Idle – Pilot P Idle – Pilot + G –L dBm Idle pilot EIRP of an interfering cell
TX  jc 
  I j TX j  jc 
 Interference from the empty part of
TX j  jc 
 TX  jc  
Idle – Pilot- 
-----------------------------  the frame transmitted using the
N
10  Log  1 – TL DL   10  1 – ------------------------
SCa – Data  
j 10
I Idle -   dBm
TX j  jc  transmitter antenna of an interfering
   N SCa – Used   cell
  
TX i  ic  – TX j  jc  TX  ic  – TX  jc  Interference reduction factor due to

10  Log  r O 
i j dB
fO  the co- and adjacent channel overlap
TX  ic  – TX  jc  TX  ic  – TX  jc  Interference reduction factor due to

10  Log  p Collision – DL 
i j i j dB
f Seg – DL   downlink segmentation
TX k
 P 
 --------------------------------------
-

Inter – Tech DL – Rec
I DL  F  TX i  ic  TX k 
W
interference
TX k  ICP DL 
10.2.10 Traffic and Pilot C/N Calculation (DL)

TX i  ic  C Traffic – n DL
CNR Traffic dB Traffic C/N for a cell
TX i  ic  Mi
DL
With MIMO: CNR Traffic + G Div – DL + G Div
TX i  ic  C Pilot – n DL
CNR Pilot dB Pilot C/N for a cell
TX i  ic  Mi
DL
With MIMO: CNR Pilot + G Div – DL + G Div
10.2.11 Traffic and Pilot C/(I+N) Calculation (DL)

   TXj  jc   TX i  ic   

TX i  ic    IDL  n DL  
   ------------------  + I Inter – Tech + --------------------  + NR Inter – Tech

- -
C Traffic – 10  Log  10 10
TX i  ic  10  DL 10  DL 
CINR Traffic   All TX j  jc     dB  Traffic C/(I+N) for a cell
     
TX i  ic  Mi
DL
With MIMO: CINR Traffic + G Div – DL + G Div
   I j  n DL
i  
TX i  ic     ------------------
DL
-  ---------------------  
    +I  + NR 

C Pilot – 10  Log 10 +10
TX i  ic      DL  DL 
CINR Pilot   All TXj  jc     dB  Pilot C/(I+N) for a cell
     
TX  ic  M
i i DL
With MIMO: CINR Pilot + G Div – DL + G Div
  I j  n DL
i 
TX i  ic    ------------------
DL
- ---------------------
 I + N  DL 10  Log    10 10  + I Inter – Tech + 10 10  + NR Inter – Tech dBm Traffic Total Noise (I+N) for a cell
   DL  DL
 All TXj  jc    
   
682
10.2.12 Traffic Signal Level Calculation (UL)

M TX
i i
M
i EIRP UL – L Path – M Shadowing – Model – L Indoor + G
C UL TX i Mi Mi
dBm Received uplink signal level
Mi Mi Mi
P +G –L
Mi
EIRP UL Mi Mi Mi Mi dBm Uplink EIRP of a user equipment
With P = P Max without power control and P = P Eff after
power control
10.2.13 Traffic Noise Calculation (UL)

Mi
 PZ UL 
TX i  ic   TXi  ic  N SCa – Used 
n 0 – UL n 0 + 10  Log  F Sampling  -------------------------
TX i  ic 
 dBm Thermal noise for a cell
 N 
 SCa – Total
n UL n 0 – UL + nf dBm Uplink noise for a cell
10.2.14 Traffic Interference Calculation (UL)

M M TX  ic  – TX  jc  M TX  ic  – TX  jc 
j j i j j i j dBm Uplink interference received at a cell
I UL C UL + f O + f TL – UL + f Seg – UL
TX i  ic  – TX j  jc  TX i  ic  – TX j  jc  Interference reduction factor due to

fO 10  Log  r O  dB
 the co- and adjacent channel overlap

Mj M
10  Log  TL UL
j dB the interfering mobile’s uplink traffic
f TL – UL   load
TX  ic  – TX  jc  TX  ic  – TX  jc  Interference reduction factor due to
10  Log  p Collision – UL 
i j i j db
f Seg – UL   uplink segmentation
SC Com Uplink segmentation collision
------------------
- None
p Collision – UL TX i  ic 
SC probability
  IMj  TX i  ic 
  UL  non-seg M  n UL 
TX i  ic 
  --------------------------------------------i --------------------- TX i  ic 
 10  Non-segmented zone uplink noise at

Inter – Tech
NR UL 10  Log   10 10  + 10 + NR UL – n UL dB
 a cell without smart antennas
 All M j  
 
   
 All TX j  jc  
  IMj  TX i  ic 
  UL  n UL 
TX i  ic 
   seg M
 --------------------- TX i  ic 
i Segmented zone uplink noise at a
10  Log 
10 

--------------------------------- Inter – Tech
NRUL – Seg   10 10  + 10 
+ NR UL – n UL dB
cell without smart antennas
 All M j  
 
   
 All TX
j
 jc  
 I + N  UL NR UL + n UL or NR UL – Seg + n UL dBm Total Noise (I+N) for a cell
2
I UL    +  n  I Uplink noise at a cell with smart
NR UL    --------------------------------- dB
2
n  I antenna
683

TX  ic  2 Total Noise (I+N) for a cell in case of
 I + N  UL
i
 I UL    +  n  I dBm
smart antennas
10.2.15 Traffic C/N Calculation (UL)

Mi TX i  ic 
Mi C UL – n UL
CNR UL dB Uplink C/N at a cell
Mi TX i  ic 
UL
With MIMO: CNR UL + G Div – UL + G Div
10.2.16 Traffic C/(I+N) Calculation (UL)

Without smart antennas: CNR UL – NR UL or CNR UL – NR UL – Seg
CINR UL With smart antennas: CNR UL – NR UL  dB Uplink C/(I+N) at a cell
Mi TX i  ic 
UL
With MIMO: CINR UL + G Div – UL + G Div
10.2.17 Calculation of Total Cell Resources

TX  ic 
i 6
TX  ic 
i  W Channel  10 
F Sampling Floor  f Sampling  -----------------------------------
-  8000 Hz Sampling frequency
 8000 
TX i  ic  –3
TX i  ic  F Sampling  10
F -------------------------------------- kHz Inter-subcarrier distance
TX i  ic 
N SCa – Total
TX i  ic  1
------------------- ms Useful symbol duration
D Sym – Useful TX i  ic 
F
TX i  ic 
TX i  ic  r CP
D CP --------------
- ms Cyclic prefix duration
F
D Symbol D Sym – Useful + D CP ms Symbol duration
Used TDD TDD

D Frame D Frame – D TTG – D RTG ms Used frame duration
 D Used 
Floor  -----------------
TX i  ic  Frame  Frame duration in terms of symbol
- SD
N  SD – Used   Frame  TXi  ic   durations
 D Symbol
If DL:UL ratio is defined in percentage:

TX  ic 
RoundUp  N  SD – Used   Frame  r DL – Frame – O Fixed
i TDD DL
TX  ic  Downlink subframe duration in

i SD
N  SD – DL   Subframe If DL:UL ratio is defined in fraction: terms of symbol durations
TDD
 TXi  ic  N SD – DL  DL
RoundUp  N SD – Used   Frame  ------------------------------------------
TDD TDD
 – O Fixed
 N SD – DL + N SD – UL
684
TX  ic 
R DL
i
=  M
i DL  Total downlink cell resources, i.e.,
 TXi  ic  PZ O Variable 
Floor  N  SD – DL   Subframe  N SCa – Data   1 – --------------------
DL Symbols the number of symbols in the
TX  ic 
- 
i  100  downlink subframe
N  Sym – DL   Subframe
 
If DL:UL ratio is defined in percentage:
TX  ic 
RoundDown  N SD – Used   Frame   1 – r DL – Frame  – O Fixed
i TDD UL
TX  ic  Uplink subframe duration in terms of

i SD
N  SD – UL   Subframe If DL:UL ratio is defined in fraction: symbol durations
TDD
 TX i  ic  N SD – UL  UL
RoundDown  N SD – Used   Frame  ------------------------------------------
TDD TDD
 – O Fixed
TX i  ic 
R UL =  M
i UL  Total uplink cell resources, i.e., the
 TXi  ic  PZ UL
 O Variable 
TX i  ic 
Floor  N SD – UL   Subframe  N SCa – Data   1 – ---------------------  Symbols number of symbols in the uplink
 100  subframe
N  Sym – UL   Subframe
 
10.2.18 Channel Throughput, Cell Capacity, Allocated Bandwidth

Throughput, and Per-user Throughput Calculation
TX i  ic 
R DL  M
i
B
DL
---------------------------------
D Frame
TX  ic 
i
R DL  M
i
B DL TX  ic 
i
For proportional fair schedulers: ---------------------------------  G MUG – DL
D Frame
Mi M
i Downlink peak MAC channel
CTP P – DL With downlink segmentation: CTP P – DL  f Segment – DL kbps
throughput
With MIMO (SU-MIMO):
Max
 Mi =  Mi   1 + f SU – MIMO  G SU – MIMO – 1  
B DL B DL
Max
With MIMO (AMS):  Mi =  Mi   1 + f SU – MIMO  G SU – MIMO – 1  
B DL B DL
if CNR Preamble  T AMS or CINR Preamble  T AMS
M M M Downlink effective MAC channel

CTP P – DL   1 – BLER  B DL 
i i i kbps
CTP E – DL    throughput
Mi
Mi Mi f TP – Scaling Mi Downlink application channel
CTP A – DL CTP E – DL  ------------------------
100 throughput
Cap P – DL CTP P – DL  TL DL – Max kbps Downlink peak MAC cell capacity
M M M
Cap P – DL   1 – BLER  BDL 
i i i kbps Downlink effective MAC cell capacity
Cap E – DL   
Mi
Mi Mif TP – Scaling Mi
Cap A – DL Cap E – DL  ------------------------
- – TP Offset kbps Downlink application cell capacity
100
Mi
M Cap P – DL Downlink peak MAC throughput per
i
PUTP P – DL -----------------------
- kbps
TX i  ic  user
N Users – DL
685

M
i
M Cap E – DL Downlink effective MAC throughput
i
PUTP E – DL -----------------------
- kbps
TX  ic  per user
i
N Users – DL
M
Mi i Downlink application throughput per
f TP – Scaling
Mi Mi
kbps
PUTP A – DL PUTP E – DL  ------------------------
- – TPOffset user
100
TX  ic 
i
R UL  Mi
B UL
---------------------------------
D Frame
TX i  ic 
R UL  Mi
B TX  ic 
UL i
For proportional fair schedulers: ---------------------------------  G MUG – UL
D Frame
With MIMO (SU-MIMO):
Max
Mi  M =  M   1 + f SU – MIMO  G SU – MIMO – 1   Uplink peak MAC channel
CTP P – UL i
B UL
i
B UL kbps
throughput
Max
With MIMO (AMS):  Mi =  Mi   1 + f SU – MIMO  G SU – MIMO – 1  
B B
UL UL
if CNR Preamble  T AMS or CINR Preamble  T AMS
With MIMO (MU-MIMO) in uplink throughput coverage predictions:
TX i  ic 
R UL  Mi
B UL TX  ic 
---------------------------------  G MUi – MIMO
D Frame
Mi M M Uplink effective MAC channel

CTP P – UL   1 – BLER  B UL 
i i kbps
CTP E – UL throughput
M
M i Uplink application channel
i Mif TP – Scaling Mi
kbps
CTP A – UL CTP E – UL  ------------------------- – TP Offset throughput
100
M M TX  ic 
i i i kbps Uplink peak MAC cell capacity
Cap P – UL CTP P – UL  TL UL – Max
Mi M M
Cap P – UL   1 – BLER  B UL 
i i kbps Uplink effective MAC cell capacity
Cap E – UL
M
M i
i Mif TP – Scaling Mi
kbps Uplink application cell capacity
Cap A – UL Cap E – UL  ------------------------- – TP Offset
100
M
i
Mi N SC – UL
Mi CTP P – UL  ----------------- M kbps
Uplink peak MAC allocated
ABTP P – UL i bandwidth throughput
PZ UL
N SC
Mi M M Uplink effective MAC allocated

ABTP P – UL   1 – BLER  B UL 
i i kbps
ABTPE – UL    bandwidth throughput
Mi
Mi M f TP – Scaling M Uplink application allocated
ABTP A – UL i
ABTP E – UL  ------------------------
i
100 bandwidth throughput
 Cap Mi M 
Min  -----------------------
- ABTP P –i UL
M
i P – UL Uplink peak MAC throughput per
PUTP P – UL  TXi  ic   kbps
user
 Cap Mi M 
Min  -----------------------
M
i E – UL Uplink effective MAC throughput per
PUTP E – UL TX i  ic   kbps
user
686

M
M i Uplink application throughput per
i M
i f TP – Scaling M
i kbps
PUTP A – UL PUTP E – UL  ------------------------- – TP Offset user
100

Sel
Mi
Sel TPD Min – DL Resources allocated to a mobile to
R Min – DL Sel
CTP P – DL
Sel
M
Sel TPD Min – UL
i Resources allocated to a mobile to
M
i --------------------------- None satisfy its minimum throughput
R Min – UL M
Sel
i demand in uplink
CTP P – UL
Sel
TX i  ic  Mi Remaining downlink cell resources
TX i  ic 
R Rem – DL
TL DL – Max –  R Min – DL None after allocation for minimum
Sel throughput demands
Mi
Sel
TX i  ic  Mi Remaining uplink cell resources after
TX i  ic 
R Rem – UL
TL DL – Max –  R Min – UL None allocation for minimum throughput
Sel demands
M
i
Sel Sel Sel Remaining throughput demand for a

Mi Mi Mi
kbps
TPD Rem – DL TPD Max – DL – TPD Min – DL mobile in downlink
Sel Sel Sel Remaining throughput demand for a
Mi Mi Mi
kbps
TPD Rem – UL TPD Max – UL – TPD Min – UL mobile in uplink
Sel Sel Downlink peak channel throughput

CTP P – DL  G MUG – DL kbps with multi-user diversity gain
CTP P – DL Without MUG (Proportional Fair)
Sel Sel Uplink peak channel throughput with

Mi M TX  ic 
i i kbps multi-user diversity gain
CTP P – UL CTP P – UL  G MUG – UL
Without MUG (Proportional Fair)
Sel
Mi
Sel TPD Rem – DL Remaining resource demand for a
Mi ---------------------------- None
RD Rem – DL Sel
Mi mobile in downlink
CTP P – DL
Sel
M
i
Sel TPD Rem – UL Remaining resource demand for a
M
i ---------------------------- None
RD Rem – UL M
Sel mobile in uplink
i
CTP P – UL
687
Sel TX  ic 
i
Proportional Fair: Min  RD Rem – DL --------------------
-
 N 
Sel
Mi
TX i  ic 
RD Rem – DL
Proportional Demand: R Eff – Rem – DL  ----------------------------------
Sel
-
Mi
 RDRem – DL
Sel
M
i
Sel Resources allocated to a mobile to
M TX i  ic 
i
 Mi
Sel
R QoS – DL None satisfy its maximum throughput
R Max – DL
Biased (QoS Class): Min  RD Rem – DL -------------------- demand in downlink
 N QoS 
Sel
Mi
TPD Rem – DL
Max Aggregate Throughput: ---------------------------
Sel
-
Mi
CTP P – DL
Sel TX i  ic 
-
 N 
Sel TX i  ic 
Proportional Fair: Min  RD Rem – UL --------------------
-
 N 
Sel
Mi
TX i  ic 
RD Rem – UL
Proportional Demand: R Eff – Rem – UL  ----------------------------------
Sel
-
Mi
 RDRem – UL
Sel
M
i
Sel Resources allocated to a mobile to
M TX i  ic 
i
 R QoS – UL
Sel
Mi None satisfy its maximum throughput
R Max – UL
Biased (QoS Class): Min  RD Rem – UL -------------------
- demand in uplink
 N QoS 
Sel
Mi
TPD Rem – UL
Max Aggregate Throughput: ---------------------------
Sel
-
Mi
CTP P – UL
Sel TX i  ic 
Round Robin: Min  RD Rem – UL --------------------
-
 N 
 Sel 
 TXi  ic  Mi  Effective remaining downlink
TX i  ic 
R Eff – Rem – DL
Min  R Rem – DL
 
RD Rem – DL

None resources in a cell
 M
i
 Sel 
 TX i  ic  Mi  Effective remaining uplink resources
TX  ic 
i
R Eff – Rem – UL
Min  R Rem – UL
 
RD Rem – UL

None in a cell
 Sel
Mi  (Proportional Demand)
Sel Sel Sel

Mi Mi Mi
QoS
f Bias R Max – ErtPS R Max – rtPS R Max – nrtPS QoS class bias
 1 + ---------
- = -------------------------- = -------------------------- = -------------------------- None
100 Mi
Sel
Mi
Sel
Mi
Sel (Biased (QoS Class))
R Max – rtPS R Max – nrtPS R Max – BE
r
1 QoS
N QoS   --- Remaining downlink cell resources
TX i  ic  
TX i  ic  R Rem – DL  ------------------------------------------------------- None
after allocation for minimum
R QoS – DL r throughput demands for a QoS class
1 QoS
 N QoS   --- (Biased (QoS Class))

All QoS
688

r
1 QoS
N QoS   --- Remaining downlink cell resources
TX  ic 
i 
TX  ic 
i R Rem – UL  ------------------------------------------------------- None
after allocation for minimum
R QoS – UL r throughput demands for a QoS class
1 QoS
N QoS   ---
 (Biased (QoS Class))
 
All QoS
  Mi
Sel
Mi 
Sel 
 
  
 Sel
M i  Site

Max  1 -------------------------------------------------------------------------------------------------------
Site Site backhaul overflow ratio in
BHOF DL  None
Sel Sel  downlink
  Mi Mi  

Site
 TP BH – DL –  R Min – DL  CTP E – DL 
  
 Sel
M i  Site 
  Mi
Sel Sel
Mi  
 
   Max – UL

R  CTP E – UL 
 
 Sel
 Site

Max  1 -------------------------------------------------------------------------------------------------------
Site M
i
BHOF DL None Site backhaul overflow ratio in uplink
Sel Sel 
  M M  

Site i i
 TP BH – UL –  R Min – UL  CTP E – UL 
  
 M
Sel
i
 Site 
Sel Sel
Mi Mi
 Mi
Sel
Mi 
Sel R 
Sel Sel  R Min – DL  CTP P – DL +  ------------------------------------------------ Site 
Mi Mi    BHOF DL  None in downlink
TL DL = R DL ------------------------------------------------------------------------------------------------------------------
Sel (Downlink traffic load of the mobile)
Mi
CTP P – DL
Sel Sel
Mi Mi
 Mi
Sel
Mi 
Sel R 
Sel Sel  R Min – UL  CTP P – UL +  ------------------------------------------------ Site
-

Mi Mi    BHOF UL  None in uplink
TL UL = R UL ------------------------------------------------------------------------------------------------------------------
Sel (Uplink traffic load of the mobile)
Mi
CTP P – UL
10.2.20 User Throughput Calculation

Sel Sel Sel
Mi Mi Mi
kbps Downlink peak MAC user throughput
UTP P – DL R DL  CTP P – DL
Sel Sel
Sel
Mi M
i   Mi   Downlink effective MAC user
UTP E – DL UTP P – DL   1 – BLER  B DL   kbps
throughput
  
Sel
Sel Mi
M Sel Sel Downlink application user
i Mif TP – Scaling Mi kbps
UTP A – DL UTP E – DL  ------------------------- – TP Offset throughput
100
Sel Sel Sel
Mi Mi Mi
kbps Uplink peak MAC user throughput
UTP P – UL R UL  CTP P – UL
Sel Sel
Sel
Mi Mi   Mi   Uplink effective MAC user
UTP E – UL UTP P – UL   1 – BLER  B UL   kbps
throughput
  
Sel
Sel Sel Mi Sel
Mi Mif TP – Scaling Mi kbps Uplink application user throughput
UTP A – UL UTP E – UL  ------------------------
- – TP Offset
100
689
10.3 Available Calculations

10.3.1.1 Profile View
The point analysis profile view displays the following calculation results for the selected transmitter based on the calculation
algorithm described in "Preamble Signal Level Calculation" on page 707.
TX  ic 
i
• Preamble signal level C Preamble
• Path loss L Path
• Total losses L Total
M M M M
i i i i
L ,G , L Ant , and L Body are not used in the calculations performed for the profile view.
10.3.1.2 Reception View

Analysis provided in the reception view is based on path loss matrices. So, you can display received signal levels from the cells
for which calculated path loss matrices are available. For each cell, Atoll displays the received preamble, pilot, or traffic signal
level or C/N.
Reception level bar graphs show the signal levels or C/N in decreasing order. The maximum number of bars in the graph
depends on the preamble signal level of the best server. The bar graph displays cells whose received preamble signal levels
are higher than their preamble C/N thresholds and are within a 30 dB margin from the highest preamble signal level.
You can use a value other than 30 dB for the margin from the highest preamble signal level, for example a smaller value for
improving the calculation speed. For more information on defining a different value for this margin, see the Administrator
Manual.
10.3.1.3 Interference View

Analysis provided in the interference view is based on path loss matrices. So, you can display the received signal level from
the best server and interfering signal levels from other cells for which calculated path loss matrices are available. For each cell,
Atoll displays the best server preamble, pilot, or traffic signal level, and interference from other cells.
Interference level bar graphs show the interference levels on different channels in decreasing order. The maximum number
of bars in the graph depends on the highest interference level on the studied channel. The bar graph displays cells whose C/
N are higher than the minimum interferer C/N threshold and whose interference levels are within a 30 dB margin from the
highest interference level on the studied channel.
You can use a value other than 30 dB for the margin from the highest interference level in the studied channel, for example a
smaller value for improving the calculation speed. For more information on defining a different value for this margin, see the
10.3.1.4 Details View

Analysis provided in the details view is based on path loss matrices. So, you can display the received signal level from the best
server and interfering signal levels from other cells for which calculated path loss matrices are available. For each cell, Atoll
displays the best server preamble signal level and interference from other cells.
The results for the best server (first row) are displayed using bold italic characters. Other cells are listed in the decreasing order
of preamble signal level. All the cells from which the received preamble signal level is higher than their preamble C/N
thresholds are listed in the table. As well, interference values are listed for all the cells whose C/N are higher than the
minimum interferer C/N threshold and whose interference levels are within a 30 dB margin from the highest interference
level on the preamble.
You can use a value other than 30 dB for the margin from the highest interference level on the preamble, for example a
smaller value for improving the calculation speed. For more information on defining a different value for this margin, see the
690
10.3.2 Coverage Predictions

10.3.2.1 Preamble Signal Level Coverage Predictions
The following coverage predictions are based on the received preamble signal levels:
• Coverage by Transmitter
• Coverage by Signal Level
• Overlapping Zones
For these calculations, Atoll calculates the received preamble signal level. Then, Atoll determines the selected display
parameter on each pixel inside the cell’s calculation area. Each pixel within the calculation area of TXi(ic) is considered a non-
interfering receiver. For these calculations, the best server calculation is always based on preamble signal level.
Mi Mi Mi Mi
L ,G , L Ant , and L Body are not considered in the calculations performed for the preamble signal level based coverage
predictions.
For more information on preamble signal level calculations, see "Preamble Signal Level Calculation" on page 707
areas to display. There are three possibilities.
• All Servers
MinimumThreshold  C Preamble  or L Total or L Path   MaximumThreshold
i i i
 
• Best Signal Level and a Margin

MinimumThreshold  C Preamble  or L Total or L Path   MaximumThreshold
i i i
 
AND
C Preamble  Best  C Preamble – M
ji  
• If M = 0 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is the highest.
• If M = 2 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is either the highest or
within a 2 dB margin from the highest.
• If M = -2 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is 2 dB higher than the
received preamble signal levels from the cells which are 2nd best servers.
• Second Best Signal Level and a Margin
MinimumThreshold  C Preamble  or L Total or L Path   MaximumThreshold
i i i
AND
691
TX  ic  nd TX  jc 
C Preamble  2 Best  C Preamble – M
i j
 
ji
Where M is the specified margin (dB). The 2nd Best function considers the second highest value from a list of values.
• If M = 0 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is the second highest.
• If M = 2 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is either the second highest
or within a 2 dB margin from the second highest.
received preamble signal levels from the cells which are 3rd best servers.
It is possible to display the coverage predictions with colours depending on any transmitter or cell attribute, and other criteria
such as:
• Signal Level (dBm, dBµV, dBµV/m)
• Best Signal Level (dBm, dBµV, dBµV/m): Where cell coverage areas overlap, Atoll keeps the highest value of the signal
level.
• Path Loss (dB)
• Total Losses (dB)
• Best Server Path Loss (dB): Where cell coverage areas overlap, Atoll determines the best cell (i.e., the cell with the
highest preamble signal level) and evaluates the path loss from this cell.
• Best Server Total Losses (dB): Where cell coverage areas overlap, Atoll determines the best cell (i.e., the cell with the
highest preamble signal level) and evaluates the total losses from this cell.
• Number of Servers: Atoll evaluates the number of cells that cover a pixel (i.e., the pixel falls within the coverage areas
of these cells).
10.3.2.2 Effective Signal Analysis Coverage Predictions

The following coverage predictions are based on the received preamble, traffic, or pilot signal levels and noise, and take into
Mi Mi Mi Mi
account the receiver characteristics ( L ,G , L Ant , and L Body ) when calculating the required parameter:
• Effective Signal Analysis (DL)

• Effective Signal Analysis (UL)
For these calculations, Atoll calculates the received signal level or C/N level at each pixel for the channel type being studied,
i.e., preamble, traffic, or pilot. Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. The
properties of the non-interfering probe receiver are set by selecting a terminal, a mobility type, and a service.
For more information on signal level calculations, see:
• "Preamble Signal Level Calculation" on page 707.
• "Traffic and Pilot Signal Level Calculation (DL)" on page 715.
• "Traffic Signal Level Calculation (UL)" on page 729
For more information on permutation zone selection, see "Permutation Zone Selection" on page 714.
For more information on C/N level calculations, see:
• "Preamble C/N Calculation" on page 712.
• "Traffic and Pilot C/N Calculation (DL)" on page 726
• "Traffic C/N Calculation (UL)" on page 734.
692
page 713.
It is possible to display the Effective Signal Analysis (DL) coverage prediction with colours depending on the following display
options:
• Preamble Signal Level (DL) (dBm)
• Pilot Signal Level (DL) (dBm)
• Traffic Signal Level (DL) (dBm)
• Preamble C/N Level (DL) (dB)
• Pilot C/N Level (DL) (dB)
• Traffic C/N Level (DL) (dB)
• Permutation Zone (DL)
• Segment
It is possible to display the Effective Signal Analysis (UL) coverage prediction with colours depending on the following display
options:
• Signal Level (UL) (dBm)
• C/N Level (UL) (dB)
• Permutation Zone (UL)
10.3.2.3 C/(I+N)-based Coverage Predictions

The following coverage predictions are based on the received signal levels, total noise, and interference.
• Coverage by C/(I+N) Level (DL)
• Service Area Analysis (DL)
• Coverage by Quality Indicator (DL)
• Coverage by C/(I+N) Level (UL)
• Service Area Analysis (UL)
• Coverage by Throughput (UL)
• Coverage by Quality Indicator (UL)
Mi Mi Mi Mi
These coverage predictions take into account the receiver characteristics ( L ,G , L Ant , and L Body ) when calculating the
required parameter. For these calculations, Atoll calculates the received signal level, noise, and interference at each pixel.
Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. The properties of the non-interfering
probe receiver are set by selecting a terminal, a mobility type, and a service.
The downlink coverage predictions are based on the downlink traffic loads of the cells, and the uplink coverage predictions
are based on the uplink noise rise values. These parameters can either be calculated by Atoll during the Monte Carlo
simulations, or set manually by the user for all the cells.
For more information on C/(I+N), (I+N), and bearer calculations, see:
• "Preamble C/(I+N) Calculation" on page 712.
• "Traffic and Pilot C/(I+N) and Bearer Calculation (DL)" on page 727.
• "Traffic C/(I+N) and Bearer Calculation (UL)" on page 737.
• "Noise Rise Calculation (UL)" on page 733
For more information on thoughput calculations, see:
• "Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-User Throughput Calculation" on
page 743.
693
page 713.
It is possible to display the Coverage by C/(I+N) Level (DL) coverage prediction with colours depending on the following display
options:
• Preamble C/(I+N) Level (DL) (dB)
• Preamble Total Noise (I+N) (DL) (dBm)
• Traffic C/(I+N) Level (DL) (dB)
• Traffic Total Noise (I+N) (DL) (dBm)
• Pilot C/(I+N) Level (DL) (dB)
It is possible to display the Service Area Analysis (DL) coverage prediction with colours depending on the following display
options:
• Bearer (DL)
• Modulation (DL): Modulation used by the bearer
• Service
It is possible to display the Coverage by Throughput (DL) coverage prediction with colours depending on the following display
options:
• Peak MAC Channel Throughput (DL) (kbps)
• Effective MAC Channel Throughput (DL) (kbps)
• Application Channel Throughput (DL) (kbps)
• Peak MAC Cell Capacity (DL) (kbps)
• Effective MAC Cell Capacity (DL) (kbps)
• Application Cell Capacity (DL) (kbps)
• Peak MAC Throughput per User (DL) (kbps)
• Effective MAC Throughput per User (DL) (kbps)
• Application Throughput per User (DL) (kbps)
It is possible to display the Coverage by Quality Indicator (DL) coverage prediction with colours depending on the following
display options:
• Quality indicators available in the document (Quality Indicators table): Atoll calculates the downlink traffic C/(I+N)
levels received from the best serving cells at each pixel of their coverage areas. From the C/(I+N), Atoll determines the
best bearer available on each pixel. Then, for the calculated C/(I+N) and bearer, it determines the value of the selected
quality indicator from the quality graphs defined in the WiMAX equipment of the selected terminal.
It is possible to display the Coverage by C/(I+N) Level (UL) coverage prediction with colours depending on the following display
options:
• C/(I+N) Level (UL) (dB)
• Total Noise (I+N) (UL) (dBm)
• Allocated Bandwidth (UL) (No. of Subchannels)
• C/(I+N) Level for 1 Subchannel (UL) (dB)
• Transmission Power (UL) (dBm)
It is possible to display the Service Area Analysis (UL) coverage prediction with colours depending on the following display
options:
• Bearer (UL)
• Modulation (UL): Modulation used by the bearer
• Service
It is possible to display the Coverage by Throughput (UL) coverage prediction with colours depending on the following display
options:
• Peak MAC Channel Throughput (UL) (kbps)
• Effective MAC Channel Throughput (UL) (kbps)
• Application Channel Throughput (UL) (kbps)
• Peak MAC Cell Capacity (UL) (kbps)
• Effective MAC Cell Capacity (UL) (kbps)
• Application Cell Capacity (UL) (kbps)
• Peak MAC Allocated Bandwidth Throughput (UL) (kbps)
• Effective MAC Allocated Bandwidth Throughput (UL) (kbps)
694
• Application Allocated Bandwidth Throughput (UL) (kbps)

• Peak MAC Throughput per User (UL) (kbps)
• Effective MAC Throughput per User (UL) (kbps)
• Application Throughput per User (UL) (kbps)
It is possible to display the Coverage by Quality Indicator (UL) coverage prediction with colours depending on the following
display options:
• Quality indicators available in the document (Quality Indicators table): Atoll calculates the uplink traffic C/(I+N) levels
received at the best serving cells from each pixel of their coverage areas. From the C/(I+N), Atoll determines the best
bearer available on each pixel. Then, for the calculated C/(I+N) and bearer, it determines the value of the selected
quality indicator from the quality graphs defined in the WiMAX equipment of the best serving cell.
10.3.2.4 Cell Identifier Collision Zones Coverage Prediction

The Cell Identifier Collision Zones coverage prediction is based on the received preamble signal levels. Atoll calculates the
received preamble signal level then Atoll determines the selected display parameter on each pixel inside the cell’s calculation
M M M M
i i i i
area. Each pixel within the calculation area of TXi(ic) is considered a non-interfering receiver. L ,G , L Ant , and L Body are
not considered in the calculations.
For more information on preamble signal level calculations, see "Preamble Signal Level Calculation" on page 707
areas to display. It is possible to determine the coverage area based on the best signal level. The coverage area of each cell
TXi(ic) corresponds to the pixels where:
TX i  ic  TX i  ic  TX i  ic  TX i  ic  TX j  jc 
MinimumThreshold  C Preamble  or L Total or L Path   MaximumThreshold AND C Preamble  Best  C Preamble – M
  ji  
• If M = 0 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is the highest.
• If M = 2 dB, Atoll considers pixels where the received preamble signal level from TXi(ic) is either the highest or within
a 2 dB margin from the highest.
received preamble signal levels from the cells which are 2nd best servers.
It is possible to display the coverage predictions with colours per cell or:
• Number of interferers
• Number of interferers per cell
10.3.3 Calculations on Subscriber Lists

When calculations are performed on a list of subscribers by running the Automatic Server Allocation, Atoll calculates the path
loss again for the subscriber locations and heights because the subscriber heights can be different from the default receiver
height used for calculating the path loss matrices.
Atoll calculates the following parameters for each subscriber in the list whose Lock Status is set to None.
• Serving Base Station and Reference Cell as described in "Best Server Determination" on page 713.
695
Atoll calculates the following parameters for each subscriber in the list that has a serving base station assigned and whose
Lock Status is set to None or Server.
• Azimuth (  ): Angle with respect to the north for pointing the subscriber terminal antenna towards its serving base
station.
• Mechanical Downtilt (  ): Angle with respect to the horizontal for pointing the subscriber terminal antenna towards
its serving base station.
Atoll calculates the remaining parameters for each subscriber in the list that has a serving base station assigned, using the
properties of the default terminal and service. For more information, see:
• "Permutation Zone Selection" on page 714.
• "Traffic Signal Level Calculation (UL)" on page 729.
• "Noise Rise Calculation (UL)" on page 733.
• "Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and Per-User Throughput Calculation" on
page 743.
10.3.4 Monte Carlo Simulations

10.3.4.1 User Distribution

• "Simulations Based on User Profile Traffic Maps and Subscriber Lists" on page 696.
[Simulation]
RandomTotalUsers=0
10.3.4.1.1 Simulations Based on User Profile Traffic Maps and Subscriber Lists
Fixed subscribers listed in subscriber lists have a user profile assigned to each of them.
696
points.
duration of each voice call, or the volume of the data transfer in the uplink and the downlink in each data session.
Voice Service (v)
User profile parameters for voice type services are:

• The average number of calls per hour N Call .
N Call  D Call
Calculation of the service usage duration per hour ( p 0 : probability of an active call): p 0 = ----------------------------
-
3600
Calculation of the number of users trying to access the service v ( n v ): n v = N Users  p 0
UL DL
factors defined for the voice type service v, f Act and f Act .

UL DL
UL UL DL
DL DL UL
UL + DL UL DL
Number of inactive users: n v – Inactive = n v  p Inactive
UL UL
Number of users active in the uplink: n v – Active = n v  p Active
DL DL
Number of users active in the downlink: n v – Active = n v  p Active
UL + DL UL + DL
Number of users active in the uplink and downlink both: n v – Active = n v  p Active
Therefore, a user can be either active on both links, inactive on both links, active on UL only, or active on DL only.
Data Service (d)
User profile parameters for data type services are:

• The average number of data sessions per hour N Session .
DL UL
• The average data volume (in kBytes) transferred in the downlink V and the uplink V during a session.
DL UL
• The average throughputs in the downlink TP Average and the uplink TP Average for the service d.
UL DL
UL N Session  V  8 DL N Session  V  8
Calculation of activity probabilities: f = ------------------------------------------
- and f = ------------------------------------------
-
UL DL
TP Average  3600 TP Average  3600
UL DL
Probability of being inactive: p Inactive =  1 – f    1 – f 
697
UL UL DL
Probability of being active in the uplink: p Active = f  1 – f 
DL DL UL
Probability of being active in the downlink: p Active = f  1 – f 
UL + DL UL DL
Probability of being active in the uplink and downlink both: p Active = f f
Calculation of number of users:
Number of inactive users: n d – Inactive = N Users  p Inactive
UL UL
Number of users active in the uplink: n d – Active = N Users  p Active
DL DL
Number of users active in the downlink: n d – Active = N Users  p Active
UL + DL UL + DL
Number of users active in the uplink and downlink both: n d – Active = N Users  p Active
Calculation of the number of active users trying to access the service d (nd):
UL DL UL + DL
n d = n d – Active + n d – Active + n d – Active
are average distributions. The service and the activity status of each user are randomly
drawn in each simulation. Therefore, if you calculate several simulations at once, the
average number of users per service and average numbers of inactive, active on UL,
active on DL and active on UL and DL users, respectively, will correspond to calculated
distributions. But if you check each simulation, the user distribution between services as
well as the activity status distribution between users can be different in each of them.

Sector traffic maps per sector are also referred to as live traffic maps. Live traffic data from the OMC is spread over the best
server coverage areas of the transmitters included in the traffic map. Either throughput demands per service or the number
of active users per service are assigned to the coverage areas of each transmitter.
For each transmitter TXi and each service s,
• Sector Traffic Maps (Throughputs)
Atoll calculates the number of active users of each service s on UL and DL in the coverage area of TXi as follows:
UL DL
UL TP Cell DL TP Cell
N = ----------------------
- and N = ----------------------
-
UL DL
TP Average TP Average
UL
Where TP Cell is the total uplink throughput demand defined in the map for any service s for the coverage area of the
DL
transmitter, TP Cell is the total downlink throughput demand defined in the map for any service s for the coverage
UL DL
area of the transmitter, TP Average is the average uplink requested throughput of the service s, and TP Average is the
average downlink requested throughput of the service s.
• Sector Traffic Maps (# Active Users)
UL DL
Atoll directly uses the defined N and N values, i.e., the number of active users on UL and DL in the transmitter
coverage area using the service s.
At any given instant, Atoll calculates the probability for a user being active in the uplink and in the downlink as follows:
UL DL
Users active in the uplink and downlink both are included in the N and N values. Therefore, it is necessary to
UL DL UL + DL
accurately determine the number of active users in the uplink ( n Active ), in the downlink ( n Active ), and both ( n Active ).
As for the other types of traffic maps, Atoll considers both active and inactive users.
UL DL
factors defined for the service, f Act and f Act .
698
UL DL
UL UL DL
DL DL UL
UL + DL UL DL
Calculation of the number of active users trying to access the service:

UL UL UL + DL DL DL UL + DL
We have: N =  p Active + p Active   n and N =  p Active + p Active   n
Where, n is the total number of active users in the transmitter coverage area using the service.
UL UL + DL DL UL + DL
UL + DL  N  p Active N  p Active 
Number of users active in the uplink and downlink both: n Active = Min  ---------------------------------------
UL + DL
 ---------------------------------------
+ DL
 or
UL DL
Active 
UL + DL UL DL DL UL
simply, n Active = Min  N  f Act N  f Act 
UL UL UL + DL
Number of users active in the uplink: n Active = N – n Active
DL DL UL + DL
Number of users active in the downlink: n Active = N – n Active
UL DL UL + DL
And, n = n Active + n Active + n Active
Calculation of the number of inactive users attempting to access the service:
nv
-  p Inactive
Number of inactive users: n Inactive = ---------------------------
1 – p Inactive
simulation, the activity status of each user is randomly drawn. Therefore, if you calculate
simulation, the activity status distribution between users can be different in each of
them.
10.3.4.2 Simulation Process

WiMAX cells include intelligent schedulers and radio resource management features for regulating network traffic loads,
optimising spectral efficiency, and satisfying the QoS demands of the users. Each Monte Carlo simulation in the Atoll WiMAX
BWA module is a snap-shot of the network with resource allocation carried out over a duration of 1 second. The number of
WiMAX frames in 1 second depends on the selected frame duration, D Frame . The steps of this algorithm are listed below.
The simulation process can be summed up into the following iterative steps.
1. Generates mobiles according to the input traffic data as explained in "User Distribution" on page 696.
2. Sets initial values for the following parameters:
• Cell transmission powers and reductions ( P Preamble , P Traffic , P Pilot , and P Idle – Pilot ) are set to the values
defined by the user.
Mi
• Mobile transmission power is set to the maximum mobile power ( P Max ).
• Cell loads ( TL DL , TL UL , NRUL , NR UL – Seg , SU DL , and AU DL ) are set to their current values in the
Cells table.
3. Determines the best servers for all the mobiles generated for the simulation as explained in "Best Server
699
Figure 10.1: WiMAX Simulation Algorithm
For each iteration k, the simulation process,

4. Determines the mobiles which are within the service areas of their best serving cells as explained in "Service Area
5. Determines the permutation zone assigned to each mobile as explained in "Permutation Zone Selection" on page 714.
6. Determines the downlink and uplink traffic C/(I+N) and bearers for each of these mobiles as explained in "Traffic and
Pilot C/(I+N) and Bearer Calculation (DL)" on page 727 and "Traffic C/(I+N) and Bearer Calculation (UL)" on page 737
respectively.
7. Determines the channel throughputs at the mobile as explained in "Channel Throughput, Cell Capacity, Allocated
Bandwidth Throughput, and Per-User Throughput Calculation" on page 743.
8. Performs radio resource management and scheduling to determine the amount of resources to allocate to each
mobile according to the QoS and throughput demands of each mobile using the selected scheduler as explained in
"Scheduling and Radio Resource Allocation" on page 748.
9. Calculates the user throughputs after allocating resources to each mobile as explained in "User Throughput
10. Updates the traffic loads, and noise rise values of all the cells according to the resources in use and the total resources
as follows:
Calculation of Traffic Loads:
Atoll calculates the traffic loads for all the cells TXi(ic).
TL DL =  RDL and TLUL =  RUL
Mi Mi
700
MU – MIMO
TX  ic  M

i i
For uplink MU-MIMO, TL UL = RC UL
MU – MIMO
M
i
Calculation of Uplink Noise Rise:

For each victim cell TXi(ic), the uplink noise rise is calculated and updated by considering each interfering mobile Mj
as explained in "Noise Rise Calculation (UL)" on page 733.
Calculation of Downlink Segmentation Usage:
Atoll calculates the segmentation usages for all the cells as follows:
Mi
 R DL
Mi
M PZ DL = Seg
i
Mi
TX i  ic  PZ = Seg
SU DL = ------------------------------------------------------------
DL -
TX i  ic 
TL DL
M

i
Where R DL is the sum of the percentages of the downlink cell resources allocated to mobiles
M
i
Mi PZ DL = Seg
Mi
PZ DL = Seg
served by the downlink segmented permutation zone.

Calculation of Downlink AAS Usage:
Atoll calculates the downlink AAS usages for all the cells as follows:
Mi
 R DL
AAS
TX i  ic  Mi
AU DL = -------------------------------
AAS -
TX i  ic 
TL DL
Mi
Where  R DL
AAS
is the sum of the percentages of the downlink cell resources allocated to mobiles served by the
M
i
AAS
smart antennas.
Calculation of Uplink MU-MIMO Gain:
Atoll calculates the uplink MU-MIMO gain for all the cells as follows:
MU – MIMO
Mi
 R UL
TX i  ic  MU – MIMO
Mi
G MU – MIMO = -------------------------------------------------------------
-
MU – MIMO
Mi
 RC UL
MU – MIMO
M
i
MU – MIMO
Mi
Where  R UL is the sum of the percentages of the uplink cell resources allocated to MU-MIMO
MU – MIMO
Mi
MU – MIMO
Mi
mobiles and  RC UL is the sum of the real resource consumption of MU-MIMO mobiles.
MU – MIMO
Mi
11. Performs the convergence test to see whether the differences between the current and the new loads are within the
Max  TL DL 
i
TL DL = – TL DL
i
701
Max  TL UL 
i i i
TL UL = – TL UL
i
Max  NR UL 
i i i
NR UL = – NR UL
i
If TL DL , TL UL , and NR UL are the simulation convergence thresholds defined when creating
Req Req Req
the simulation, Atoll stops the simulation in the following cases.
Convergence: Simulation has converged between iteration k - 1 and k if:
TL DL  TL DL AND TL UL  TL UL AND NR UL  NR UL
k Req k Req k Req
No convergence: Simulation has not converged even after the last iteration, i.e., k = Max Number of Iterations defined
when creating the simulation, if:
i i i i i i
TL DL  TL DL OR TL UL  TL UL OR NR UL  NR UL
k Req k Req k Req
12. Repeats the above steps (from step 3.) for the iteration k+1 using the new calculated loads as the current loads.
Simulation Results
At the end of the simulation process, the main results obtained are:
• Downlink traffic loads
• Uplink traffic loads
• Uplink noise rise received at the main antenna
• Ssegmented zone uplink noise rise received at the main antenna
• Angular distributions of downlink traffic power density for cells with smart antennas
• Angular distributions of uplink noise rise for cells with smart antennas
• Downlink AAS usage
• Downlink segmentation usage
• Uplink MU-MIMO capacity gain
These results can be used as input for C/(I+N)-based coverage predictions.
In addition to the above parameters, the simulations also list the connection status of each mobile. Mobiles can be rejected
due to:
• No Coverage: If the mobile does not have any best serving cell (step 3.) or if the mobile is not within the service area
of its best server (step 4.).
• No Service: If the mobile is not able to access a bearer in the direction of its activity (step 6.), i.e., UL, DL, or DL+UL.
• Scheduler Saturation: If the mobile is not in the list of mobiles selected for scheduling (step 8.)
• Resource Saturation: If all the cell resources are used up before allocation to the mobile or if, for a user active in
uplink, the minimum uplink throughput demand is higher than the uplink allocated bandwidth throughput (step 8.)
• Backhaul Saturation: If allocating resources to a mobile makes the effective MAC aggregate site throughputs exceed
the maximum backhaul throughputs defined for the site. This condition is only verified if the simulation was created
with the Backhaul capacity check box selected (step 8.)
Connected mobiles (step 8.) can be:
• Connected UL: If a mobile active in UL is allocated resources in UL.
• Connected DL: If a mobile active in DL is allocated resources in DL.
• Connected DL+UL: If a mobile active in DL+UL is allocated resources in DL+UL.
10.4 Calculation Details

The following sections describe all the calculation algorithms used in point analysis, calculation of coverage predictions,
calculations on subscriber lists, and Monte Carlo simulations.

A WiMAX network can consist of cells that use different channel bandwidths. Therefore, the start and end frequencies of all
the channels may not exactly coincide. Channel bandwidths of cells can overlap each other with different ratios.
702
Figure 10.2: Co-Channel and Adjacent Channel Overlaps
The following sections describe how the co- and adjacent channel overlaps are calculated between the channels used by any
studied cell TXi(ic) and any other cell TXj(jc) of the network. In terms of interference calculation, the studied cell can be
considered a victim of interference received from the other cells that might be interfering the studied cell.
TX i  ic 
If the studied cell is assigned a channel number N Channel , it receives co-channel interference on the channel bandwidth of
N Channel , and adjacent channel interference on the adjacent channel bandwidths, i.e., corresponding to N Channel – 1 and
TX i  ic 
N Channel + 1 .
In order to calculate the co- and adjacent channel overlaps between two channels, it is necessary to calculate the start and
end frequencies of both channels (explained in "Conversion From Channel Numbers to Start and End Frequencies" on
page 703). Once the start and end frequencies are known for the studied and other cells, the co- and adjacent overlaps and
the total overlap ratio are calculated as respectively explained in:
• "Co-Channel Overlap Calculation" on page 704.
• "Adjacent Channel Overlap Calculation" on page 705.
• "Total Overlap Ratio Calculation" on page 706.
10.4.1.1 Conversion From Channel Numbers to Start and End Frequencies

Input
• F Start – FB and F Start – FB : Start frequency of the frequency band assigned to the cells TXi(ic) and TXj(jc).
F Start – FB can be the start frequency of a TDD frequency band ( F Start – FB – TDD ), or the uplink or the downlink start
frequency of an FDD frequency band ( F Start – FB – FDD – UL or F Start – FB – FDD – DL ).
First – TX  ic  First – TX  jc 
i j
• N Channel and N Channel : First channel numbers the frequency band assigned to the cells TXi(ic) and TXj(jc).
• N Channel and N Channel : Channel numbers assigned to cells TXi(ic) and TXj(jc).
For FDD networks, Atoll considers that the same channel number is assigned to a cell in the downlink and uplink, i.e.,
the channel number you assign to a cell is considered for uplink and downlink both.
• ICS FB and ICS FB : Inter-channel spacing of the frequency bands assigned to cells TXi(ic) and TXj(jc).
• CN FB and CN FB : Channel number step of the frequency bands assigned to cells TXi(ic) and TXj(jc).
703
Calculations
Channel numbers are converted into start and end frequencies as follows:
For cell TXi(ic):
TX  ic  TX  ic  TX  ic  TX  ic   TXi  ic  – N First – TX i  ic 
i
F Start =
i
F Start – FB +  W Channel + ICS FB
i i   N Channel Channel
-
-------------------------------------------------------
   TX i  ic  
 CN FB 
For cell TXj(jc):
TX j  jc  TX j  jc  TX  jc  TX  jc   N TXj  jc  – N First – TX j  jc 
= F Start – FB +  W Channel + ICS FB    -------------------------------------------------------
-
j j Channel Channel
F Start    TX  ic 
i 
 CN FB 
Output
i j
• F Start and F Start : Start frequencies for the cells TXi(ic) and TXj(jc).
• F End and F End : End frequencies for the cells TXi(ic) and TXj(jc).
10.4.1.2 Co-Channel Overlap Calculation

Input
TX  ic 
i
Calculations
Atoll first verifies that co-channel overlap exists between the cells TXi(ic) and TXj(jc).
Co-channel overlap exists if:
TX i  ic  TX j  jc  TX i  ic  TX j  jc 
F Start  F End AND F End  F Start
Otherwise there is no co-channel overlap.

Atoll calculates the bandwidth of the co-channel overlap as follows:
= Min  F End  F End  – Max  F Start  F Start 
j i j i
W CCO
   
The co-channel overlap ratio is given by:
TX  ic  – TX  jc 
TX i  ic  – TX j  jc  i j
W CCO
r CCO = ----------------------------------
-
TX i  ic 
W Channel
Output
• r CCO : Co-channel overlap ratio between the cells TXi(ic) and TXj(jc).
704
10.4.1.3 Adjacent Channel Overlap Calculation

Input
i j
• F Start
and F Start : Start frequencies for the cells TXi(ic) and TXj(jc) as calculated in "Conversion From Channel
TX i  ic 
Calculations
Atoll first verifies that adjacent channel overlaps exist between (the lower-frequency and the higher-frequency adjacent
channels of) the cells TXi(ic) and TXj(jc).
Adjacent channel overlap exists on the lower-frequency adjacent channel if:
TX i  ic  TX i  ic  TX j  jc  TX i  ic  TX j  jc 
F Start – W Channel  F End AND F Start  F Start
Adjacent channel overlap exists on the higher-frequency adjacent channel if:

TX i  ic  TX j  jc  TX i  ic  TX i  ic  TX j  jc 
F End  F End AND F End + W Channel  F Start
Otherwise there is no adjacent channel overlap.

Atoll determines the adjacent channel overlap ratio as follows:
Bandwidth of the lower-frequency adjacent channel overlap:
TX i  ic  – TX j  jc  TX  jc  TX  ic  TX  jc  TX  ic  TX  ic 
= Min  F End  F Start  – Max  F Start  F Start – W Channel
j i j i i
W ACO
L
The lower-frequency adjacent channel overlap ratio is given by:

L
r ACO = ----------------------------------
-
L TX i  ic 
W Channel
Bandwidth of the higher-frequency adjacent channel overlap:
TX i  ic  – TX j  jc  TX j  jc  TX i  ic  TX  ic  TX  jc  TX  ic 
= Min  F End  F End + W Channel – Max  F Start  F End 
i j i
W ACO
H    
The higher-frequency adjacent channel overlap ratio is given by:

H
r ACO = ----------------------------------
-
H TX  ic 
i
W Channel
The adjacent channel overlap ratio is given by:

r ACO = r ACO + r ACO
L H
Output
TX  ic  – TX  jc 
i j
• r ACO : Adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc).
10.4.1.4 FDD – TDD Overlap Ratio Calculation

There are many different interference scenarios possible in a WiMAX network depending on the type of duplexing used by
the cells of the network. The most common interference scenarios are FDD-only and TDD-only interferences. However, co-
existing FDD and TDD cells may also exist and interfere each other. Atoll models the co-existence of FDD and TDD cells in a
network by determining the FDD – TDD overlap ratio as follows:
705
Input
TDD
• r DL – Frame : Downlink subframe ratio defined in the global network settings.
Calculations
The FDD – TDD overlap ratio is calculated as follows depending on the frequency bands assigned to the cells TXi(ic) and TXj(jc):
Frequency Band TX i  ic  – TX j  jc 
TXi(ic) TXj(jc) Overlap Ratio r FDD – TDD
TDD TDD 1
TDD FDD 1
TDD
r DL – Frame
FDD TDD -----------------------
100
FDD FDD 1
Output
TX  ic  – TX  jc 
i j
• r FDD – TDD : FDD – TDD overlap ratio between the cells TXi(ic) and TXj(jc).
10.4.1.5 Total Overlap Ratio Calculation

Input
• r CCO : Co-channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Co-Channel Overlap
• r ACO : Adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "Adjacent Channel
Overlap Calculation" on page 705.
• r FDD – TDD : FDD – TDD overlap ratio between the cells TXi(ic) and TXj(jc) as calculated in "FDD – TDD Overlap Ratio
TX i  ic 
• f ACS – FB : Adjacent channel suppression factor defined for the frequency band of the cell TXi(ic).
i j
• W Channel and W Channel : Bandwidths of the channels assigned to the cells TXi(ic) and TXj(jc).
Calculations
The total overlap ratio is:
 TX  ic 
  –f
i 
  TX  ic  – TX  jc  TX  ic  – TX  jc  ACS – FB-
--------------------------- TX  ic  – TX j  jc  TX i  ic  TX j  jc 
 r i j i
+ r ACO
j
 10
10  r i if W Channel  W Channel
  CCO  FDD – TDD
  
TX i  ic  – TX j  jc    
rO = 
 TX  ic 
  –f
i  TX i  ic 
  TX  ic  – TX  jc  TX  ic  – TX  jc  ----------------------------
ACS – FB
TX  ic  – TX j  jc  W TX i  ic  TX j  jc 
 r i j
+ r
i j
 10
10  r i 
Channel
--------------------- if W Channel  W Channel
  CCO ACO  FDD – TDD TX j  jc 
   W Channel
  
TX i  ic 
W Channel
The multiplicative factor ---------------------
TX j  jc 
is used to normalise the transmission power of the interfering cell TXj(jc). This means that
W Channel
if the interfering cell transmits at X dBm over a bandwidth of W Channel , and it interferes over a bandwidth less than W Channel ,
706
TX  ic 
i
W Channel
the interference from this cell should not be considered at X dBm but less than that. The factor ---------------------
TX  jc 
converts X dBm over
j
W Channel
j j
W Channel to Y dBm (which is less than X dBm) over less than W Channel .
Output
• rO : Total co- and adjacent channel overlap ratio between the cells TXi(ic) and TXj(jc).
10.4.2 Preamble Signal Level and Quality Calculations

These calculations include the calculation of the received preamble signal level, and the noise and interference on the
preamble. The following sections also describe how the received preamble signal level, the noise and interference, C/N, and
C/(I+N) ratios are calculated in Atoll:
• "Preamble Noise Calculation" on page 708.
• "Preamble C/N Calculation" on page 712.
• "Preamble Interference Calculation" on page 710.
10.4.2.1 Preamble Signal Level Calculation

Input
TX i  ic 
• P Preamble : Preamble transmission power of the cell TXi(ic).
TX i
Combining
Div
TX i
• G : Transmitter antenna gain for the antenna used by the transmitter TXi.
TX TX
i i
TX
i
is selected.
Mi
Mi
Mi
Mi
Mi
Mi
707
M M M M
i i i i
L , G , L Ant , and L Body are not used in the calculations performed for the point
analysis tool’s profile tab and the preamble signal level based coverage predictions.
Calculations
The received preamble signal level (dBm) from any cell TXi(ic) is calculated for a pixel, subscriber, or mobile Mi as follows:
C Preamble = EIRP Preamble – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body
Where EIRP is the effective isotropic radiated power of the cell calculated as follows:
i i i i
• Without smart antennas: EIRP Preamble = P Preamble + G –L
+ 10  Log  E SA  + G SA
Combining Div
• With smart antennas: EIRP Preamble = P Preamble + G –L + G SA
 
L Path is the path loss (dB) calculated as follows:
TX i
Furthermore, the total losses between the cell and the pixel, subscriber, or mobile Mi can be calculated as follows:
TX TX M M M M
i i i i i i
L Total = L Path + L + L Indoor + M Shadowing – Model – G +L –G + L Ant + L Body
If you wish to exclude the the energy corresponding to the cyclic prefix part of the total
symbol duration from the useful signal level, you must add the following lines in the
Atoll.ini file:
[WiMAX]
ExcludeCPFromUsefulPower = 1
TX i  ic 
When this option is active, the cyclic prefix energy is excluded from C Preamble . In other
words, the factor 10  Log  1 – r CP  is added to C i
i
 Preamble .
Independant of the option, interference levels are calculated for the total symbol
durations, i.e., the energy of the useful symbol duration and the cyclic prefix energy.
Output
TX i  ic 
• C Preamble : Received preamble signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
• L Path : Path loss between the cell TXi(ic) and the pixel, subscriber, or mobile Mi.
• L Total : Total losses between the cell TXi(ic) and the pixel, subscriber, or mobile Mi.
10.4.2.2 Preamble Noise Calculation

For determining the preamble C/N and C/(I+N), Atoll calculates the preamble noise over the bandwidth used by the cell. The
used bandwidth depends on the number of subcarriers used by the preamble.The number of subcarriers used by the
preamble can be different from the number of subcarriers used by the permutation zones.
The preamble noise comprises thermal noise and the noise figure of the equipment. The thermal noise density depends on
the temperature, i.e., it remains constant for a given temperature. However, the value of the thermal noise varies with the
used bandwidth.
Input
TX i  ic 
• N SCa – Preamble : Number of subcarriers used by the preamble defined for the frame configuration of the cell TXi(ic).
TX i  ic 
• N SCa – Total : Total number of subcarriers defined for the frame configuration of the cell TXi(ic).
708
TX  ic 
i
• F Sampling : Sampling frequency for the cell TXi(ic) as calculated in "Calculation of Sampling Frequency" on page 740.
M
i
Calculations

n 0 = 10  Log  K  T  1000  = – 174 dBm/Hz
The thermal noise over the preamble for a cell is calculated as:
TX i  ic 
TX i  ic   TX  ic  N SCa – Preamble Preamble
n 0 – Preamble  i
= n 0 + 10  Log  F Sampling  ---------------------------------
-  f Segment
TX i  ic  
 N SCa – Total 
Effect of Segmentation:
The preamble is segmented and one of the three preamble carrier sets is used for transmission. Each preamble carrier
set uses 1/3rd of the total number of preamble subcarriers. The power transmitted over the preamble has higher
spectral density than the power transmitted over the entire channel bandwidth. This power concentration due to
segmentation on the C/N and C/(I+N) results in an increase in the coverage footprint of the preamble. Hence, the
thermal noise at the pixel, subscriber, or mobile Mi covered by the preamble is reduced by a factor of f Segment = 1
Preamble
--- .
3
The following table shows the different types of subcarriers and their numbers for preamble transmission in WiMAX.
Guard Subcarriers DC Preamble

N SCa – Total Segment N SCa – Preamble f Segment
Left Right Total Subcarrier
All 1 (54) 107 1
0 1 (54) 35 0.3271
128 10 10 20
1 None 36 0.3364
2 None 36 0.3364
All 1 (214) 428 1
0 None 143 0.3341
512 42 41 83
1 1 (214) 142 0.3318
2 None 143 0.3341
All 1 (426) 851 1
0 1 (426) 283 0.3325
1024 86 86 172
1 None 284 0.3337
2 None 284 0.3337
All 1 (852) 1703 1
0 1 (852) 567 0.3329
2048 172 172 344
1 None 568 0.3335
2 None 568 0.3335
The preamble noise is the sum of the thermal noise and the noise figure of the terminal used for the calculations by the pixel,
subscriber, or mobile Mi.
n Preamble = n 0 – Preamble + nf
Output
TX i  ic 
• n Preamble : Preamble noise for the cell TXi(ic).
709
10.4.2.3 Preamble Interference Calculation

the preamble signal levels received from interfering cells TXj(jc) depending on the overlap that exists between the channels
used by the cells TXi(ic) and TXj(jc) and which preamble carrier sets are used by the two cells.
Input
TX j  jc 
• C Preamble : Preamble signal level received from an interfering cell TXj(jc) as calculated in "Preamble Signal Level
Calculation" on page 707 at the pixel, subscriber, or mobile Mi covered by the cell TXi(ic).
In Monte Carlo simulations, interfering signal levels already include M Shadowing – Model , as explained in "Preamble
Signal Level Calculation" on page 679.
information, see "Shadow Fading Model" on page 90). As the received interfering signal levels already include
j j
C Preamble = C Preamble + M Shadowing – C  I
is selected.
• N Seg and N Seg : Segment numbers assigned to the cells TXi(ic) and TXj(jc) calculated from their respective
preamble indexes ( n Preamble and n Preamble ) as follows:
n Preamble N Seg
0 to 31, 96, 99, 102, 105, 108, 111 0

32 to 63, 97, 100, 103, 106, 109, 112 1
64 to 95, 98, 101, 104, 107, 110, 113 2
Inter – Tech
• f IRF : Inter-technology interference reduction factor.
Calculations
The received preamble interference (dBm) from any cell TXj(jc) is calculated for a pixel, subscriber, or mobile Mi as follows:
Inter – Tech
I Preamble = C Preamble + f O + f Seg – Preamble + I DL
TX  ic  – TX  jc 
i j
Where f O is the interference reduction factor due to channel overlap between the cells TXi(ic) and TXj(jc),
fO = 10  Log  r O 

f Seg – Preamble is the interference reduction factor due to preamble segmentation, calculated as follows:
f Seg – Preamble = 10  Log  p Collision 
 
710
TX  ic  – TX  jc  TX  ic  TX  jc 
i j i j
The probability of preamble subcarrier collision p Collision between the cells TXi(ic) and TXj(jc) is 0 if N Seg  N Seg and
i j
1 if N Seg = N Seg .
j j
In case of smart antennas, C Preamble in I Preamble already includes the effect of the
TX
j
number of antenna elements ( E SA ). If you wish to include the effect of the number of
antennas in case of MIMO, you must add the following lines in the Atoll.ini file:
[WiMAX]
MultiAntennaInterference
When the multi-antenna interference option is active, and TXj(jc) does not have a smart
antenna equipment assigned, the interference is incremented by
TX  jc 
+ 10  Log  N Ant – TX .
j
 
TX j  jc 
Where N Ant – TX is the number of MIMO transmission (downlink) antennas defined for
the cell TXj(jc).
Inter – Tech
I DL is the inter-technology downlink interference from transmitters of an external network (linked document of any
technology) calculated as follows:
Mi Mi Mi Mi

Inter – Tech TX – External Inter – Tech
I DL = EIRP DL – L Path – L Indoor + G –L – L Ant – L Body – f IRF
All External TXs
TX – External
Where EIRP DL is the downlink EIRP of the external transmitter, L Path is the path loss from the external transmitters
to the pixel, subscriber, or mobile location, L Indoor is the indoor losses taken into account when the option "Indoor coverage"
Mi Mi
is selected, L is the receiver terminal losses for the pixel, subscriber, or mobile Mi, G is the receiver terminal’s antenna
Mi
gain for the pixel, subscriber, or mobile Mi, L Ant is the receiver terminal’s antenna attenuation calculated for the pixel,
Mi
subscriber, or mobile Mi, and L Body is the body loss defined for the service used by the pixel, subscriber, or mobile Mi.

TX
 k
P DL – Rec 
 -

Inter – Tech
I DL = --------------------------------------
 F  TX  ic  TX 
i k
TX k  ICP DL 
TX k
F  TX i  ic  TX k 
TX k
Output
TX j  jc 
• I Preamble : Preamble interference received from any interfering cell TXj(jc) at the pixel, subscriber, or mobile Mi
covered by a cell TXi(ic).
Inter – Tech
711
10.4.2.4 Preamble C/N Calculation

Input
TX  ic 
i
• C Preamble : Received preamble signal level from the cell TXi(ic) as calculated in "Preamble Signal Level Calculation" on
page 707.
TX i  ic 
• n Preamble : Preamble noise for the cell TXi(ic) as calculated in "Preamble Noise Calculation" on page 708.
Mi
• G Div – Preamble : Preamble diversity gain defined in the WiMAX equipment of the terminal used by the pixel, subscriber,
or mobile Mi.
DL
located.
Calculations
The preamble C/N for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX i  ic  TX i  ic  TX i  ic  Mi
DL
CNR Preamble = C Preamble – n Preamble + G Div – Preamble + G Div
The preamble diversity gain is applied to the preamble C/N when the cell and the terminal both support any form of MIMO in
downlink. The additional downlink diversity gain defined per clutter is also applied.
Output
TX i  ic 
• CNR Preamble : Preamble C/N from the cell TXi(ic) at any pixel, subscriber, or mobile Mi.
10.4.2.5 Preamble C/(I+N) Calculation

The carrier signal to interference and noise ratio is calculated in three steps. First Atoll calculates the received preamble signal
level from the studied cell (as explained in "Preamble Signal Level Calculation" on page 707) at the pixel, subscriber or mobile
under study. Next, Atoll calculates the interference received at the same studied pixel, subscriber, or mobile from all the
interfering cells (as explained in "Preamble Interference Calculation" on page 710). Interference from each cell is weighted
according to the co- and adjacent channel overlap between the studied and the interfering cells, and the probabilities of
subcarrier collision. Finally, Atoll takes the ratio of the preamble signal level, and the sum of the total interference from all
interfering cells and the noise (as calculated in "Preamble Noise Calculation" on page 708).
Input
TX i  ic 
• C Preamble : Preamble signal level received from the cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated in
"Preamble Signal Level Calculation" on page 707.
TX i  ic 
• n Preamble : Preamble noise for the cell TXi(ic) as calculated in "Preamble Noise Calculation" on page 708.
TX j  jc 
• I Preamble : Preamble interference received from any cell TXj(jc) at a pixel, subscriber, or mobile Mi covered by a cell
TXi(ic) as calculated in "Preamble Interference Calculation" on page 710.
Inter – Tech
• NRDL : Inter-technology downlink noise rise.
M
i
• G Div – Preamble : Preamble diversity gain defined in the WiMAX equipment of the terminal used by the pixel, subscriber,
or mobile Mi.
DL
located.
Inter – Tech
• I DL : Downlink inter-technology interference as calculated in "Preamble Interference Calculation" on page 710.
712
Calculations
The preamble C/(I+N) for a cell TXi(ic) is calculated as follows at any pixel, subscriber, or mobile Mi:
TX  ic 
   TX j  jc   n
i  
TX  ic  TX  ic     IPreamble -----------------------------
Preamble  M
= C Preamble –  10  Log   ---------------------------
-  + NR Inter – Tech + G i

i i Inter – Tech 10 DL
CINR Preamble
   10
10

+ I DL + 10
 DL  Div – Preamble + G Div
  All TX j  jc     
     
The preamble diversity gain is applied to the preamble C/(I+N) when the cell and the terminal both support any form of MIMO.
The additional downlink diversity gain defined per clutter is also applied.
The preamble total noise (I+N) for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX  ic 
  TXj  jc   n Preamble
i
TX i  ic    IPreamble -----------------------------
= 10  Log   ---------------------------
-  + NRInter – Tech

Inter – Tech 10
 I + N  Preamble 10 +I + 10
  10  DL  DL
 All TXj  jc    
   
Output
TX i  ic 
• CINR Preamble : Preamble C/(I+N) from the cell TXi(ic) at a pixel, subscriber, or mobile Mi.
TX i  ic 
•  I + N  Preamble : Preamble total noise from the interfering cells TXj(jc) at the pixel, subscriber, or mobile Mi covered
by a cell TXi(ic).
10.4.3 Best Server Determination

In WiMAX, best server refers to a cell ("serving transmitter"-"reference cell" pair) from which a pixel, subscriber, or mobile Mi
gets the highest preamble signal level or preamble C/(I+N). This calculation also determines whether the pixel, subscriber, or
mobile Mi is within the coverage area of any transmitter or not.
Input
TX i  ic 
• C Preamble : Preamble signal level received from any cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated in
Mi Mi Mi
"Preamble Signal Level Calculation" on page 707 using the terminal and service parameters ( L , G , L Ant , and
M
i
L Body ) of Mi. "Preamble C/(I+N) Calculation" on page 712
TX i  ic 
• CINR Preamble : Preamble C/(I+N) received from any cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated in
"Preamble C/(I+N) Calculation" on page 712.
Calculations
The best server of any pixel, subscriber, or mobile Mi, BS M , is the cell from which the received preamble signal level or C/
i
(I+N) is the highest among all the cells. The best server is determined as follows:
BSM = TX i  ic  or BS M = TX i  ic 
i TX i  ic   TX i  ic  
i TX i  ic   TX i  ic  
C Preamble = Best C  CINR Preamble = Best  CINR Preamble 
All TX i  ic   Preamble  All TX i  ic   
Here ic is the cell of the transmitter TXi with the highest preamble power. However, if more than one cell of the same
transmitter covers the pixel, subscriber, or mobile, the final reference cell ic might be different from the initial cell ic (the one
with the highest power) depending on the serving cell selection method:
• Random: In coverage prediction calculations and in calculations on subsriber lists, the cell of the highest priority layer
is selected as the serving (reference) cell. In Monte Carlo simulations, a random cell is selected as the serving
(reference) cell.
• Distributive: In coverage prediction calculations and in calculations on subsriber lists, the cell of the highest priority
layer is selected as the serving (reference) cell. In Monte Carlo simulations, mobiles are distributed among cell layers
one by one, i.e., if more than one cell layer covers a set of mobiles, the first mobile is assigned to the highest priority
layer, the 2nd mobile to the second highest priority layer, and so on.
713
When using either the Random or the Distributive cell selection method, the reference cell once assigned to a mobile
does not change during Monte Carlo simulations.
Output
• BS M : Best serving cell of the pixel, subscriber, or mobile Mi.

i
10.4.4 Service Area Calculation

In WiMAX, a pixel, subscriber, or mobile Mi can be covered by a cell (as calculated in "Best Server Determination" on page 713)
but can be outside the service area. A pixel, subscriber, or mobile Mi is said to be within the service area of its best serving cell
TXi(ic) if the preamble C/N from the cell at the pixel, subscriber, or mobile is greater than or equal to the preamble C/N
threshold defined for the cell.
Input
TX i  ic 
• CNR Preamble : Preamble C/N from the cell TXi(ic) at a pixel, subscriber, or mobile Mi as calculated in "Preamble C/N
TX i  ic 
• T Preamble : Preamble C/N threshold defined for the cell TXi(ic).
Calculations
A pixel, subscriber, or mobile Mi is within the service area of its best serving cell TXi(ic) if:
i i
CNR Preamble  T Preamble
Output
• True: If the calculation criterion is satisfied.

• False: Otherwise.
10.4.5 Permutation Zone Selection

In order to be able to calculate the traffic C/(I+N) and the throughputs, a permutation zone is assigned to each pixel,
subscriber, or mobile Mi located within the service area (as calculated in "Service Area Calculation" on page 714) of its best
serving cell. The permutation zone assigned to Mi is one which covers Mi in terms of distance and preamble C/N or C/(I+N),
and accepts user speeds equal to or higher than Mi’s speed selected for the calculation.
A pixel, subscriber, or mobile Mi which is unable to get a permutation zone is considered to be outside the service area.
Input
TX i  ic 
• d Max – PZ : Maximum distance covered by a permutation zone of a cell TXi(ic).
TX  ic 
i
• QT PZ : Minimum preamble C/N or C/(I+N) required at the pixel, subscriber, or mobile Mi to connect to a
permutation zone of a cell TXi(ic).
TX i  ic 
• Speed Max – PZ : Maximum speed supported by a permutation zone of a cell TXi(ic).
M – TX  ic 
i i
• d : Distance between the pixel, subscriber, or mobile Mi and a cell TXi(ic).
TX  ic 
i
• CNR Preamble : Preamble C/N from the cell TXi(ic) as calculated in "Preamble C/N Calculation" on page 712.
TX i  ic 
• CINR Preamble : Preamble C/(I+N) from the cell TXi(ic) as calculated in "Preamble C/(I+N) Calculation" on page 712.
• Mobility  M i  : Speed of the pixel, subscriber, or mobile Mi.
Calculations
Mi is assigned the permutation zone with the highest priority among the permutation zones whose selection criteria Mi
satisfies. Mi satisfies the selection criteria of a permutation zone if:
• The distance between Mi and TXi(ic) is less than or equal to the maximum distance covered by the permutation zone:
714
M – TX  ic  TX  ic 
i i i
d  d Max – PZ
• The preamble C/N or C/(I+N) at Mi is better than or equal to the quality threshold defined for the permutation zone:
i i i i
CNR Preamble  QT PZ or CINR Preamble  QT PZ
• The mobility of Mi is less than or equal to the maximum mobile speed supported by the permutation zone:
TX i  ic 
Mobility  M i   Speed Max – PZ
Therefore, the permutation zones assigned to a pixel, subscriber, or mobile Mi in the downlink and uplink are:
 
 
 
 
Mi  TX  ic  
= Highest Priority  PZ DL 
i
PZ DL
  TX i  ic  TX i  ic   
  CNR  QT  
  M i – TX i  ic  TX i  ic    Preamble PZ   TX i  ic   
 d  d Max – PZ AND  OR  AND  Mobility  M i   Speed Max – PZ 
 
    TX i  ic  TX i  ic 
 
  CINR
  QT 


Preamble PZ
 
 
 
 
Mi  TX  ic  
= Highest Priority  PZ UL 
i
PZ UL
 TX i  ic  TX i  ic   
  CNR  QT  
  M i – TX i  ic  TX i  ic    Preamble PZ   TX i  ic   
  
Max – PZ 
    
Max – PZ 
d d AND OR AND Mobility M Speed
  i
   TX  ic  TX  ic 
 
  CINR i
  QT
i 


Preamble PZ
If more than 1 permutation zone satisfies the distance, speed, and quality threshold criteria, and all have the same priority,
the permutation zone assigned to the pixel, subscriber, or mobile will be the first in the list of permutation zones (frame
configuration) among these zones.
Output
Mi Mi
• PZ DL and PZ UL : Downlink and uplink permutation zones assigned to the pixel, subscriber, or mobile Mi.
10.4.6 Traffic and Pilot Signal Level and Quality Calculations

Traffic and pilot subcarriers can be transmitted with different transmission powers than the preamble power of a cell, and do
not suffer the same interference and noise as the preamble. The following sections describe how traffic and pilot signal levels,
noise and interference, C/N, and C/(I+N) ratios are calculated on the downlink and uplink.
• "Traffic and Pilot Noise Calculation (DL)" on page 717.
• "Traffic and Pilot Interference Calculation (DL)" on page 718.
• "Traffic and Pilot C/N Calculation (DL)" on page 726.
• "Traffic Signal Level Calculation (UL)" on page 729.
• "Traffic Noise Calculation (UL)" on page 730.
• "Traffic Interference Calculation (UL)" on page 731.
• "Traffic C/N Calculation (UL)" on page 734.
10.4.6.1 Traffic and Pilot Signal Level Calculation (DL)

Input
TX i  ic 
• P Preamble : Preamble transmission power of the cell TXi(ic).
TX i  ic 
• P Traffic : Traffic power reduction of the cell TXi(ic).
715
TX  ic 
i
• P Pilot : Pilot power reduction of the cell TXi(ic).
TX
i
TX TX TX
i i i
• Without smart antennas: G is the transmitter antenna gain, i.e., G = G Ant .
TX i
• With smart antennas: G is the smart antenna gain in the direction of the pixel, subscriber, or mobile Mi, i.e.,
TX i
G = G SA    . Where  is the direction in which Mi is located. For more information on the calculation of
G SA    , refer to section "Beamforming Smart Antenna Models" on page 43.
Array
• G SA : Smart antenna array gain offset defined per clutter class.
Combining
Div
TX i TX i
TX
i
TX i
is selected.
Mi
Mi
Mi
Mi
Mi
Mi
Calculations
The received traffic and pilot signal levels (dBm) from any cell TXi(ic) are calculated for a pixel, subscriber, or mobile Mi as
follows:
C Traffic = EIRP Traffic – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body and
C Pilot = EIRP Pilot – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body
Array Combining Div
EIRP Traffic = P Traffic + G + G SA + G SA + G SA – L and
Array Combining Div
EIRP Pilot = P Pilot + G + G SA + G SA + G SA – L
i i
With P Traffic and P Pilot being the traffic and pilot transmission powers of the cell TXi(ic) calculated as follows:
i i i i i i
P Traffic = P Preamble – P Traffic and P Pilot = P Preamble – P Pilot
716
If you wish to exclude the the energy corresponding to the cyclic prefix part of the total
symbol duration from the useful signal level, you must add the following lines in the
Atoll.ini file:
[WiMAX]
ExcludeCPFromUsefulPower = 1
TX i  ic 
When this option is active, the cyclic prefix energy is excluded from C Preamble . In other
TX i  ic  TX  ic 
words, the factor 10  Log  1 – r CP  is added to C i
 Preamble .
Independant of the option, interference levels are calculated for the total symbol
durations, i.e., the energy of the useful symbol duration and the cyclic prefix energy.
Output
TX i  ic 
• C Traffic : Received traffic signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX i  ic 
• C Pilot : Received pilot signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
10.4.6.2 Traffic and Pilot Noise Calculation (DL)

For determining the traffic and pilot C/N and C/(I+N), Atoll calculates the downlink noise over the channel bandwidth used by
the cell. The used bandwidth depends on the number of used subcarriers. The numbers of subcarriers used by different
permutation zones can be different.
The downlink noise comprises thermal noise and the noise figure of the equipment. The thermal noise density depends on
the temperature, i.e., it remains constant for a given temperature. However, the value of the thermal noise varies with the
used bandwidth.
Input
M
i
PZ DL
• N SCa – Used : Number of subcarriers used by the downlink permutation zone of a cell TXi(ic) assigned to Mi.
TX i  ic 
• N SCa – Total : Total number of subcarriers defined for the frame configuration of a cell TXi(ic).
TX i  ic 
M
i
Calculations

n 0 = 10  Log  K  T  1000  = – 174 dBm/Hz
The thermal noise for a cell is calculated as:

Mi
 PZ DL 
n 0 – DL = n 0 + 10  Log  F Sampling  -------------------------
TX i  ic 

 N 
 SCa – Total 
The downlink noise is the sum of the thermal noise and the noise figure of the terminal used for the calculations by the pixel,
subscriber, or mobile Mi.
n DL = n 0 – DL + nf
Effect of Segmentation:
If you select downlink segmentation support for the frame configuration used by the cell, it means that the first
downlink PUSC permutation zone is segmented. All other zones are pooled together to form a non-segmented zone.
The downlink segmenting factor, f Segment – DL , is calculated from the number of secondary subchannel groups
assigned to the permutation zone in the Permutation Zones table.
717
 PSG + 2  SSG-
f Segment – DL = 3
--------------------------------------------
15
Where, PSG is the number of primary subchannel groups and SSG is the number of used secondary subchannel groups.
The multiplicative coefficients of 3 and 2 are derived from the ratio of the numbers of
subchannels that belong to the primary and to the secondary subchannel gourps. For
example, for the FFT size of 1024 (or 2048), each primary subchannel group contains 6
(or 12) subchannels, and each secondary subchannel group contains 4 (or 8)
subchannels, which gives the ratio of 3:2. And, the denominator of 15 = 3 x 3 + 2 x 3.
f Segment – DL represents the fraction of the channel bandwidth used by a downlink segment. The power transmitted
1
over a segment has ---------------------------- times the spectral density of the power transmitted over the entire channel
f Segment – DL
1
bandwidth. When calculating the downlink C/N and C/(I+N) ratios, the increase in power by ---------------------------- due to this
f Segment – DL
power concentration is equivalent to a reduction in the noise level by f Segment – DL . Hence, if downlink segmentation
is used, the thermal noise power at the pixel, subscriber, or mobile Mi covered by the downlink segmented
permutation zone is reduced by the factor f Segment – DL . Which means that the thermal noise for the a segment of the
channel used by a cell is calculated as:
Mi
 PZ DL 
TX i  ic   TX i  ic  N SCa – Used 
n 0 – DL = n 0 + 10  Log  FSampling  -------------------------
TX i  ic 
 f Segment – DL
 N SCa – Total 
 
Output
TX i  ic 
• n DL : Downlink noise for the cell TXi(ic).
10.4.6.3 Traffic and Pilot Interference Calculation (DL)

the traffic and pilot signal levels received from interfering cells TXj(jc) depending on the overlap that exists between the
channels used by the cells TXi(ic) and TXj(jc), on the traffic loads of the interfering cells TXj(jc), and whether the cells use
downlink segmentation or not. Moreover, the interference can come from cells using simple as well as smart antennas.
The calculation can be divided into the two parts.
• "Traffic and Pilot Interference Signal Levels Calculation (DL)" on page 718.
• "Effective Traffic and Pilot Interference Calculation (DL)" on page 722.
10.4.6.3.1 Traffic and Pilot Interference Signal Levels Calculation (DL)

The traffic and pilot signal levels received from interfering cells TXj(jc) at a pixel, subscriber, or mobile Mi, covered by a cell
TXi(ic), are calculated in a different manner than the traffic and pilot signal levels from the studied cell TXi(ic). This section
explains how these interfering signals are calculated.
Input
TX j  jc 
• P Preamble : Preamble transmission power of the cell TXj(jc).
TX j  jc 
• P Pilot : Pilot power reduction of the interfering cell TXj(jc).
TX j  jc 
• P Traffic : Traffic power reduction of the interfering cell TXj(jc).
TX j  jc 
• P Idle – Pilot : Idle pilot power reduction of the interfering cell TXj(jc).
TX TX
j j
• L : Total transmitter losses for the transmitter TXj ( L = L Total – DL ).
TX j
718
TX
j
• L Ant : Antenna attenuation (from antenna patterns) calculated for the antenna used by the transmitter TXj.
is selected.
Mi
Mi
Mi
M
i
L Ant is determined in the direction of TXj(jc) from the antenna patterns of the antenna used by Mi while the antenna
is pointed towards TXi(ic).
Mi
TX  jc 
j
• TL DL : Downlink traffic load of the interfering cell TXj(jc).
Traffic loads can either be calculated using Monte Carlo simulations, or entered manually for each cell. Calculation of
traffic loads is explained in "Simulation Process" on page 699.
TX  jc 
j
• AU DL : Downlink AAS usage ratio of the interfering cell TXj(jc).
Downlink AAS usage ratios are calculated using Monte Carlo simulations as explained in "Simulation Process" on
page 699.
TX j  jc 
• N SCa – Used : Number of used subcarriers defined for the first downlink permutation zone in the frame configuration
assigned to the interfering cell TXj(jc).
TX j  jc 
• N SCa – Data : Number of data subcarriers defined for the first downlink permutation zone in the frame configuration
assigned to the interfering cell TXj(jc).
Calculations
WiMAX cells can transmit different powers on pilot (NUsed – NData) and data (NData) subcarriers for the part of the frame with
traffic, and a different pilot power for the part of the frame that does not have traffic bursts. Data subcarriers are off during
the empty part of the frame. Therefore, the interference received from a cell depends on the traffic load and the different
powers of the cell, i.e., pilot, traffic, and idle pilot powers.
Monte Carlo simulations and coverage prediction calculations present different scenarios for interference calculations in the
case of smart antennas.
• Monte Carlo Simulations:
In the case of Monte Carlo simulations, the interferer is either using the transmitter antenna or the smart antenna at
any given moment. So, for each interfered pixel, subscriber, or mobile, Atoll already knows the type of the
interference source. Therefore, the interference received from any cell TXj(jc) can be given by:
 I j I
j 
TX j  jc   ----------------------------
Non – AAS
- -------------------
Idle
Without smart antennas: I Total 
= 10  Log 10
10
+ 10
10 
 
 
 
TX  jc 
 I j 
TX  jc   ------------------
AAS -
With smart antennas:
j
I Total 
= 10  Log  10
10 

 
 
• Coverage Predictions:
719
In the case of coverage prediction calculations, the interferer could either be transmitting using the transmitter
antenna, or using the smart antenna, or it could be empty, or not transmitting.Therefore, the interference received
from any cell TXj(jc) can be given by:
 I j I
j
I
j 
TX  jc   ----------------------------
Non – AAS
-
Idle
------------------- -------------------
AAS
= 10  Log  10
j 10 10 10 
I Total + 10 + 10
 
 
 
Where, the three components of the interference are:

TX j  jc 
• I Non – AAS : Interference from the loaded part of the frame transmitted using the main antenna,
TX  jc 
j
• I AAS : Interference from the loaded part of the frame transmitted using the smart antenna,
TX j  jc 
• I Idle : Interference from the empty, or idle, part of the frame.
The above components of the interference are calculated as follows:

The interference from the loaded part of the frame transmitted using the main antenna is calculated as follows:
The received interfering traffic and pilot signal levels (dBm) from any cell TXj(jc) are calculated for a pixel, subscriber,
or mobile Mi as follows:
In Monte Carlo simulations:

TX j  jc  TX j  jc  Mi Mi Mi Mi
I Traffic = EIRP Traffic – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body
I Pilot = EIRP Pilot – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body
In coverage prediction:
I Traffic = EIRP Traffic – L Path – M Shadowing – Model + M Shadowing – C  I – L Indoor + G –L – L Ant – L Body
I Pilot = EIRP Pilot – L Path – M Shadowing – Model + M Shadowing – C  I – L Indoor + G –L – L Ant – L Body
TX j  jc  TX j  jc  TX j TX j TX j  jc  TX j  jc  TX j TX j
EIRP Traffic = P Traffic + G –L and EIRP Pilot = P Pilot + G –L
With P Traffic and P Pilot being the traffic and pilot transmission powers of the cell TXj(jc) calculated as follows:
TX  jc  TX  jc  TX  jc  TX  jc  TX  jc  TX  jc 
j j j j j j
P Traffic = P Preamble – P Traffic and P Pilot = P Preamble – P Pilot
TX j TX j
And G = G Ant , i.e., the transmitter antenna gain for the antenna used by the transmitter TXj.
The interference from the loaded part of the frame transmitted using the main antenna is given as:
  I j TX j  jc  I Pilot
j
TX j  jc 

TX j  jc   TX  jc  TX  jc   -------------------
Traffic
- N - 
------------------
N 
= 10  Log  TL DL   1 – AU DL    10   1 – -------------------------  
j j 10 SCa – Data 10 –
I Non – AAS  ------------------------
- + 10 SCa Data
    TX j  jc   TX j  jc  
  N SCa – Used  N SCa – Used  
  
If you wish to include the effect of the number of antennas in case of MIMO, you must
[WiMAX]
When the multi-antenna interference option is active, the interference is incremented by
+ 10  Log  N Ant – TX . Where N Ant – TX is the number of MIMO transmission
j j
 
(downlink) antennas defined for the cell TXj(jc).
720
The interference from the loaded part of the frame transmitted using the smart antenna is calculated as follows:
The received interfering traffic signal level (dBm) from any cell TXj(jc) is calculated for a pixel, subscriber, or mobile Mi
as follows:
In Monte Carlo simulations:
I AAS = EIRPAAS – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body
In coverage prediction:
I AAS = EIRPAAS – L Path – M Shadowing – Model + M Shadowing – C  I – L Indoor + G –L – L Ant – L Body
EIRP AAS = P Traffic + G –L
TX j  jc 
With P Traffic being the traffic transmission power of the cell TXj(jc) calculated as follows:
P Traffic = P Preamble – P Traffic
TX j
And, G = G SA    is the smart antenna gain in the direction of the victim mobile Mi, calculated from the angular
distributions of the downlink traffic power density of the interfering cells. The angular distribution of the downlink
traffic power density is determined from the array correlation matrices calculated during Monte Carlo simulations.
 is the direction in which the victim pixel, subscriber, or mobile Mi is located. For more information on the
calculation of G SA    , see "Beamforming Smart Antenna Models" on page 43.
The gain of the interfering signal, G SA    , transmitted in the direction of each pixel  is given by:
H
G SA    = g n     S   R Avg  S 
Where S  is the steering vector in the direction  (probe mobile/pixel), H denotes the Hilbert transform, R Avg is the
average array correlation matrix, and g n    is the gain of the nth antenna element in the direction  .
The interference from the empty, or idle, part of the frame transmitted using the transmitter antenna is calculated
as follows:
The received interfering pilot signal level (dBm) from any cell TXj(jc) is calculated for a pixel, subscriber, or mobile Mi
as follows:
I Idle – Pilot = EIRP Idle – Pilot – L Path – L Indoor + G –L – L Ant – L Body
EIRP Idle – Pilot = P Idle – Pilot + G –L
TX j  jc 
With P Idle – Pilot being the idle pilot transmission power of the cell TXj(jc) calculated as follows:
P Idle – Pilot = P Preamble – P Idle – Pilot
TX j TX j
And, G = G Ant , i.e., the transmitter antenna gain for the antenna used by the transmitter TXj.
The interference from the empty, or idle, part of the frame transmitted using the transmitter antenna is given as:
TX  jc 
  I j TX j  jc 

TX j  jc   TX j  jc   -----------------------------
Idle – Pilot 
-  
N
= 10  Log   1 – TL DL    10   1 – ------------------------
SCa – Data  
10
I Idle -
    TX j  jc   
   N SCa – Used  
  
721
If you wish to include the effect of the number of antennas in case of MIMO, you must
[WiMAX]
When the multi-antenna interference option is active, the interference is incremented by
+ 10  Log  N Ant – TX . Where N Ant – TX is the number of MIMO transmission
j j
 
(downlink) antennas defined for the cell TXj(jc).
Output
TX j  jc 
• I Total : Interference received at the pixel, subscriber, or mobile Mi from any interfering cell TXj(jc).
10.4.6.3.2 Effective Traffic and Pilot Interference Calculation (DL)

The effective downlink traffic and pilot interference received at a pixel, subscriber, or mobile Mi covered by a cell TXi(ic) from
interfering cells TXj(jc) depends on the co- and adjacent channel overlap that exists between the channel used by the studied
cell and the interfering cells, and the downlink segmentation parameters of the studied and interfering cells. The first
downlink PUSC zone can be segmented at the studied and the interfering cells. The probability of subcarrier collision depends
on the lengths of the segmented zones and on the subchannel groups used at both sides.
Input
TX j  jc 
• I Total : Interference received at the pixel, subscriber, or mobile Mi from any interfering cell TXj(jc) as calculated in
"Traffic and Pilot Interference Signal Levels Calculation (DL)" on page 718.
• SU DL and SU DL : Downlink segmentation usage ratios defined for cells TXi(ic) and TXj(jc).
Calculations
The total traffic and pilot interference (dBm) from any cell TXj(jc) is calculated for a pixel, subscriber, or mobile Mi as follows:
Inter – Tech
I DL = I Total + f O + f Seg – DL + I DL
Calculations for the interference reduction factors due to channel overlapping and downlink segmentation are explained
below:
as follows:
fO = 10  Log  r O 
 
Interference reduction due to downlink segmentation:

If you select downlink segmentation support for the frame configuration that you are using, it means that the first
zone in the downlink, i.e., the DL PUSC zone, is segmented. All other zones are pooled together to form a group of
non-segmented zones. There are two effects of segmentation:
1. Power concentration, which means that the spectral density of the power transmitted over one segment is higher
than the spectral density of the same power transmitted over the entire channel bandwidth. The effect of power
concentration is visible when calculating the downlink C/(I+N). The power transmitted over a segmented zone has
1
---------------------------
- times the spectral density of the power transmitted over the entire channel bandwidth. When
f Segment – DL
1
calculating the C/(I+N) ratio, the increase in power by ---------------------------- is equivalent to decreasing the noise and
f Segment – DL
722
interference by f Segment – DL . Hence, if downlink segmentation is used, the interference received at the pixel,
subscriber, or mobile Mi covered by the segmented zone is reduced by a factor of f Segment – DL .
2. Collision probability between the subcarriers used by the subchannels belonging to the segment of the studied cell
and the subcarriers used by other sectors, segmented or not. The following paragraphs explain how the collision
probability is calculated.
The downlink segmentation usage (SU) ratio is the percentage of the total downlink traffic load present in the
segmented downlink PUSC zone. For example, if the downlink traffic load is 80 %, and the downlink segmentation
usage ratio is 50 %, then this means that the downlink traffic load of the segmented zone is 40 % (i.e., 50 % of 80 %),
and the downlink traffic load of the non-segmented zones is 40 %.
In coverage predictions, Atoll uses the downlink segmentation usage ratios stored in the cell properties for
determining the interference. In simulations, Atoll resets the downlink segmentation usage ratios for all the cells to
0, and then calculates the downlink segmentation usage ratios according to the traffic loads of the mobiles allocated
to the segmented zone and in the non-segmented zones.
Figure 10.3: Downlink Segmentation
Atoll determines the switching point between the segmented and the non-segmented zones using the downlink
segmentation usage ratio. The switching points between the segmented and non-segmented zones of the victim and
interfering cells, TXi(ic) and TXj(jc) respectively, are calculated as follows:
TX i  ic 
TX i  ic  SU DL
SP = ----------------------------------------------------------------------------------------------- and
SU DL + f Segment – DL  1 – SU DL  
 
TX  jc 
TX j  jc  j
SU DL
SP = -----------------------------------------------------------------------------------------------
SU DL + f Segment – DL   1 – SU DL 
 
Where, SP is the switching point between the segmented and the non-segmented zones, SU is the downlink
segmentation usage ratios of the cells, and f Segment – DL is downlink segmenting factor, which gives the bandwidth
used by a segment.
The downlink segmenting factor, f Segment – DL , is calculated from the number of secondary subchannel groups
assigned to the first downlink PUSC permutation zone in the Permutation Zones table.
 PSG + 2  SSG-
f Segment – DL = 3
--------------------------------------------
15
Where, PSG is the number of primary subchannel groups and SSG is the number of secondary subchannel groups.
The multiplicative coefficients of 3 and 2 are derived from the ratio of the numbers of
subchannels that belong to the primary and to the secondary subchannel gourps. For
example, for the FFT size of 1024 (or 2048), each primary subchannel group contains 6
(or 12) subchannels, and each secondary subchannel group contains 4 (or 8)
subchannels, which gives the ratio of 3:2. And, the denominator of 15 = 3 x 3 + 2 x 3.
723
If the downlink segmentation usage ratio is set to 0, it means that the segmented zone does not exist. Setting SU to 0
gives SP = 0, and setting SU to 1 gives SP = 1 (or 100%), which shows how the switching point varies with the downlink
segmentation usage ratio.
Derivation of the switching point formula: The downlink segmentation usage ratio is used
to partition the total downlink traffic load into segmented and non-segmented zones.
Therefore, the switching point formula is derived from the equation:
SU DL  TL DL  1 – SU DL   TL DL
--------------------------------------------------------------------
- = -----------------------------------------------
-
SP  fSegment – DL  W Channel  1 – SP   W Channel
With cells using downlink segmentation, there can be four different interference scenarios.
• Between the segmented zone of the victim and the segmented zone of the interferer.
• Between the segmented zone of the victim and the non-segmented zone of the interferer.
• Between the non-segmented zone of the victim and the segmented zone of the interferer.
• Between the non-segmented zone of the victim and the non-segmented zone of the interferer.
Figure 10.4: Downlink Segmentation Interference Scenarios
Therefore, Atoll calculates the probabilities of collision for each scenario and weights the total interference according
to the total collision probability. The probability of collision p Coll for each scenario is given by the following formula:
3  PSG Com + 2  SSG Com

p Coll = ----------------------------------------------------------------------
-
3  PSG + 2  SSG
Where, PSGCom is the number of primary subchannel groups common in TXi(ic) and TXj(jc), SSGCom is the number of
TX i  ic 
secondary subchannel groups common in TXi(ic) and TXj(jc), PSG is the number of primary subchannel groups
TX i  ic 
in the cell TXi(ic), and SSG is the number of secondary subchannel groups in the cell TXi(ic).
The segment numbers and the cell permutation base numbers (Cell PermBase) are determined from the cell’s
preamble index. The mapping between the preamble index, the segment number, and Cell PermBase is available in
the IEEE specifications. This mapping is performed in Atoll as follows:
Preamble Index ( PI )
PI  96 96  PI  114
Range: 0 to 113
Cell PermBase ( PB )
PI Modulo 32 PI – 96
Range: 0 to 31
Segment Number ( N Seg )
Floor  ------
PI
 PI – 96  Modulo 3
Range: 0, 1, 2 32
There can be 2 cases for calculating the total probability of collision.
724
• Case 1: If the pixel, subscriber, or mobile Mi is covered by the segmented zone of TXi(ic), the total collision
probability for the pixel, subscriber, or mobile Mi is calculated as follows:
 TX  jc  TX  ic 
 SS
p Coll If SP
j
 SP
i

TX  ic  – TX  jc 
i j  TX  jc  TX  ic  TX  jc 
p Collision – DL =  SS
+ p Coll   SP 
j SN i j
 p Coll  SP 
– SP
 TX j  jc  TX i  ic 
 -------------------------------------------------------------------------------------------------------------
TX i  ic 
If SP  SP
 SP

• Case 2: If the pixel, subscriber, or mobile Mi is covered by the non-segmented zone of TXi(ic), the total collision
probability for the pixel, subscriber, or mobile Mi is calculated as follows:
 NN
 p Coll If SP  SP

TX i  ic  – TX j  jc   TX j  jc  TX  jc  TX  ic 
=  p NN   + p NS   SP j – SP i 
Coll   1 – SP
p Collision – DL
 ---------------------------------------------------------------------------------------------------------------------------- If SP j  SP i
 Coll  TX  jc  TX  ic 
 TX  ic 
  1 – SP i 
  
The interference reduction factor due to downlink segmentation for the pixel, subscriber, or mobile Mi is calculated
as follows:
TX i  ic  – TX j  jc  TX  ic  – TX  jc 
= 10  Log  p Collision – DL 
i j
f Seg – DL
 
Inter – Tech
I DL is the inter-technology downlink interference from transmitters of an external network (linked document of any
technology) calculated as follows:
M M M M

Inter – Tech TX – External i i i i Inter – Tech
I DL = EIRP DL – L Path – L Indoor + G –L – L Ant – L Body – f IRF
All External TXs
TX – External
Where EIRP DL is the downlink EIRP of the external transmitter, L Path is the path loss from the external transmitters
to the pixel, subscriber, or mobile location, L Indoor is the indoor losses taken into account when the option "Indoor coverage"
Mi Mi
is selected, L is the receiver terminal losses for the pixel, subscriber, or mobile Mi, G is the receiver terminal’s antenna
Mi
gain for the pixel, subscriber, or mobile Mi, L Ant is the receiver terminal’s antenna attenuation calculated for the pixel,
Mi
subscriber, or mobile Mi, and L Body is the body loss defined for the service used by the pixel, subscriber, or mobile Mi.

TX k
 ---------------------------------------

Inter – Tech
I DL =
 F  TX i  ic  TX k 
TX  ICP DL 
k
TX
k
F  TX i  ic  TX k 
TX k
Output
TX j  jc 
• I DL : Effective downlink traffic and pilot interference received at the pixel, subscriber, or mobile Mi from any
interfering cell TXj(jc).
Inter – Tech
725
10.4.6.4 Traffic and Pilot C/N Calculation (DL)

Input
TX  ic 
i
• C Traffic : Received traffic signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as calculated in "Traffic
and Pilot Signal Level Calculation (DL)" on page 715.
TX i  ic 
• C Pilot : Received pilot signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as calculated in "Traffic
TX i  ic 
• n DL : Downlink noise for the cell TXi(ic) as calculated in "Traffic and Pilot Noise Calculation (DL)" on page 717.
TX i  ic 
• T AMS : AMS threshold defined for the cell TXi(ic).
M
i
• T B : Bearer selection thresholds of the bearers defined in the WiMAX equipment used by Mi’s terminal.
Mi
or mobile Mi.
M
i
or mobile Mi.
TX i  ic 
• N Ant – TX : Number of MIMO transmission (downlink) antennas defined for the cell TXi(ic).
Mi
• N Ant – RX : Number of MIMO reception (downlink) antennas defined for the terminal used by the pixel, subscriber, or
mobile Mi.
Mi
• Subchannel allocation mode used by the downlink permutation zone PZ DL assigned to the pixel, subscriber, or mobile
Mi as calculated in "Permutation Zone Selection" on page 714.
M
BLER  BDL : Downlink block error rate read from the graphs available in the WiMAX equipment assigned to the
i
•
 
Calculations
The traffic and pilot C/N for a cell TXi(ic) are calculated as follows for any pixel, subscriber, or mobile Mi:
i i i
CNR Traffic = C Traffic – n DL
CNR Pilot = C Pilot – n DL
The bearers available for selection in the pixel, subscriber, or mobile Mi’s WiMAX equipment are the ones:
• Whose selection thresholds are less than the traffic or pilot C/N at Mi: T B  CNR Traffic or T B  CNR Pilot
Mi
If the cell supports MIMO, the STTD/MRC or SU-MIMO diversity gain, G Div – DL , corresponding to the bearer is
applied to its selection threshold. The gain is read from the properties of the WiMAX equipment assigned to the
TX i  ic  Mi Mi
pixel, subscriber, or mobile Mi for N Ant – TX , N Ant – RX , the subchannel allocation mode of PZ DL , Mobility  M i  ,
M
BLER  B DL .
i
726
DL
The additional diversity gain defined for the clutter class of the pixel, subscriber, or mobile Mi G Div is also
applied. Therefore, the bearers available for selection are all the bearers defined in the WiMAX equipment for
M M TX  ic 
i i DL i
T B – G Div – DL – G Div  CNR Traffic
DL
T B – G Div – DL – G Div  CNR Pilot
• Bearer Index
• Peak MAC Throughput
From among the bearers available for selection, the selected bearer is the one with the highest downlink peak
MAC channel throughput as calculated in "Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput,
and Per-User Throughput Calculation" on page 743.
• Effective MAC Throughput
MIMO – STTD/MRC and SU-MIMO Diversity Gain:
Once the bearer is known, the traffic and pilot C/N calculated above become:
DL
CNR Traffic = CNR Traffic + G Div – DL + G Div
DL
CNR Pilot = CNR Pilot + G Div – DL + G Div
Mi
Where G Div – DL is the STTD/MRC or SU-MIMO diversity gain corresponding to the selected bearer.
Output
TX  ic 
i
• CNR Traffic : Traffic C/N from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX i  ic 
• CNR Pilot : Pilot C/N from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
10.4.6.5 Traffic and Pilot C/(I+N) and Bearer Calculation (DL)

The carrier signal to interference and noise ratio is calculated in three steps. First Atoll calculates the received signal level from
the studied cell (as explained in "Traffic and Pilot Signal Level Calculation (DL)" on page 715) at the pixel, subscriber, or mobile
under study. Next, Atoll calculates the interference received at the same studied pixel, subscriber, or mobile from all the
interfering cells (as explained in "Traffic and Pilot Interference Calculation (DL)" on page 718). Interference from each cell is
weighted according to the co- and adjacent channel overlap between the studied and the interfering cells, the traffic loads of
the interfering cells, and the probabilities of subcarrier collision if downlink segmentation is used. Finally, Atoll takes the ratio
of the signal level and the sum of the total interference from other cells and the downlink noise (as calculated in "Traffic and
Pilot Noise Calculation (DL)" on page 717).
Input
TX  ic 
i
• C Traffic : Received traffic signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as calculated in "Traffic
TX  ic 
i
• C Pilot : Received pilot signal level from the cell TXi(ic) at the pixel, subscriber, or mobile Mi as calculated in "Traffic
TX  ic 
i
• n DL : Downlink noise for the cell TXi(ic) as calculated in "Traffic and Pilot Noise Calculation (DL)" on page 717.
727
TX  jc 
j
• I DL : Effective downlink traffic and pilot interference from any cell TXj(jc) calculated for a pixel, subscriber, or
mobile Mi covered by a cell TXi(ic) as explained in "Traffic and Pilot Interference Calculation (DL)" on page 718.
Inter – Tech
• NRDL : Inter-technology downlink noise rise.
TX  ic 
i
Mi
• T B : Bearer selection thresholds of the bearers defined in the WiMAX equipment used by Mi’s terminal.
M
i
or mobile Mi.
Mi
or mobile Mi.
TX i  ic 
Mi
• N Ant – RX : Number of MIMO reception (downlink) antennas defined for the terminal used by the pixel, subscriber, or
mobile Mi.
Mi
• Subchannel allocation mode used by the downlink permutation zone PZ DL assigned to the pixel, subscriber, or mobile
M
BLER  BDL : Downlink block error rate read from the graphs available in the WiMAX equipment assigned to the
i
•
Inter – Tech
• I DL : Downlink inter-technology interference as calculated in "Traffic and Pilot Interference Calculation (DL)"
on page 718.
Calculations
The traffic and pilot C/(I+N) for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
   TXj  jc  TX i  ic  
TX i  ic  
TX i  ic    IDL  n
DL
 
   ------------------- --------------------- Inter – Tech

Inter – Tech
CINR Traffic = C Traffic – 10  Log 10 + I DL + 10 + NR DL and
   10  10  
  All TXj  jc     
     
   TXj  jc  TX i  ic  
TX i  ic  TX i  ic     IDL  n DL  
   ------------------- --------------------- Inter – Tech

Inter – Tech
CINR Pilot = C Pilot – 10  Log 10 + I DL + 10 + NR DL
   10  10  
  All TXj  jc     
     
The Traffic Total Noise (I+N) for a cell TXi(ic) is calculated as follows for any pixel, subscriber, or mobile Mi:
TX  ic 
  TX j  jc  n DL
i 
TX i  ic    I DL  ---------------------
= 10  Log   ------------------
10  + I DL  + NR Inter – Tech
- Inter – Tech 10
 I + N  DL + 10
  10   DL
 All TX j  jc    
   
The bearers available for selection in the pixel, subscriber, or mobile Mi’s WiMAX equipment are the ones:
• Whose selection thresholds are less than the traffic or pilot C/(I+N) at Mi: T B  CINR Traffic or T B  CINR Pilot
728
M
i
If the cell supports MIMO, the STTD/MRC or SU-MIMO diversity gain, G Div – DL , corresponding to the bearer is
applied to its selection threshold. The gain is read from the properties of the WiMAX equipment assigned to the
TX  ic  M M
i i i
pixel, subscriber, or mobile Mi for N Ant – TX , N Ant – RX , the subchannel allocation mode of PZ DL , Mobility  M i  ,
M
BLER  B DL .
i
 
DL
DL
T B – G Div – DL – G Div  CINR Traffic
DL
T B – G Div – DL – G Div  CINR Pilot
• Bearer Index
From among the bearers available for selection, the selected bearer is the one with the highest downlink peak
MIMO – STTD/MRC and SU-MIMO Diversity Gain:
Once the bearer is known, the traffic and pilot C/(I+N) calculated above become:
DL
CINR Traffic = CINR Traffic + G Div – DL + G Div
DL
CINR Pilot = CINR Pilot + G Div – DL + G Div
M
i
Where G Div – DL is the STTD/MRC or SU-MIMO diversity gain corresponding to the selected bearer.
Output
TX i  ic 
• CINR Traffic : Traffic C/(I+N) from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX i  ic 
• CINR Pilot : Pilot C/(I+N) from the cell TXi(ic) at the pixel, subscriber, or mobile Mi.
TX i  ic 
•  I + N  DL : Traffic Total noise from the interfering cells TXj(jc) at the pixel, subscriber, or mobile Mi covered by a cell
TXi(ic).
Mi
• B DL : Bearer assigned to the pixel, subscriber, or mobile Mi in the downlink.
10.4.6.6 Traffic Signal Level Calculation (UL)

Input
M
i
• P Max : Maximum transmission power of the terminal used by the pixel, subscriber, or mobile Mi without power
control.
Mi
• P Eff : Effective transmission power of the terminal used by the pixel, subscriber, or mobile Mi after power control as
calculated in "Traffic C/(I+N) and Bearer Calculation (UL)" on page 737.
729
TX
i
TX
i
TX TX TX
i i i
• Without smart antennas: G is the transmitter antenna gain, i.e., G = G Ant .
TX i TX i TX i
• With smart antennas: G is the uplink smart antenna beamforming gain, i.e., G = G SA = 10  Log  E SA  .
 
For more information on the calculation of G SA , refer to section "Beamforming Smart Antenna Models" on
page 43.
TX i TX i
• L : Total transmitter losses for the transmitter TXi ( L = L Total – UL ).
TX i
TX i
is selected.
Mi
Mi
Mi
M
i
Mi
Mi
Calculations
The received traffic signal level (dBm) from a pixel, subscriber, or mobile Mi at its serving cell TXi(ic) is calculated as follows:
Mi Mi TX i TX i Mi Mi
C UL = EIRP UL – L Path – M Shadowing – Model – L Indoor + G –L – L Ant – L Body
Where EIRP is the effective isotropic radiated power of the terminal calculated as follows:
Mi Mi Mi Mi
EIRP UL = P +G –L
Mi Mi Mi Mi
With P = P Max without power control at the start of the calculations, and is the P = P Eff after power control.
Output
Mi
• C UL : Received uplink signal level from the pixel, subscriber, or mobile Mi at a cell TXi(ic).
10.4.6.7 Traffic Noise Calculation (UL)

For determining the uplink C/N and C/(I+N), Atoll calculates the uplink noise over the channel bandwidth used by the cell. The
used bandwidth depends on the number of used subcarriers. The numbers of subcarriers used by different permutation zones
can be different.
The uplink noise comprises thermal noise and the noise figure of the equipment. The thermal noise density depends on the
temperature, i.e., it remains constant for a given temperature. However, the value of the thermal noise varies with the used
bandwidth.
730
Input
M
i
PZ
UL
• N SCa – Used : Number of subcarriers used by the uplink permutation zone of a cell TXi(ic) assigned to Mi.
TX i  ic 
TX i  ic 
TX  ic 
i
• nf : Noise figure of the cell TXi(ic).
Calculations

n 0 = 10  Log  K  T  1000  = – 174 dBm/Hz
The thermal noise for a cell is calculated as:

Mi
 PZ UL 
n 0 – UL = n 0 + 10  Log  F Sampling  -------------------------
TX i  ic 

 N 
 SCa – Total 
The uplink noise is the sum of the thermal noise and the noise figure of the cell TXi(ic).
n UL = n 0 – UL + nf
Output
TX i  ic 
• n UL : Uplink noise for the cell TXi(ic).
10.4.6.8 Traffic Interference Calculation (UL)

The uplink traffic interference is only calculated during Monte Carlo simulations. In coverage predictions, the uplink noise rise
values already available in simulation results or in the Cells table are used.
The interference received by a cell TXi(ic) from an interfering mobile covered by a cell TXj(jc) can be defined as the uplink signal
level received from interfering mobiles Mj depending on the overlap that exists between the channels used by the cells TXi(ic)
and TXj(jc), on the traffic loads of the interfering mobile Mj.
The calculation of uplink interference can be divided into two parts:
• Calculation of the uplink interference from each individual interfering mobile as explained in "Traffic Interference
Signal Levels Calculation (UL)" on page 731.
• Calculation of the uplink noise rise which represents the total uplink interference from all the interfering mobiles as
explained in "Noise Rise Calculation (UL)" on page 733.
10.4.6.8.1 Traffic Interference Signal Levels Calculation (UL)

Input
Mj
• C UL : Uplink signal level received at a cell TXi(ic) from an interfering mobile Mj covered by a cell TXj(jc) as calculated in
"Traffic Signal Level Calculation (UL)" on page 729.
In Monte Carlo simulations, interfering signal levels already include M Shadowing – Model , as explained in "Traffic Signal
Level Calculation (UL)" on page 729.
information, see "Shadow Fading Model" on page 90). As the interfering signal levels already include
731
M M
j j
C UL = C UL + M Shadowing – C  I
is selected.
Mj
• TL UL : Uplink traffic load of the interfering mobile Mj.
Traffic loads are calculated during Monte Carlo simulations as explained in "Scheduling and Radio Resource
Calculations
The uplink interference received at a cell TXi(ic) from an interfering mobile Mj covered by a cell TXj(jc) is calculated as follows:
Mj Mj TX i  ic  – TX j  jc  Mj TX i  ic  – TX j  jc 
I UL = C UL + f O + f TL – UL + f Seg – UL
Calculations for the interference reduction factors due to channel overlapping, uplink traffic load, and uplink segmentation
are explained below:
as follows:
fO = 10  Log  r O 

Interference reduction due to interfering mobile’s traffic load:

The interference reduction factor due to the interfering mobile’s uplink traffic load is calculated as follows:
M M
f TL – UL = 10  Log  TL UL
j j
 
Interference reduction due to uplink segmentation:

If you select uplink segmentation support for the frame configuration that you are using, it means that the first zone
in the uplink, i.e., the UL PUSC zone, is segmented. All other zones are pooled together to form a group of non-
segmented zones. The interference reduction factor due to uplink segmentation is calculated as follows:
= 10  Log  p Collision – UL 
i j i j
f Seg – UL
Where p Collision – UL is the collision probability between the subcarriers of the uplink segments being used by the
interfered and interfering cells. It is determined during Monte Carlo simulations as follows:
TX i  ic  – TX j  jc  SC Com
p Collision – UL = ------------------
-
TX i  ic 
SC
TX i  ic 
Where, SCCom is the number of subchannels common in TXi(ic) and TXj(jc), SC is the number of subchannels in
the cell TXi(ic).
The segment numbers and the cell permutation base numbers (Cell PermBase) are determined from the cell’s
preamble index. The mapping between the preamble index, the segment number, and Cell PermBase is available in
the IEEE specifications. This mapping is performed in Atoll as follows:
Preamble Index ( PI )
PI  96 96  PI  114
Range: 0 to 113
Cell PermBase ( PB )
PI Modulo 32 PI – 96
Range: 0 to 31
732
Segment Number ( N Seg )

Floor  ------
PI
 32  PI – 96  Modulo 3
Range: 0, 1, 2
In Monte Carlo simulations, Atoll calculates two separate noise rise values; for the mobiles served by the segmented
zone of the interfered cell Atoll calculates the uplink segmented noise rise, and for the mobiles served by the non-
segmented zones of the interfered cell Atoll calculates the uplink noise rise.
In coverage predictions, point analysis, and calculations on subscriber lists, according to the zone, segmented or non-
segmented, that covers the pixel, receiver, or subscriber, Atoll uses either the uplink segmented noise rise or the
uplink noise rise to calculate the C/(I+N). For more information on the calculation of the uplink noise rise, see "Noise
Rise Calculation (UL)" on page 733.
Output
Mj
• I UL : Uplink interference signal level received at a cell TXi(ic) from an interfering mobile Mj covered by a cell TXj(jc).
10.4.6.8.2 Noise Rise Calculation (UL)

The uplink noise rise is defined as the ratio of the total uplink interference received by any cell TXi(ic) from interfering mobiles
Mj present in the coverage areas of other cells TXj(jc) to the uplink noise of the cell TXi(ic). In other words, it is the ratio (I+N)/N.
Input
Mj
• I UL : Uplink interference signal levels received at a cell TXi(ic) from interfering mobiles Mj covered by other cells TXj(jc)
as calculated in "Traffic Interference Signal Levels Calculation (UL)" on page 731.
TX i  ic 
• n UL : Uplink noise for the cell TXi(ic) as calculated in "Traffic Noise Calculation (UL)" on page 730.
Inter – Tech
• NR UL : Inter-technology uplink noise rise.
Calculations
The uplink noise rise and total noise (I+N) for the cell TXi(ic) are calculated as follows:
• Without smart antennas:
For any mobile Mi covered by a non-segmented zone in the interfered cell TXi(ic), Atoll calculates the UL noise rise as
follows:
  IMj  TX i  ic 
  UL  n 
TX i  ic     non-seg M i

UL -
-------------------- TX i  ic 
= 10  Log 
10 

-------------------------------------------- Inter – Tech
NR UL
  10 10  + 10 
+ NRUL – n UL
 All Mj  
 
   
 All TX
j
 jc  
For any pixel, subscriber, or mobile Mi covered by the non-segmented zone in the interfered cell TXi(ic), Atoll
calculates the uplink total noise (I+N) as follows:
 I + N  UL = NR UL + n UL
For any mobile Mi covered by the segmented zone in the interfered cell TXi(ic), Atoll calculates the segmented zone
UL noise rise as follows:
  IMj  TX i  ic 
  UL  n UL 
TX i  ic     seg M i
 --------------------- TX i  ic 
= 10  Log 
10 

--------------------------------- Inter – Tech
NR UL – Seg  10 10  + 10  + NR UL – n UL
 All M j  
 
   
 All TX
j
 jc  
For any pixel, subscriber, or mobile Mi covered by the segmented zone in the interfered cell TXi(ic), Atoll calculates
the uplink total noise (I+N) as follows:
 I + N  UL = NR UL – Seg + n UL
733
• With smart antennas:

The angular distribution of the uplink noise rise is calculated during Monte Carlo simulations and can be stored in the
Cells table in order to be used in coverage predictions. The angular distribution of the uplink noise rise is given by:
2
I UL    +  n  I
NRUL    = ---------------------------------
2
n  I
TX i  ic 
2
 I + N  UL    = I UL    +  n  I
Output
TX i  ic 
• NRUL : Non-segmented uplink noise rise for the cell TXi(ic).
TX i  ic 
• NRUL – Seg : Segmented uplink noise rise for the cell TXi(ic).
TX i  ic 
• NRUL    : Angular distribution of the uplink noise rise for the cell TXi(ic).
•  I + N  UL or  I + N  UL    : Total Noise for a cell TXi(ic) calculated for any pixel, subscriber, or mobile Mi.
10.4.6.9 Traffic C/N Calculation (UL)

Input
M
i
• C UL : Received uplink signal level from the pixel, subscriber, or mobile Mi at its serving cell TXi(ic) as calculated in
"Traffic Signal Level Calculation (UL)" on page 729.
TX i  ic 
• n UL : Uplink noise for the cell TXi(ic) as calculated in "Traffic Noise Calculation (UL)" on page 730.
TX i  ic 
TX  ic 
i
• T B – Lowest : Bearer selection threshold of the lowest bearer in the WiMAX equipment assigned to the cell TXi(ic).
Mi
PZ UL
• N SC : Number of subchannels per channel defined for the uplink permutation zone assigned to the pixel, subscriber,
or mobile Mi as calculated in "Permutation Zone Selection" on page 714.
Mi
PZ UL = 8
• N SC  Seg : Number of subchannels per segment for the first uplink PUSC permutation zone.
Mi
Mi
• M PC : Power control margin defined in the global network settings.
M
i
• T B : Bearer selection thresholds of the bearers defined in the WiMAX equipment used bythe cell TXi(ic).
Mi
or mobile Mi.
M
i
or mobile Mi.
Mi
• N Ant – TX : Number of MIMO transmission (uplink) antennas defined for the terminal used by the pixel, subscriber, or
mobile Mi.
TX i  ic 
• N Ant – RX : Number of MIMO reception (uplink) antennas defined for the cell TXi(ic).
M
i
• Subchannel allocation mode used by the uplink permutation zone PZ UL assigned to the pixel, subscriber, or mobile
734
M
BLER  B UL : Uplink block error rate read from the graphs available in the WiMAX equipment assigned to the cell
i
•
TXi(ic).
Calculations
The uplink C/N from a pixel, subscriber, or mobile Mi at its serving cell TXi(ic) is calculated as follows:
M M TX  ic 
i i i
CNR UL = C UL – n UL
The bearers available for selection in the cell TXi(ic)’s WiMAX equipment are the ones:
Mi Mi
• Whose selection thresholds are less than the uplink C/N at Mi: T B  CNR UL
TX i  ic 
If the cell supports MIMO, the STTD/MRC, SU-MIMO diversity or MU-MIMO diversity gain, G Div – UL ,
WiMAX equipment assigned to the cell TXi(ic) for N Ant – RX , N Ant – TX , the subchannel allocation mode of PZ UL ,
M
Mobility  M i  , BLER  B UL .
i
UL
UL
T B – G Div – UL – G Div  CNR UL
• Bearer Index
From among the bearers available for selection, the selected bearer is the one with the highest uplink peak MAC
Per-User Throughput Calculation" on page 743.
MIMO – STTD/MRC, SU-MIMO Diversity, and MU-MIMO Diversity Gain:
Once the bearer is known, the uplink C/N calculated above becomes:
UL
CNR UL = CNR UL + G Div – UL + G Div
TX i  ic 
Where G Div – UL is the STTD/MRC, SU-MIMO diversity, or MU-MIMO diversity gain corresponding to the selected
bearer.
Uplink Subchannelisation:
The uplink subchannelisation depends on the uplink bandwidth allocation target defined for the scheduler used by the
cell TXi(ic). The uplink C/N calculated above is given for the total number of subchannels associated with the
735
M
i
PZ
UL
permutation zone, i.e., N SC . Subchannelisation is performed for all the pixels, subscribers, or mobiles in the uplink,
and may reduce the number of used subchannels in order to satisfy the selected target.
• Full Bandwidth
transmission, there is no gain in the uplink C/N.
The bandwidth used for transmission by a mobile is reduced only if the uplink C/N is not enough to even access
the lowest bearer. For example, as a mobile moves from good to bad radio conditions, the number of subchannels
used by it for transmission in uplink are reduced one by one in order to improve the uplink C/N. The calculation of
the gain introduced by the subchannelisation is explained below.
• Best Bearer
The bandwidth used for transmission by a mobile is reduced in order to improve the uplink C/N enough to access
the best bearer. For example, if using 5 subchannels, a mobile is able to access the best bearer, and using 6 it would
only get access to the second best, it will be assigned 5 subchannels as the used uplink bandwidth. Although using
4 subchannels, its uplink C/N will be better than when using 5, the uplink bandwidth is not reduced to 4 because
it does not provide any gain in terms of the bearer, i.e., the mobile already has the best bearer using 5
subchannels. The calculation of the gain introduced by the bandwidth reduction is explained below.
i.e., bearer with the highest index, with the highest peak MAC throughput, or with the highest effective MAC
throughput.
The uplink subchannelisation may result in the use of a number of subchannels which is less than the total number of
subchannels associated with the permutation zone. The gain related to this bandwidth reduction is applied to the
uplink C/N:
Mi
 PZUL 
Mi Mi  N SC 
CNR UL = CNR UL+ 10  Log  -----------------
Final All SC  N Mi 
 SC – UL
Mi
M PZ
Min i UL
Where N SC – UL  Service   N SC – UL  N SC for any pixel, subscriber, or mobile Mi covered by a non-segmented
Mi
Mi PZ UL = 8
Min
permutation zone in the interfered cell TXi(ic), and N SC – UL  Service   N SC – UL  N SC  Seg for any pixel, subscriber,
or mobile Mi covered by the segmented uplink PUSC zone in the interfered cell TXi(ic).
Uplink Power Control:

Once the subchannelisation is performed, Atoll continues to work with the C/N given by the subchannelisation, i.e.,
M M
i i
CNR UL = CNR UL .
Final
The pixel, subscriber, or mobile Mi reduces its transmission power so that the uplink C/N from it at its cell is just
enough to get the selected bearer.
If with P = P Max AND CNR UL  T Mi + M PC , where T Mi is the bearer selection threshold, from the WiMAX
B UL B UL
equipment assigned to the cell TXi(ic), for the bearer selected for the pixel, subscriber, or mobile Mi.
Mi  Mi  M i  TX i  ic    Mi 
P Eff = Max  PMax –  CNR UL –  T M + M PC   P Min
   B i  
UL
Mi Mi
CNR UL is calculated again using P Eff .
736
Output
M
i
• CNR UL : Uplink C/N from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic).
10.4.6.10 Traffic C/(I+N) and Bearer Calculation (UL)

The carrier signal to interference and noise ratio is calculated in three steps. First, Atoll calculates the received signal level
from each pixel, subscriber, or mobile at its serving cell using the effective power of the terminal used by the pixel, subscriber,
or mobile as explained in "Traffic Signal Level Calculation (UL)" on page 729. Next, Atoll calculates the uplink carrier to noise
ratio as explained in "Traffic C/N Calculation (UL)" on page 734. Finally, determines the uplink C/(I+N) by dividing the
previously calculated uplink C/N by the uplink noise rise value of the cell as calculated in "Noise Rise Calculation (UL)" on
page 733.
The uplink noise rise can be set by the user manually for each cell or calculated using Monte Carlo simulations.
Input
Mi
• CNR UL : Uplink C/N from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic) as calculated in "Traffic C/N
TX i  ic 
• NR UL : Non-segmented uplink noise rise for the cell TXi(ic) as calculated in "Noise Rise Calculation (UL)" on
page 733.
TX i  ic 
• NR UL – Seg : Segmented uplink noise rise for the cell TXi(ic) as calculated in "Noise Rise Calculation (UL)" on page 733.
TX i  ic 
• NR UL    : Angular distribution of the uplink noise rise for the cell TXi(ic) as calculated in "Noise Rise Calculation
(UL)" on page 733.
TX  ic 
i
TX i  ic 
• T B – Lowest : Bearer selection threshold of the lowest bearer in the WiMAX equipment assigned to the cell TXi(ic).
Mi
PZ UL
M
i
PZ UL = 8
• N SC  Seg : Number of subchannels per segment for the first uplink PUSC permutation zone.
Mi
Mi
• M PC : Power control margin defined in the global network settings.
Mi
• T B : Bearer selection thresholds of the bearers defined in the WiMAX equipment used bythe cell TXi(ic).
Mi
or mobile Mi.
Mi
or mobile Mi.
M
i
• N Ant – TX : Number of MIMO transmission (uplink) antennas defined for the terminal used by the pixel, subscriber, or
mobile Mi.
TX i  ic 
Mi
• Subchannel allocation mode used by the uplink permutation zone PZ UL assigned to the pixel, subscriber, or mobile
737
M
BLER  BUL : Uplink block error rate read from the graphs available in the WiMAX equipment assigned to the cell
i
•
TXi(ic).
Calculations
The uplink C/(I+N) for any pixel, subscriber, or mobile Mi at a cell TXi(ic) is calculated as follows:
• Without smart antennas:
For any pixel, subscriber, or mobile Mi covered by the non-segmented zone in the interfered cell TXi(ic):
CINR UL = CNR UL – NRUL
For any pixel, subscriber, or mobile Mi covered by the segmented zone in the interfered cell TXi(ic):
CINR UL = CNR UL – NRUL – Seg
• With smart antennas:

• Monte Carlo simulations: The uplink C/(I+N) is calculated as described in the section "Beamforming Smart
Antenna Models" on page 43. Victim and interfering mobiles are generated by a time-slot scenario as explained
in "Simulation Process" on page 699.
M M TX  ic 
i i i
• Coverage predictions: CINR UL    = CNR UL – NR UL 
The bearers available for selection in the cell TXi(ic)’s WiMAX equipment are the ones:
Mi Mi Mi Mi
• Whose selection thresholds are less than the uplink C/(I+N) at Mi: T B  CINR UL and T B  CINR UL   
TX  ic 
i
If the cell supports MIMO, the STTD/MRC, SU-MIMO diversity or MU-MIMO diversity gain, G Div – UL ,
M TX  ic  M
i i i
WiMAX equipment assigned to the cell TXi(ic) for N Ant – RX , N Ant – TX , the subchannel allocation mode of PZ UL ,
M
Mobility  M i  , BLER  B UL .
i
 
UL
UL
T B – G Div – UL – G Div  CINR UL and
UL
T B – G Div – UL – G Div  CINR UL   
• Bearer Index
From among the bearers available for selection, the selected bearer is the one with the highest uplink peak MAC
738
MIMO – STTD/MRC, SU-MIMO Diversity, and MU-MIMO Diversity Gain:
Once the bearer is known, the uplink C/(I+N) calculated above becomes:
UL
CNR UL = CNR UL + G Div – UL + G Div
UL
CINR UL = CINR UL + G Div – UL + G Div and
UL
CINR UL    = CINR UL    + G Div – UL + G Div
TX i  ic 
Where G Div – UL is the STTD/MRC, SU-MIMO diversity, or MU-MIMO diversity gain corresponding to the selected
bearer.
Uplink Subchannelisation:
The uplink subchannelisation depends on the uplink bandwidth allocation target defined for the scheduler used by the
cell TXi(ic). The uplink C/(I+N) calculated above is given for the total number of subchannels associated with the
Mi
PZ UL
permutation zone, i.e., N SC . Subchannelisation is performed for all the pixels, subscribers, or mobiles in the uplink,
and may reduce the number of used subchannels in order to satisfy the selected target.
• Full Bandwidth
transmission, there is no gain in the uplink C/(I+N).
The bandwidth used for transmission by a mobile is reduced only if the uplink C/(I+N) is not enough to even access
the lowest bearer. For example, as a mobile moves from good to bad radio conditions, the number of subchannels
used by it for transmission in uplink are reduced one by one in order to improve the uplink C/(I+N). The calculation
of the gain introduced by the subchannelisation is explained below.
• Best Bearer
The bandwidth used for transmission by a mobile is reduced in order to improve the uplink C/(I+N) enough to
access the best bearer. For example, if using 5 subchannels, a mobile is able to access the best bearer, and using
6 it would only get access to the second best, it will be assigned 5 subchannels as the used uplink bandwidth.
Although using 4 subchannels, its uplink C/(I+N) will be better than when using 5, the uplink bandwidth is not
reduced to 4 because it does not provide any gain in terms of the bearer, i.e., the mobile already has the best
bearer using 5 subchannels. The calculation of the gain introduced by the bandwidth reduction is explained below.
i.e., bearer with the highest index, with the highest peak MAC throughput, or with the highest effective MAC
throughput.
The uplink subchannelisation may result in the use of a number of subchannels which is less than the total number of
subchannels associated with the permutation zone. The gain related to this bandwidth reduction is applied to the
uplink C/(I+N):
Mi
 PZUL 
Mi Mi  N SC 
CINR UL = CINR UL+ 10  Log  -----------------
Final All SC  NMi 
 SC – UL 
Mi
M PZ
Min i UL
Where N SC – UL  Service   N SC – UL  N SC for any pixel, subscriber, or mobile Mi covered by a non-segmented
Mi
Min Mi PZ UL = 8
permutation zone in the interfered cell TXi(ic), and N SC – UL  Service   N SC – UL  N SC  Seg for any pixel, subscriber,
or mobile Mi covered by the segmented uplink PUSC zone in the interfered cell TXi(ic).
Uplink Power Control:
739
Once the subchannelisation is performed, Atoll continues to work with the C/(I+N) given by the subchannelisation,
M M
i i
i.e., CINR UL = CINR UL .
Final
The pixel, subscriber, or mobile Mi reduces its transmission power so that the uplink C/(I+N) from it at its cell is just
enough to get the selected bearer.
If with P = P Max AND CINR UL  T Mi + M PC , where T Mi is the bearer selection threshold, from the WiMAX
B UL B UL
equipment assigned to the cell TXi(ic), for the bearer selected for the pixel, subscriber, or mobile Mi.
Mi  Mi  M i  TX i  ic    Mi 
P Eff = Max  PMax –  CINR UL –  T M + M PC   P Min
   B i  
UL
Mi Mi
CINR UL is calculated again using P Eff .
Output
Mi Mi
• CINR UL or CINR UL    : Uplink C/(I+N) from a pixel, subscriber, or mobile Mi at it serving cell TXi(ic).
Mi
• N SC – UL : Number of subchannels used by the pixel, subscriber, or mobile Mi in the uplink after subchannelisation.
Mi
• P Eff : Effective transmission power of the terminal used by the pixel, subscriber, or mobile Mi.
M
i
• B UL : Bearer assigned to the pixel, subscriber, or mobile Mi in the uplink.
10.4.7 Throughput Calculation

Throughputs are calculated in two steps.
• Calculation of uplink and downlink total resources in a cell as explained in "Calculation of Total Cell Resources" on
page 740.
• Calculation of throughputs as explained in "Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput, and
10.4.7.1 Calculation of Total Cell Resources

The total amount of resources in a cell is the number of modulation symbols that can be used for data transfer in each frame.
The total cell resources can be calculated separately for the downlink and the uplink subframes. The following sections
describe how the cell capacities are calculated for TDD and FDD networks.
10.4.7.1.1 Calculation of Sampling Frequency

Input
TX  ic 
i
• f Sampling : Sampling factor defined for the frequency band of the cell TXi(ic).
TX i  ic 
• W Channel : Channel bandwidth of the cell TXi(ic).
Calculations
Atoll determines the sampling frequency as follows:

TX  ic 
i 6
TX i  ic   W Channel  10 
F Sampling = Floor  f Sampling  -----------------------------------
-  8000
 8000 
Output
TX i  ic 
• F Sampling : Sampling frequency for the cell TXi(ic).
740
10.4.7.1.2 Calculation of Symbol Duration

Input
TX  ic 
i
TX i  ic 
TX i  ic 
• r CP : Cyclic prefix ratio defined for the frame configuration of TXi(ic) or, otherwise, in the global network settings.
Calculations
From the sampling frequency, Atoll determines the inter-subcarrier spacing.

TX i  ic  –3
TX  ic 
i F Sampling  10
F = -------------------------------------
-
TX i  ic 
N SCa – Total
Atoll calculates the useful symbol duration.

TX i  ic 
1
D Sym – Useful = -------------------
TX  ic 
i
F
And, the duration of the cyclic prefix.
TX i  ic 
TX i  ic  r CP
D CP = --------------
-
F
Adding the Cyclic prefix ratio to the useful symbol duration, Atoll determines the total symbol duration.
D Symbol = D Sym – Useful + D CP
Output
TX i  ic 
• D Symbol : Total symbol duration of one modulation symbol for a cell TXi(ic).
10.4.7.1.3 Calculation of Total Cell Resources - TDD Networks

Input

TDD
• D TTG : TTG duration.
TDD
• D RTG : RTG duration.
TX i  ic 
• D Symbol : Total symbol duration of one modulation symbol for a cell TXi(ic) as calculated in "Calculation of Symbol
Duration" on page 741.
TDD
• r DL – Frame : DL ratio.
TDD
• N SD – DL : Number of symbol durations that correspond to the downlink subframe.
TDD
• N SD – UL : Number of symbol durations that correspond to the uplink subframe.
DL
• O Fixed : Downlink fixed overhead.
DL
• O Variable : Downlink variable overhead.
UL
• O Fixed : Uplink fixed overhead.
UL
• O Variable : Uplink variable overhead.
Mi
PZ DL
• N SCa – Data : Number of data subcarriers of the downlink permutation zone of a cell TXi(ic) assigned to Mi.
Mi
PZ UL
• N SCa – Data : Number of data subcarriers of the uplink permutation zone of a cell TXi(ic) assigned to Mi.
741
Calculations
The downlink and the uplink subframes of a TDD frame are separated in time by the TTG and the RTG time guards.
First of all, Atoll calculates the useful frame duration by removing the TTG and RTG from the frame duration:
Used TDD TDD
D Frame = D Frame – D TTG – D RTG
Then, Atoll calculates the frame duration in terms of number of symbol durations:
TX  ic   D Used 
N  SD – Used   Frame = Floor  -----------------
i Frame 
-
TX i  ic  
 D Symbol
Next, Atoll calculates the downlink and uplink cell capacities as follows:
Downlink Subframe:
Atoll calculates the number of symbol durations in the downlink subframe excluding the fixed overhead defined in the
global network settings:
N  SD – DL   Subframe = RoundUp  N SD – Used   Frame  r DL – Frame – O Fixed if DL:UL ratio is defined in percentage.
i i TDD DL
TDD
TX i  ic   TXi  ic  N SD – DL  DL
Or N  SD – DL   Subframe = RoundUp  N  SD – Used   Frame  ------------------------------------------
TDD TDD
 – O Fixed if DL:UL ratio is defined in
fraction.
The RoundUp function rounds a float value up to the nearest integer value.
The total number of symbols in the downlink subframe after removing the variable overhead is:
 Mi
DL 
TX i  ic  TX i  ic   TXi  ic  PZ DL O Variable 
R DL = N  Sym – DL   Subframe = Floor  N  SD – DL   Subframe  N SCa – Data   1 – --------------------
- 
 100  

 
Uplink Subframe:
Atoll calculates the number of symbol durations in the uplink subframe excluding the fixed overhead defined in the
global network settings:
N  SD – UL   Subframe = RoundDown  N SD – Used   Frame   1 – r DL – Frame  – O Fixed
i i TDD UL
if DL:UL ratio is defined in
percentage.
TDD
TX i  ic   TX i  ic  N SD – UL  UL
Or N  SD – UL   Subframe = RoundDown  N SD – Used   Frame  ------------------------------------------
TDD TDD
 – O Fixed if DL:UL ratio is defined in
fraction.
The RoundDown function rounds a float value down to the nearest integer value.
The total number of symbols in the uplink subframe after removing the variable overhead is:
 Mi
UL 
TX  ic 
i
TX  ic 
i  TX i  ic  PZ
UL  O Variable 
R UL = N  Sym – UL   Subframe = Floor  N  SD – UL   Subframe  N SCa – Data   1 – --------------------- 
 100 
 
Output
• R DL = N  Sym – DL   Subframe : Amount of downlink resources in the cell TXi(ic).
• R UL = N  Sym – UL   Subframe : Amount of uplink resources in the cell TXi(ic).
10.4.7.1.4 Calculation of Total Cell Resources - FDD Networks

The total cell resources calculation is the same for downlink and uplink subframes in FDD networks. Therefore, the symbol X
is used to represent DL or UL in the expressions below.
742
Input

TX  ic 
i
• D Symbol : Total symbol duration of one modulation symbol for a cell TXi(ic) as calculated in "Calculation of Symbol
Duration" on page 741.
X
• O Fixed : Downlink or uplink fixed overhead.
X
• O Variable : Downlink or uplink variable overhead.
Mi
PZ X
• N SCa – Data : Number of data subcarriers of the downlink or uplink permutation zone of a cell TXi(ic) assigned to Mi.
Calculations
There are no transmit and receive time guards in FDD systems. Therefore, the downlink and the uplink subframe durations
are the same as the frame duration.
X
D Subframe = D Frame
The subframe durations in terms of the number of symbol durations excluding the fixed overheads are:
TX  ic   DX 
N  SD – X   Subframe = Floor  ----------------------
i Subframe
- – O XFixed
TX i  ic  
 D Symbol 
The total numbers of symbols in the downlink or uplink subframes after removing the variable overheads are:
 Mi
X 
TX i  ic  TX i  ic   TXi  ic  PZ X O Variable 
RX = N  Sym – X   Subframe = Floor  N  SD – X   Subframe  N SCa – Data   1 – --------------------
- 
 100  

 
Output
• RX = N  Sym – X   Subframe : Amount of downlink or uplink resources in the cell TXi(ic).
10.4.7.2 Channel Throughput, Cell Capacity, Allocated Bandwidth Throughput,

and Per-User Throughput Calculation
Channel throughputs are calculated for the entire channel resources allocated to the pixel, subscriber, or mobile Mi. Cell
capacities are similar to channel throughputs but upper-bound by the maximum downlink and uplink traffic loads. Allocated
bandwidth throughputs are calculated for the number of used subchannels in uplink allocated to the pixel, subscriber, or
mobile Mi. Per-user throughputs are calculated by dividing the downlink cell capacities or uplink allocated bandwidth
throughputs by the average number of downlink or uplink users defined for the cell, respectively.
Input
TX i  ic 
TX i  ic 
TX i  ic 
• R DL : Amount of downlink resources in the cell TXi(ic) as calculated in "Calculation of Total Cell Resources" on
page 740.
TX i  ic 
• R UL : Amount of uplink resources in the cell TXi(ic) as calculated in "Calculation of Total Cell Resources" on
page 740.
•  M : Bearer efficiency (bits/symbol) of the bearer assigned to the pixel, subscriber, or mobile Mi in the downlink in
i
B DL
"Traffic and Pilot C/(I+N) and Bearer Calculation (DL)" on page 727.
•  M : Bearer efficiency (bits/symbol) of the bearer assigned to the pixel, subscriber, or mobile Mi in the uplink in
i
B UL
"Traffic C/(I+N) and Bearer Calculation (UL)" on page 737.
743

• f Segment – DL : Downlink segmenting factor for the first downlink PUSC zone as calculated in "Effective Traffic and Pilot
Interference Calculation (DL)" on page 722.
TX  ic 
i
• CNR Preamble : Preamble C/N the cell TXi(ic) as calculated in "Preamble C/N Calculation" on page 712.
TX  ic 
i
TX i  ic 
• T MU – MIMO : MU-MIMO threshold defined for the cell TXi(ic).
TX i  ic 
• G MU – MIMO : MU-MIMO gain defined for the cell TXi(ic).
M TX  ic 
BLER  BDL : Downlink block error rate read from the BLER vs. CINR Traffic graph available in the WiMAX equipment
i i
•
 
assigned to the terminal used by the pixel, subscriber, or mobile Mi.
M M
BLER  BUL : Uplink block error rate read from the BLER vs. CINR UL graph available in the WiMAX equipment assigned
i i
•
 
to the cell TXi(ic).
Mi
• f TP – Scaling : Throughput scaling factor defined in the properties of the service used by the pixel, subscriber, or mobile
Mi.
M
i
• TP Offset : Throughput offset defined in the properties of the service used by the pixel, subscriber, or mobile Mi.
Mi
PZ UL
Mi
• N SC – UL : Number of uplink subchannels after subchannelisation with which the pixel, subscriber, or mobile Mi can get
the highest available bearer, as calculated in "Traffic C/(I+N) and Bearer Calculation (UL)" on page 737.
TX i  ic 
• N Users – DL : Number of users connected to the cell TXi(ic) in downlink.
TX i  ic 
• N Users – UL : Number of users connected to the cell TXi(ic) in uplink.
Calculations
Downlink:
TX i  ic 
R DL  M
Mi i
B DL
• Peak MAC Channel Throughput: CTP P – DL = ---------------------------------
D Frame
TX  ic 
i
For proportional fair schedulers, the channel throughput is increased by the multi-user diversity gain G MUG – DL read
from the scheduler properties for the Mobility  M i  and the number of users connected to the cell in downlink.
TX i  ic 
R DL  M
i
Mi B TX  ic 
i
DL
CTP P – DL = ---------------------------------  G MUG – DL
D Frame
TX  ic  M
i i Max
G MUG – DL = 1 if CINR Traffic  CINR MUG
number of users.
Downlink Segmentation:
M
i
If the permutation zone assigned to the pixel, subscriber, or mobile Mi is the first downlink PUSC zone ( PZ DL = 0 )
and it is segmented, the channel throughput is calculated as:
744
M M
i i
CTP P – DL = CTP P – DL  f Segment – DL

If the permutation zone assigned to the pixel, subscriber, or mobile Mi supports SU-MIMO or AMS, SU-MIMO gain
Max
G SU – MIMO is applied to the bearer efficiency. The gain is read from the properties of the WiMAX equipment assigned
to the pixel, subscriber, or mobile Mi for:
TX  ic 
i
Mi
• N Ant – RX : Number of MIMO reception (downlink) antennas defined for the terminal used by the pixel, subscriber,
or mobile Mi.
Mi
• Subchannel allocation mode used by the downlink permutation zone PZ DL assigned to the pixel, subscriber, or
mobile Mi as calculated in "Permutation Zone Selection" on page 714.
M
i
• B DL : Bearer assigned to the pixel, subscriber, or mobile Mi in the downlink as explained in "Traffic and Pilot C/
(I+N) and Bearer Calculation (DL)" on page 727.
M
BLER  B DL : Downlink block error rate read from the graphs available in the WiMAX equipment assigned to the
i
•
TX  ic 
i
terminal used by the pixel, subscriber, or mobile Mi. BLER is determined for CINR Traffic .
Max
In case of SU-MIMO:  Mi =  Mi   1 + f SU – MIMO  G SU – MIMO – 1  
B DL B DL
Max
In case of AMS:  Mi =  Mi   1 + f SU – MIMO  G SU – MIMO – 1   if CNR Preamble  T AMS or CINR Preamble  T AMS
B DL B DL
M M M
Effective MAC Channel Throughput: CTP E – DL = CTP P – DL   1 – BLER  B DL 
i i i
•
  
Mi
M
i i
i
• Application Channel Throughput: CTP A – DL = CTP E – DL  ------------------------
- – TP Offset
100
• Peak MAC Cell Capacity: Cap P – DL = CTP P – DL  TL DL – Max
M M M
Effective MAC Cell Capacity: Cap E – DL = Cap P – DL   1 – BLER  B DL 
i i i
•
  
Mi
i
M
i
i
• Application Cell Capacity: Cap A – DL = Cap E – DL  ------------------------
- – TP Offset
100
Mi
Mi Cap P – DL
• Peak MAC Throughput per User: PUTP P – DL = -----------------------
-
TX i  ic 
N Users – DL
Mi
Mi Cap E – DL
• Effective MAC Throughput per User: PUTP E – DL = -----------------------
-
TX i  ic 
N Users – DL
Mi
Mi f TP – Scaling
Mi Mi
• Application Throughput per User: PUTP A – DL = PUTP E – DL  ------------------------
- – TP Offset
100
745
Uplink:
TX  ic 
i
R UL  M
M i
i B
UL
• Peak MAC Channel Throughput: CTP P – UL = ---------------------------------
D Frame
TX i  ic 
For proportional fair schedulers, the channel throughput is increased by the multi-user diversity gain G MUG – UL read
from the scheduler properties for the Mobility  M i  and the number of users connected to the cell in uplink.
TX i  ic 
R UL  Mi
Mi B TX  ic 
UL i
CTP P – UL = ---------------------------------  G MUG – UL
D Frame
TX i  ic  Mi
Max
G MUG – UL = 1 if CINR UL  CINR MUG
number of users.
If the permutation zone assigned to the pixel, subscriber, or mobile Mi supports SU-MIMO or AMS, SU-MIMO gain
Max
G SU – MIMO is applied to the bearer efficiency. The gain is read from the properties of the WiMAX equipment assigned
to the cell TXi(ic) for:
Mi
• N Ant – TX : Number of MIMO transmission (uplink) antennas defined for the terminal used by the pixel, subscriber,
or mobile Mi.
TX  ic 
i
Mi
• Subchannel allocation mode used by the uplink permutation zone PZ UL assigned to the pixel, subscriber, or
mobile Mi as calculated in "Permutation Zone Selection" on page 714.
Mi
• B UL : Bearer assigned to the pixel, subscriber, or mobile Mi in the uplink as explained in "Traffic C/(I+N) and Bearer
M
BLER  B UL : Uplink block error rate read from the graphs available in the WiMAX equipment assigned to the cell
i
•
 
Mi
TXi(ic). BLER is determined for CINR UL .
Max
In case of SU-MIMO:  Mi =  Mi   1 + fSU – MIMO  G SU – MIMO – 1  
B UL B UL
Max
In case of AMS:  Mi =  Mi   1 + f SU – MIMO  G SU – MIMO – 1   if CNR Preamble  T AMS or CINR Preamble  T AMS
B UL B UL
MIMO – MU-MIMO Gain (for uplink throughput coverage predictions only):
If the permutation zone assigned to the pixel, subscriber, or mobile Mi supports MU-MIMO and
i i i i
CNR Preamble  T MU – MIMO and N Ant – RX  2 , the MU-MIMO gain G MU – MIMO is applied to the channel throughput.
The MU-MIMO gain is read from the properties of the cell TXi(ic).
746
M M TX  ic 
i i i
CTP P – UL = CTP P – UL  G MU – MIMO
M M M
Effective MAC Channel Throughput: CTP E – UL = CTP P – UL   1 – BLER  B UL 
i i i
•
  
M
i
• Application Channel Throughput: CTP A – UL = CTP E – UL  ------------------------
- – TP Offset
100
• Peak MAC Cell Capacity: Cap P – UL = CTP P – UL  TL UL – Max
M M M
Effective MAC Cell Capacity: Cap E – UL = Cap P – UL   1 – BLER  B UL 
i i i
•
  
Mi
• Application Cell Capacity: Cap A – UL = Cap E – UL  ------------------------
- – TP Offset
100
Mi
Mi Mi N SC – UL
• Peak MAC Allocated Bandwidth Throughput: ABTP P – UL = CTP P – UL  -----------------
M
i
PZ
UL
N SC
M M M
Effective MAC Allocated Bandwidth Throughput: ABTP E – UL = ABTP P – UL   1 – BLER  B UL 
i i i
•
  
Mi
• Application Allocated Bandwidth Throughput: ABTPA – UL = ABTP E – UL  ------------------------
- – TPOffset
100
Mi  Cap M i M 
= Min  -----------------------
- ABTPP –i UL
P – UL
• Peak MAC Throughput per User: PUTP P – UL TX i  ic  
Mi  Cap Mi M 
= Min  -----------------------
E – UL
• Effective MAC Throughput per User: PUTP E – UL
 TXi  ic  
Mi
• Application Throughput per User: PUTP A – UL = PUTP E – UL  ------------------------
- – TP Offset
100
Output
Mi
• CTP P – DL : Downlink peak MAC channel throughput at the pixel, subscriber, or mobile Mi.
Mi
• CTP E – DL : Downlink effective MAC channel throughput at the pixel, subscriber, or mobile Mi.
M
i
• CTP A – DL : Downlink application channel throughput at the pixel, subscriber, or mobile Mi.
Mi
• Cap P – DL : Downlink peak MAC cell capacity at the pixel, subscriber, or mobile Mi.
Mi
• Cap E – DL : Downlink effective MAC cell capacity at the pixel, subscriber, or mobile Mi.
Mi
• Cap A – DL : Downlink application cell capacity at the pixel, subscriber, or mobile Mi.
Mi
• PUTP P – DL : Downlink peak MAC throughput per user at the pixel, subscriber, or mobile Mi.
M
i
• PUTP E – DL : Downlink effective MAC throughput per user at the pixel, subscriber, or mobile Mi.
Mi
• PUTP A – DL : Downlink application throughput per user at the pixel, subscriber, or mobile Mi.
Mi
• CTP P – UL : Uplink peak MAC channel throughput at the pixel, subscriber, or mobile Mi.
Mi
• CTP E – UL : Uplink effective MAC channel throughput at the pixel, subscriber, or mobile Mi.
Mi
• CTP A – UL : Uplink application channel throughput at the pixel, subscriber, or mobile Mi.
M
i
• Cap P – UL : Uplink peak MAC cell capacity at the pixel, subscriber, or mobile Mi.
747
M
i
• Cap E – UL : Uplink effective MAC cell capacity at the pixel, subscriber, or mobile Mi.
M
i
• Cap A – UL : Uplink application cell capacity at the pixel, subscriber, or mobile Mi.
M
i
• ABTP P – UL : Uplink peak MAC allocated bandwidth throughput at the pixel, subscriber, or mobile Mi.
Mi
• ABTP E – UL : Uplink effective MAC allocated bandwidth throughput at the pixel, subscriber, or mobile Mi.
Mi
• ABTP A – UL : Uplink application allocated bandwidth throughput at the pixel, subscriber, or mobile Mi.
Mi
• PUTP P – UL : Uplink peak MAC throughput per user at the pixel, subscriber, or mobile Mi.
Mi
• PUTP E – UL : Uplink effective MAC throughput per user at the pixel, subscriber, or mobile Mi.
Mi
• PUTP A – UL : Uplink application throughput per user at the pixel, subscriber, or mobile Mi.

Atoll WiMAX BWA module includes a number of scheduling methods which can be used for scheduling and radio resource
allocation during Monte Carlo simulations. These resource allocation algorithms are explained in "Scheduling and Radio
Resource Allocation" on page 748 and the calculation of user throughputs is explained in "User Throughput Calculation" on
page 757.
10.4.8.1 Scheduling and Radio Resource Allocation

Input
TX i  ic 
TX i  ic 
TX i  ic 
• N Users – Max : Maximum number of users defined for the cell TXi(ic).
M
i
• QoS : QoS class of the service (UGS, ErtPS, rtPS, nrtPS, or Best Effort) accessed by a mobile Mi.
M
i
• p : Priority of the service accessed by a mobile Mi.
M
i
• TPD Min – DL : Downlink minimum throughput demand for the service accessed by a mobile Mi.
M
i
• TPD Min – UL : Uplink minimum throughput demand for the service accessed by a mobile Mi.
Mi
• TPD Max – DL : Downlink maximum throughput demand for the service accessed by a mobile Mi.
Mi
• TPD Max – UL : Uplink maximum throughput demand for the service accessed by a mobile Mi.
M TX  ic 
BLER  BDL : Downlink block error rate read from the BLER vs. CINR Traffic graph available in the WiMAX equipment
i i
•
 
assigned to the terminal used by the mobile Mi.
M M
BLER  BUL : Uplink block error rate read from the BLER vs. CINR UL graph available in the WiMAX equipment assigned
i i
•
to the cell TXi(ic).
Mi
• f TP – Scaling : Throughput scaling factor defined in the properties of the service used by the mobile Mi.
Mi
• TP Offset : Throughput offset defined in the properties of the service used by the mobile Mi.
Mi
• CTP P – DL : Downlink peak MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
page 740.
Mi
• CTP E – DL : Downlink effective MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
page 740.
748
M
i
• CTP P – UL : Uplink peak MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
page 740.
M
i
• CTP E – UL : Uplink effective MAC channel throughput at the mobile Mi as calculated in "Throughput Calculation" on
page 740.
Mi
• ABTP P – UL : Uplink peak MAC allocated bandwidth throughput at the mobile Mi as calculated in "Throughput
QoS
• f Bias : Bias factor defined for the Biased (QoS Class) scheduling method.
Calculations
The following calculations are described for any cell TXi(ic) containing the users Mi for which it is the best server.
Mobile Selection:
TX  ic 
i
The scheduler selects N Users mobiles for the scheduling and RRM process. If the Monte Carlo user distribution has generated
TX i  ic 
a number of users which is less than N Users – Max , the scheduler keeps all the mobiles generated for the cell TXi(ic).
N Users = Min  N Users – Max N Users – Generated
i i i
 
TX  ic 
Sel i
For a cell, mobiles M i  N Users are selected for RRM by the scheduler.
Calculation of Actual Minimum and Maximum Throughput Demands:

Depending on the selected target throughput of the scheduler assigned to the cell TXi(ic), the actual minimum and maximum
throughput demands can be considered as the peak MAC, effective MAC, or application throughput. Therefore:
• Target Throughput = Peak MAC Throughput
Sel Sel
Mi Mi
Downlink: TPD Min – DL , TPD Max – DL
Sel Sel
M
i  M
i
M
i 
Uplink: TPD Min – UL , Min  TPD Max – UL ABTP P – UL
 
• Target Throughput = Effective MAC Throughput

Sel Sel
Sel Mi Sel Mi
Mi TPD Min – DL TPD Max – DL
Mi
Downlink: TPD Min – DL = ---------------------------------------------
- , TPD Max – DL = ---------------------------------------------
-
Sel Sel
  Mi     Mi  
 1 – BLER  B DL    1 – BLER  BDL  
     
Sel
Sel  Mi Mi 
Sel Mi Sel Min  TPD Max – UL ABTP P – UL
Mi TPD Min – UL Mi  
Uplink: TPD Min – UL = ---------------------------------------------
- , TPD Max – UL = -------------------------------------------------------------------------
Sel Sel
  Mi     Mi  
 1 – BLER  BUL    1 – BLER  B UL  
     
• Target Throughput = Application Throughput

Sel Sel
Sel Mi Mi Sel Mi Mi
Mi TPD Min – DL + TP Offset Mi TPD Max – DL + TP Offset
Downlink: TPD Min – DL = -----------------------------------------------------------------------------
- , TPD Max – DL = -----------------------------------------------------------------------------
-
Sel Sel
  Mi   Mi   Mi   Mi
 1 – BLER  B DL    f TP – Scaling  1 – BLER  B DL    f TP – Scaling
     
Sel
Sel Mi Mi
Mi TPD Min – UL + TP Offset
Uplink: TPD Min – UL = -----------------------------------------------------------------------------
-,
Sel
  Mi   Mi
  
749
Sel
 M
i
M
i 
M
i
Sel Min  TPD Max – UL ABTP P – UL + TP Offset
M
i  
TPD Max – UL = --------------------------------------------------------------------------------------------------
Sel
  Mi   Mi
  
The Min() function selects the lower of the two values. This calculation is performed in order to limit the maximum uplink
throughput demand to the maximum throughput that a user can get in uplink using the allocated bandwidth (number of used
subchannels) calculated for it in "Traffic C/(I+N) and Bearer Calculation (UL)" on page 737.
Resource Allocation for Minimum Throughput Demands:
TX i  ic 
Sel
1. For the QoS classes UGS, ErtPS, rtPS, and nrtPS, Atoll sorts the M i  N Users in order of decreasing service priority,
Sel
Mi
p :
Sel Sel Sel

Mi Mi
Mi
QoS p
Sel
1 UGS Mi
p =n
Sel
2 Mi
... n > p > 0 ...
Sel
: Mi
p =0
Sel
: ErtPS Mi
p =n
Sel
: Mi
... n > p > 0 ...
Sel
: Mi
p =0
Sel
: rtPS M
i
p =n
Sel
: M
i
... n > p > 0 ...
Sel
: Mi
p =0
Sel
: nrtPS Mi
p =n
Sel
N–1 Mi
... n > p > 0 ...
Sel
N Mi
p =0
Where N  N Users , if there are some Best Effort users, or N = N Users if there are no Best Effort users selected.
Sel Sel
2. Starting with M i = 1 up to M i = N , Atoll allocates the downlink and uplink resources required to satisfy each
user’s minimum throughput demands in downlink and uplink as follows:
Sel Sel
Sel Mi Sel Mi
Mi TPD Min – DL TPD Min – UL
Mi
R Min – DL = --------------------------
- and R Min – UL = --------------------------
-
Sel Sel
Mi Mi
3. Atoll stops the resource allocation in downlink or uplink,
750
Sel
M TX  ic 

i i
• When/If in downlink R Min – DL = TL DL – Max , i.e., the resources available in downlink have been used up for
Sel
M
i
Sel
M TX  ic 

i i
• When/If in uplink R Min – UL = TL UL – Max , i.e., the resources available in uplink have been used up for
Sel
Mi

4. Mobiles which are active DL+UL must be able to get their minimum throughput demands in both UL and DL in order
to be considered connected DL+UL. If an active DL+UL mobile is only able to get its minimum throughput demand in
one direction, it is rejected, and the resources, that were allocated to it in the one direction in which it was able to get
a throughput, are allocated to other mobiles.
5. Mobiles which are active UL and whose minimum throughput demand in UL is higher than the uplink allocated
Sel Sel
Mi Mi
bandwidth throughput ( TPD Min – UL  ABTP P – UL ) are rejected due to Resource Saturation.
Sel Sel
6. If  RMin – DL  TLDL – Max or  RMin – UL  TLUL – Max , and all the minimum throughput resources demanded by
Sel Sel
Mi Mi
the mobiles have been allocated, Atoll goes to the next step for allocating resources to satisfy the maximum
throughput demands.
Backhaul Saturation:
If at this stage, a site’s downlink or uplink effective MAC aggregate throughput exceeds its maximum downlink or uplink
backhaul throughput, respectively, mobiles are rejected one by one due to Backhaul Saturation, starting from the mobile with
the lowest priority service, among all the cells of the site in order to reach a downlink or uplink effective MAC aggregate site
throughput ≤ the site’s maximum downlink or uplink backhaul throughput.
Resource Allocation for Maximum Throughput Demands:
For each cell, the remaining cell resources available are:
Sel

i i i
Downlink: R Rem – DL = TL DL – Max – R Min – DL
Sel
Mi
Sel
Uplink: R Rem – UL = TL UL – Max –  R Min – UL
Sel
M
i
For each mobile, the throughput demands remaining once the minimum throughput demands have been satisfied are the
difference between the maximum and the minimum throughput demands:
Sel Sel Sel
M M M
i i i
Downlink: TPD Rem – DL = TPD Max – DL – TPD Min – DL
Sel Sel Sel

Mi Mi Mi
Uplink: TPD Rem – UL = TPD Max – UL – TPD Min – UL
For the remaining throughput demands of the mobiles belonging to the QoS classes ErtPS, rtPS, nrtPS, and Best Effort, the
following resource allocation methods are available:
• Proportional Fair:
The goal of this scheduling method is to distribute resources among users fairly in such a way that, on the average,
each user gets the highest possible throughput that it can get under the radio conditions at its location.
751
Sel
Let the total number of users belonging to the QoS classes ErtPS, rtPS, nrtPS, and Best Effort, be N  M i .
i i
a. Each user’s channel throughput is increased by the multi-user diversity gain G MUG – DL or G MUG – UL read from the
Sel
scheduler properties for the Mobility  M i  assigned to mobile M i and the number of connected users, DL or
UL, in the cell TXi(ic) in the iteration k-1.
Sel Sel Sel Sel
Mi Mi TX i  ic  Mi Mi TX i  ic 
CTP P – DL = CTP P – DL  G MUG – DL and CTP P – UL = CTP P – UL  G MUG – UL
Without MUG Without MUG
Sel Sel
Max Max
G MUG – DL = 1 if CINR Traffic  CINR MUG and G MUG – UL = 1 if CINR UL  CINR MUG .
If the multi-user diversity gain for the exact value of the number of connected users is not available in the graph,
it is interpolated from the gain values available for the numbers of users just less than and just greater than the
actual number of users.
b. Atoll divides the remaining resources in the cell into equal parts for each user:
--------------------- and --------------------
-
N N
c. Atoll converts the remaining throughput demands of all the users to their respective remaining resource
demands:
Sel Sel
Sel M Sel M
i i
RD Rem – DL = ---------------------------
- and RD Rem – UL = ---------------------------
-
Sel Sel
Mi Mi
d. The resources allocated to each user by the Proportional Fair scheduling method for satisfying its maximum

Mi  Mi R Rem – DL Mi  Mi R Rem – UL
 N   N 
e. Atoll stops the resource allocation in downlink or uplink,
Sel
Mi TX i  ic 
Sel
M
i
Sel
Mi TX i  ic 
• When/If in uplink  RMax – UL = R Rem – UL , i.e., the resources available in uplink have been used up for
Sel
Mi

f. If the resources allocated to a user satisfy its maximum throughput demands, this user is removed from the list of
remaining users.
g. Atoll recalculates the remaining resources as follows:
Sel Sel
R Rem – DL = TL DL – Max –  R Min – DL –  R Max – DL and
Sel Sel
Mi Mi
Sel Sel
R Rem – UL = TL UL – Max –  RMin – UL –  RMax – UL
Sel Sel
Mi Mi
752
h. Atoll repeats the all the above steps for the users whose maximum throughput demands have not been satisfied
i i
• Proportional Demand:
The goal of this scheduling method is to allocate resources to users weighted according to their remaining throughput
demands. Therefore, the user throughputs for users with high throughput demands will be higher than those with low
throughput demands. In other words, this scheduler distributes channel throughput between users proportionally to
their demands.
a. Atoll converts the remaining throughput demands of all the users to their respective remaining resource
demands:
Sel Sel
Sel Mi Sel Mi
RD Rem – DL = ---------------------------
- and RD Rem – UL = ---------------------------
-
Sel Sel
Mi Mi
b. Atoll calculates the amount effective remaining resources for the cell of each user to distribute among the users
as follows:
 Sel   Sel 
TX i  ic   TXi  ic  Mi  TX i  ic   TXi  ic  Mi 
R Eff – Rem – DL = Min  R Rem – DL
 
RD Rem – DL and R Eff – Rem – UL = Min  R Rem – UL
 
RD Rem – UL
 
 Sel
Mi   Sel
Mi 
c. The resources allocated to each user by the Proportional Demand scheduling method for satisfying its maximum
Sel Sel
Sel Mi Sel Mi
Mi TX i  ic  RD Rem – DL Mi TX i  ic 
RD Rem – UL
R Max – DL = R Eff – Rem – DL  ----------------------------------
Sel
- and R Max – UL = R Eff – Rem – UL  ----------------------------------
Sel
-
Mi Mi
 RDRem – DL  RDRem – UL
Sel Sel
Mi Mi
• Biased (QoS Class):

The goal of this scheduling method is to distribute resources among users of each QoS class fairly in such a way that,
on the average, each user gets the highest possible throughput that it can get under the radio conditions at its
location. The resources available for allocation to users of each QoS class depend on a bias factor. The QoS Class Bias
Factor controls the amount of resources available for each QoS class.
Calculation of the Remaining Resources per QoS Class:
QoS
The bias factor f Bias represents the bias in terms of resources allocated to 1 user of a QoS class with rank r to the
resources allocated to 1 user of a QoS class with rank r–1:
Sel Sel Sel
Mi Mi Mi
QoS
f Bias R Max – ErtPS R Max – rtPS R Max – nrtPS
 = 1 + ---------
- = -------------------------- = -------------------------- = --------------------------
100 Mi
Sel
Mi
Sel
Mi
Sel
R Max – rtPS R Max – nrtPS R Max – BE
The ranks of QoS classes are:
QoS Class QoS Class Rank r QoS
ErtPS 1
rtPS 2
nrtPS 3
Best Effort 4
The resources available for the users of each QoS class from among the remaining resources is calculated as follows:
753
r r
1 QoS 1 QoS
N QoS   --- N QoS   ---
TX  ic 
i
TX  ic 
i  TX  ic 
i
TX  ic 
i 
R QoS – DL = R Rem – DL  ------------------------------------------------------- and R QoS – UL = R Rem – UL  -------------------------------------------------------
r r
1 QoS 1 QoS
N QoS   ---
 N QoS   ---

   
All QoS All QoS
Resource Allocation:
Once the remaining resources available for the users of each QoS class have been determined, the allocation of
resources within each QoS class is performed as for the proportional fair scheduler.
Sel
Let the number of users belonging to a QoS class N QoS  M i .
a. Atoll divides the remaining resources of the QoS class into equal parts for each user:
R QoS – DL R QoS – UL
-------------------- and -------------------
-
N QoS N QoS
b. Atoll converts the remaining throughput demands of all the users to their respective remaining resource
demands:
Sel Sel
Sel Mi Sel Mi
RD Rem – DL = ---------------------------
- and RD Rem – UL = ---------------------------
-
Sel Sel
Mi Mi
c. The resources allocated to each user by the Biased scheduling method for satisfying its maximum throughput

Atoll 3.3.0 Technical Reference Guide PDF

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Atoll 3.3.0 Technical Reference Guide PDF

Uploaded by

Copyright:

Available Formats

Version 3.3.

Technical Reference Guide

Atoll 3.3.0 Technical Reference Guide for Radio Networks

Printing Help Topics

About Atoll Documentation

Contacting Technical Support

1.2 Antenna Pattern Smoothing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

1.3 Power Received From Secondary Antennas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

1.4 Transmitter Radio Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

1.5 Repeaters and Remote Antennas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

1.6 Beamforming Smart Antenna Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

1.7 Grid-of-Beams Smart Antenna Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

1.8 Adaptive Beam Smart Antenna Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

1.9 Statistical Smart Antenna Gain Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

2 Radio Propagation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57

2.2 List of Default Propagation Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

2.3 Okumura-Hata and Cost-Hata Propagation Models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

2.3.3 Calculations in Atoll . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64

2.4 ITU 529-3 Propagation Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64

2.5 Standard Propagation Model (SPM). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .65

2.6 WLL Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75

2.7 ITU-R P.526-5 Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76

2.8 ITU-R P.370-7 Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76

2.9 Erceg-Greenstein (SUI) Propagation Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78

2.10 ITU-R P.1546-2 Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .80

2.11 Sakagami Extended Propagation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84

2.12 Free Space Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86

2.13 Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86

2.14 Shadow Fading Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .90

2.15 Path Loss Matrices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

2.16 File Formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

3 GSM GPRS EDGE Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125

3.2 Interference-based DL Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

3.3 GPRS/EDGE Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136

3.3.1.1 Calculations Based on C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

3.4 Codec Mode Selection and CQI Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145

3.5 UL Coverage Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152

3.5.3.4.2 Max C/I Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

3.6 Traffic Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

3.7 Network Dimensioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

3.8 Key Performance Indicators Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181

3.8.2.2.5 Blocking Probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183

3.9 Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184

3.10 Automatic Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191

3.11 AFP Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196

4 UMTS HSPA Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211

4.2 Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214

4.3 Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225

4.3.1.1.1 Circuit Switched Service (i) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226

4.4 UMTS HSPA Prediction Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284

4.4.3.6 HSUPA Prediction Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309

4.5 Automatic Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312

4.6 Primary Scrambling Code Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319

4.7 Automatic GSM-UMTS Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329

5 CDMA2000 Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337

5.2 Definitions and Formulas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340

5.3 Active Set Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358

5.4 Simulations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358

5.5 CDMA2000 Prediction Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392

5.6 Automatic Neighbour Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423

5.7 PN Offset Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430

5.7.1.3.2 Transmitter Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435