ICoIT17 Proceeding PDF

The 1st International Conference
on Information Technology
( ICoIT'17 )
Lebanese French University

Erbil, Kurdistan Region Iraq
10th April, 2017
lfu.edu.krd/conferences/icoit17
This Page Intentionally Left Blank
Preface
Once Steve Jobs said Lets go to invent tomorrow
The organizing committee is delighted to warmly welcome you to the

1st International Conference on Information Technology (ICoIT17) which will
be held in Erbil, by Department of Information Technology in Lebanese French
University.
This specific IT conference is the first of its kind that will be held in our
university and Erbil as well. The goal of this conference is to promote the
development of cutting edge research, invent development in IT and to develop
links between researchers, academicians and experts in Kurdistan and other
universities whether in Middle East or Western universities.
The conference committee received 112 research papers from 29 universities

(Kurdistan Region of Iraq, Iraq, Malaysia, Cyprus, Iran, Turkey, Syria, USA
and UK) in different IT fields.
Since the conference committee has put strict controls to meet quality
measurements and rejected a number of submitted research papers who have
exceeded the similarity index percentage and considering the reviewers
professional feedback, the number of accepted research papers were brought
down to 49.
The International Program Committee (peer reviewers) was made of 41

members from the academic institutions and universities all over the world,
from America, Europe and Asia continent, all of whom were reputable
professors of the field.
By the end of this process, the conference committee accepted 49 papers,

representing 43% of submitted papers. The 49 accepted papers will be
presented in a one day conference, with 12 regular sessions running in 5 parallel
halls. Besides that, 10 Keynote Speeches will be presented by a reputable
professors in 7 sessions and one special panel. This process will be proceeded
by the opening ceremony and sessions.
i
We are looking forward to hear what the technology gurus, researchers will
share the technology advancement and their impact on our society and daily
lives.
Finally, the organizing committee would like to express their gratitude to the
Conference Scientific Committee and International Program Committee.
Without their contributions, this conference would not have been possible.
Thank you very much for your time and presence
ICoIT17 Conference Chair

April 2017
ii
ICoIT17
ORGANIZING COMMITTEE
Conference Chair : Asst. Prof. Dr. Mazin S. Al-Hakeem
Head of IT Department - LFU
Logistic Committee Chair : Dr. Yahya Tareq Hussein

Department of IT LFU
Logistic Committee : Bnar Fareed Ibrahim
Logistic Committee : Schwan Dlawar M. Sabir
Director of Administration Directorate
Logistic Committee : Mohammed Hamadamin Huseen
LFU Presidency
Social Activities Committee Chair : Abdulsatar Abduljabbar

Department of Business Management
Social Activities Committee : Azwar Mehdi Salih
Department of Business Management
Social Activities Committee : Farah Qasim Alyousuf
Social Activities Committee : Anmar Abdulkareem Nassar
Office of President
Publishing & Media Committee Chair : Sarah Y. Yousif

Director of Planning Directorate
Publishing & Media Committee : Abdul Qaddir Abdul Ameer Hassan
LFU Presidency
Financial Committee : Firas N. Mardan

Department of Accounting - LFU
Financial Committee Chair : Mahamad T. Sleman
Directed of Accounting Directorate
IT Committee Chair : Mohammed Salim Abdulrahman
IT Committee : Mustafa Qays Ibrahim Al-Ameen
Computer Services Center
Immigration Affairs : Rekar Shams Aldeen Abdullah
LFU Presidency
Quality Assurance Affairs : Hawre Nuraddin Sabir
Director of Quality Assurance Directorate
Secretariat : Dina Khalid Khudher

Department of IT - LFU
iii
ICoIT17
SCIENTIFIC COMMITTEE
Prof. Dr. Alaa Hussein Al-hamami Prof. Dr. Tarik A. Rashid
Princess Sumaya University for Technology University of Kurdistan Hewler
Jordan Kurdistan Region Iraq
Prof. Dr. Ali Makki Saghaer Prof. Dr. Aree Ali Mohammed
University of Anbar University of Sulaimani
Iraq Kurdistan Region Iraq
Asst. Prof. Dr. Loay E. George Asst. Prof. Dr. Kamran Faraj
University of Baghdad Cihan University Sulaimani
Iraq Kurdistan Region Iraq
Asst. Prof. Dr. Mary Agoyi Asst. Prof. Dr. Zarei Behrouz
Cyprus International University University of Tehran
Republic of Northern Cyprus Iran
Dr. Sarina Muhamad Noor Dr. Halgurd Sarhang

Universiti Utara Malaysia University of Koya
Malaysia Kurdistan Region Iraq
Dr. Huda K Aljobori Dr. Selcuk Cankurt

Ahlia University University of Ishik
Kingdom of Bahrain Kurdistan Region Iraq
Dr. Akar Taher Dr. Yahya Tareq Hussein

University of Koya Lebanese French University
Kurdistan Region Iraq Kurdistan Region Iraq
Mr. Safwan Mawlood Mr. Mohammed Bajalan

University of Ishik Cihan University Erbil
Kurdistan Region Iraq Kurdistan Region Iraq
Mrs. Mardin Anwar

Salahaddin University
Kurdistan Region Iraq
iv
ICoIT17
INTERNATIONAL TPC
(TECHNICAL PROGRAMME COMMITTEE REVIEWERS)
Abduladhem Ali, PhD University of Basrah Iraq

Baghdad College for
Abdulkareem Ibadi, PhD Iraq
Economic Sciences University
Adeeb Hamdoon Sulaiman, PhD Applied Science University Kingdom of Bahrain
Ahmed Sabah AL-Jumaily, PhD Ahlia University Kingdom of Bahrain
Princess Sumaya University
Alaa H. Al-Hamami, Prof., PhD Jordan
for Technology
Alaa Kadhim, PhD University of Technology Iraq
Ali Makki, Prof., PhD University of Anbar Iraq
Ali Ahmed, PhD University of Basrah Iraq
Ali Kadhim, PhD University of Baghdad Iraq
Ali Kattan, PhD University of Ishik Kurdistan Region - Iraq
Aree Ali Mohammed, Prof., PhD University of Sulaimani Kurdistan Region - Iraq
Akar Taher, PhD University of Koya Kurdistan Region - Iraq
Ashwaq Hameed, PhD University of Technology Iraq
Bayan Sabbar, PhD AL-Nahrain University Iraq
Bashar M Issawe, PhD Al-Mustansiriya University Iraq
Emad H. Al-Hemiary AL-Nahrain University Iraq
Faraydon Karim, Prof., PhD University at Buffalo USA
Jamila Harbi Al'Ameri, Prof., PhD Al-Mustansiriya University Iraq
Halgurd Sarhang, PhD University of Koya Kurdistan Region - Iraq
Huda K Aljobori, PhD Ahlia University Kingdom of Bahrain
Kamran Faraj, PhD Cihan University Sulaimani Kurdistan Region - Iraq
Karim Hashim, PhD Al-Mustansiriya University Iraq
Kassim Al-Shamry, PhD University of Tikrit Iraq
Loay E. George, PhD University of Baghdad Iraq
Cyprus International
Mehmet Toycan, PhD Cyprus
University
Mohammad Al-Hamami, PhD Applied Science University Kingdom of Bahrain
Al Madinah International
Najeeb Alwan, PhD Malaysia
University
v
Nidhal El Abbadi, PhD University of Kufa Iraq
Osama Abbas Hussein, PhD Middle Technical University Iraq
Reem Jafar, PhD Cihan University Erbil Kurdistan Region - Iraq
Salah Yahya, Prof., PhD University of Koya Kurdistan Region - Iraq
Saman M. Abdullah University of Koya Kurdistan Region - Iraq
Sarina Muhamad Noor, PhD Universiti Utara Malaysia Malaysia
Suha Hadi, PhD University of Baghdad Iraq
International University of
Susan Al Naqshbandi, PhD Kurdistan Region - Iraq
Erbil
Soukaena Hassan, PhD University of Technology Iraq
Taleb A. S. Obaid, Prof., PhD University of Basrah Iraq
University of Kurdistan
Tara Botani, PhD Kurdistan Region - Iraq
Hewler
University of Kurdistan
Tarik A. Rashid, Prof., PhD Kurdistan Region - Iraq
Hewler
Wesam S Bhaya, Prof., PhD University of Babil Iraq
Zarei Behrouz, PhD University of Tehran Iran
vi
ICoIT17
KEYNOTE SPEAKERS
Prof. Dr. Alaa Hussein Al-hamami Prof. Dr. Ali M. Sagher
Princess Sumaya University for Technology University of Anbar
Jordan Iraq
Asst. Prof. Dr. Faiz Al-Alawy Asst. Prof. Dr. Ali Ahme
Kent State University University of Basrah
USA Iraq
Asst. Prof. Dr. Dujan Basheer Taha Asst. Prof. Dr. Emad H. Al-Hemiary
Mosul University Al-Nahrain University
Iraq Iraq
Asst. Prof. Dr. Ali Qatan Dr. Amer S. Elameer

Ishik University IIPS-ICCI
Kurdistan Region - Iraq Iraq
Dr. Kayhan Zrar Ghafoor Dr. Saman Mamyahia

Salahaddin University Koya University
Kurdistan Region - Iraq Kurdistan Region - Iraq
vii
This Page Intentionally Left Blank
viii
TABLE OF CONTENTS
5G: MIMO BDMA SYSTEM TRENDS AND TECHNOLOGY
Ashwaq Q. Hameed, Salman Goli and Zahraa Jaafar Oleiwi . 1
A Comparative Analysis of Texture Methods for Visual Object Categorization

Hayder Ayad, Loay E. Georgeand Mamoun Jasim Mohammed 16
A Developed Realistic Urban Road Traffic in Erbil City Using Bi-directionally Coupled
Simulations
Prof. Dr. Saad Talib Hasoon and Muhammed A. Mahdi ... 28
A Hybrid Neur-Fuzzy Optimization Method for PV Steady State Improvement in

Maximum Power Point Tracking Controller
Ashty M. Aaref and Ali M. Humada . 36
A Multi Secret Sharing Approach for Vulnerability Identification in Social Media

Dr. Saman Mirza Abdullah . 45
A PAPER SURVEY ON THE IMPACT OF (IT) ON THE QUALITY OF FINANCIAL

DATA (CASE STUDY OF MUNICIPALITIES AND TOURISM MINISTRY IN
KURDISTAN REGION IRAQ)
Firas N. Mardan and Farah Qasim Ahmed . 52
A PAPER SURVEY ON THE IMPLEMENTATION OF THE PARALLEL FDTD ON

MULTIPROCESSORS USING MPI
Adamu Abubakar Isah, Asst. Prof. Dr.Oyku Akaydin and Assoc. Prof. Dr.Mehmet Kusaf .. 60
A Proposed Public Key Encryption Based on Hybrid Chaotic Maps

Prof. Dr. Ahmed T. Sadeeq, Assist Prof. Dr. Alaa K. Farhan and Shaimaa A. Hassan . 77
A Proposed Non Feistel Block Cipher Algorithm

Ahssan Ahmed Mohammed nad Dr.Abdulkareem O. Ibadi ... 86
A Survey On Unsupervised Evaluation Criteria For Image Clustering Validation

Akar Taher .. 97
Adaptive Data Collection protocol for Extending Lifetime of Periodic Sensor Networks
Ali K. M. Al-Qurabat and Ali Kadhum Idrees ... 108
Alarm System using Microcontroller , GPS and GSM Devices

Nadia Mahmood Hussein and Saad Najim AlSaad 119
An Analytical Study for Some Drawbacks and Weakness Points of the AES Cipher
(Rijndael Algorithm)
Omar A. Dawood and Othman I. Hammadi ... 129
ix
An Overview of Software Defined Environment
Ali Dheyaa Jalil and Dr. Abdulkareem A. Kadhim 134
Building New Tool for Software Project Planning Based on Meta-modeling Concepts and
CMMI Meta-Standards
Husam Bofaoor ... 143
Computer Aided Diagnosis System for Breast Cancer using ID3 and SVM Based on
Slantlet Transform
Mohammed Salih Mahdi .... 157
Design a Complete Website for the Sustainable Development Goals (SDG) in Iraq
Amer S. Elameer, Methaq Khamis and Amjed Abbas ... 165
Design and Development of a MOOC for Academic Institution in Iraq

Inas Ismail Imran, Amer S. Elameer and Rafah Al-Hamdani 174
Design and Execute a Complete Vehicle System for the MOHEDSR

Mead Ali Moter, Rafah Al-Hamdani, Israa Shehab and Wasan Ala'a alhamami .. 187
Design and Implementation of a Rehabilitation Wireless Networked System

Asst. Prof. Dr. Mahmood Zaki Abdullah and Fadia Noori Hummadi ... 198
Distance Varieties Wireless Signals Relations Connected Peer to Peer Computers

Kamaran HamaAliFaraj, Banar Fareed Ibrahim and Azhi A. Salih .. 205
Distributed Data Aggregation protocol for improving lifetime of Wireless Sensor

Networks
Ali K. M. Al-Qurabat and Ali Kadhum Idrees .. 220
DSP for Mobile Wireless LTE System: A Review

Ashwaq Q. Hameed, Salman Goli and Mohammed K. Shoubith .. 232
Effect of Doppler Shift frequency on the performance of 2x2 OSTBC-OFDM System

Diana Hayder Hussein and Thuraya Mahmood Qaradaghi 247
Energy-Efficient Sensor Activity Scheduling Protocol for Wireless Sensor Networks

Wesam Huseein and Ali Kadhum Idrees 257
E-Service Quality Factors Impacting Customers Purchase Retention in E-Retailing in

Malaysia
Azwar Mehdi Qassim Agha 266
Establishing Educational Robotics Labs in Iraqi Universities

Dr. Ghada Saleem ... 284
x
Evolving Fuzzy Neural Network Equalization of Channel Impulse Response in Optical
Mode Division Multiplexing
Anahed M. Kareem and Assoc. Prof. Dr. Angela Amphawan ... 295
Fast and Accurate Real Time Pedestrian Detection Using Convolutional Neural Network
Hayder Albehadili, Laith Alzubaidi, Jabbar Rashed, Murtadha Al-Imam and Haider A.
Alwzwazy ... 305
Finding Shortest Path Basing on Dijkstra's Algorithm For Baghdad Universities by using
Geographic Information System Applications
Roaa Ebada Saeed, Amer Saleem Elameer and Rafah Shahab Alhamdani ... 317
Image Encryption Based on Partitioning and Rearranged Pixels Position

Dr. Jamila H. Al-Ameri, Salah T. Allawi and May M. Abbas . 326
Image Enhancement By Using Homomorphic Filtering Model

Asst.Prof. Amel H.Abbas and Prof. Dr. Jamila Harbi 335
Implementation of an Economic Light Duty Three-Axis Computer Numerical Control

Machine Ali A. Abed, Murtadha A. Al-Ibraheem and Auday A. Mohammed .. 343
Intelligent Fault Recovery Controller for Power Generator at Perdawd CCGS in

Kurdistan-Iraq
Wassan Adnan Hashim and Faiz Al Alawy ... 352
Internet of Things (IoT): Readme

Istabraq M. Al-Joboury and Emad H. Al-Hemiary 363
Internet of Things Based Cloud Smart Monitoring for Asthma Patient

Tabarek G. AL-Jaf and Emad H. Al-Hemiary 380
MARKOV-BASED DEPLOYMENT APPROACH TO IMPROVE WSN COVERAGE

Abd alnasir R. Finjan and Dr. Saad T. Hasson 386
NEURAL FUZZY PETRI NET BASED ARABIC PHONEME CLASSIFIER WITH

MFCC FEATURE EXTRACTION
Abduladhem Abdulkareem Ali and Ghassaq S. Mosa ... 396
New Weights in Laplacian Smoothing on Triangular Mesh

Kawa M. A. Manmi 411
Proposed Secret Authentication Method for E-government Purposes

Wasan Ala'a alhamami, Zanbaq H.Thenon, Amer S. Elameer and Soukayna Hassan .. 418
Security Issues in Social Media: Challenges and Solutions

Mardin A. Anwer, Rina Dinkha Zarro and Kamaran HamaAli Faraj 429
xi
SMS DISTRIBUTER BASED PATIENT APPOINTMENTS SYSTEM
Mohammad Salim Abdulrahman, Rebin B. Koshnaw and Dr. Mazin S. Al-Hakeem ... 442
Survey and New Security methodology of Routing Protocol in AD-Hoc Network

Shahab Wahhab Kareem and Dr.Yahya Tareq Hussein 452
The Estimation of Wind Velocity Using Data Mining Techniques

Sattar Nabee Rasool, Ahmet Koca and Karwan Hussein Qader 465
Unbreakable Digital Watermarking Technique

Abdulkareem O. Ibadi 477
Using RSA Digital Signature to Solve Pollution Problem in Single-Source Network

Coding
Rami S. Youail 487
Virtualizing Network Services: Practical Perspectives

Emad H. Al-Hemiary .. 492
, E. Banking
........................................................ . .
( )

...........................................................................................................
xii
The 1stInternational Conference on Information Technology
(ICoIT'17)
Lebanese French University - Erbil, Kurdistan Region Iraq
10thof April, 2017.
5G: MIMO BDMA SYSTEM TRENDS AND TECHNOLOGY

Ashwaq Q. Hameed
Electrical Engineering Dept.
University of technology- Iraq
50058@uotechnology.edu.iq
Salman Goli
Computer Engineering Dept.
University ofKashan- Iran
salmangoli@kashanu.ac.ir
Zahraa Jaafar Oleiwi
Communication Engineering Dept.,
University of Kashan- Iran
zahra.j3far@gmail.com
ARTICLE INFO ABSTRACT

Beam Division Multiple access (BDMA) is the last technique
Article History: came to create the fifth-generation (5G) and solving the
Received: 17 March 2017 problem in the fourth generation (4G). In which base station
(BS) allocates private beam to the mobile station (MS) in the
Accepted: 1 April 2017
system, to increase the capacity and bandwidth of system. For
Published: 10 April 2017 that antenna beam is divided according to the speed and
location of mobile stations. In mobile wireless
Keywords: Beam communication the important thing is how to increased
division multiple access limited frequency of the system, improve data rate and good
(BDMA), Next quality of services for user etc. This paper presents
Generation, BDMA for information about BDMA and explains the different
5G technologies, and which we want to making future mobile
with more powerful and more in demand. Useful thing in
BDMA technique is the large number of users to connect at
the same time by direction of orthogonal beam from BS by
giving multiple access to MS, and increasing the ability of
system to be established very big channel. Direction of
antenna beam goes off the desired location of MS, and void
are placed in undesired direction depend on work
environment.
1. INTRODUCTION.
T he various multiple access techniques is growing at a rapid pace for multiple user to
usage efficient like is Time Division Multiple Access (TDMA), Frequency Division
Multiple Access (FDMA)... Etc. In Telecommunication system, the limitation oftime
1
(ICoIT'17)
10thof April, 2017.
and frequency are divided to be used by subscribers (user) also capacity is limited because
depend on giving frequency and time. Now we demand a development technique utilize other
resources to increase the ability of system. So the thought of Koreans in seeking for finding
anew access technique Known as BDMA for next generation (fifth generation) of mobile
wireless communication system. Short term of BDMA is Beam Division Multiple
Access.Next generation utilize (BDMA) that enable to transmit data rate greater than
100Mbps for one square kilometer to million Internetof things. 5G includes advantages such
it is the strongest and high demanding in the near future. Founding 5G technology have
come to existence high resolution for moving mobile users, higher data rates and Quality of
Service (QoS), high bandwidth. Big challenge in the telecommunication field will be to
how increase capacity of system and quality of service within the limited frequency
spectrum, in 5g will be frequency band 3-300GHz and the Bandwidth of 1 Giga bit per
second (Gbps) or highest from that[1]. Mobile wireless communication system based on
BDMA has to support better quality service to a lots of subscribers. Fifth generation (5G)
telecommunications network provided typical solutions for thecommon problems in the
fourth generation (4G). Add to that we can increase the capacity and reliability, in this
technology also we can reduce bit error rate [2].
2. Beam Division Multiple Access and Fifth Generation.
The main aim of (mobile network wireless communication system) has to availability high
quality of service to a big number of user via different beams by select user without
interference of beams with others and affordable cost. Beam Division Multiple Access
(BDMA) is foreseen as the latest multiple access technology for next or fifth generation (5G)
of wireless communication [3]. What meant by the beam division? It is dividing of beam
according to environment will towards to desired direction and nulls are placed in undesired
direction. In BDMA available beam can be divided in to 3D and improve the capacity of
system as shown in figure 1. Multiple accesses to channel depend on multiple access
technology. A channel represents a resources of data that enable the user (mobile station) to
connect with other user in network. Many user can be share a large data from their resources
by using a finite spectrum in multiple access. Multiple access is desirable to allow the
subscriber to send information at the same time from mobile station (MS) to base station
(BS) while receiving information from BS, in wireless communication system.
2
(ICoIT'17)
10thof April, 2017.
Figure 1: Beam Division Multiple Access [4]
From (1G) to the(2.5G) and from (3G) to (5G) Communications system improved
dramatically with a significant improvement in the transfer and receipt of data. Because of
the huge volume of data and sources and increasing numbers of subscribers, it has become
necessary to develop communication system and meet the real requirements of users. 5G (5th
generation mobile networks or 5th generation wireless systems) Designed for the future of
mobile telecommunications standards, and it is supposed to outperform significantly on the
current 4G standards. Due to the frequency band of 5G will be 3-300GHz and the bandwidth
more than 1Gbps, so that will become more capacity and quality of service than previous
generations[5]. 5G provides many features such as photo gallery, messenger, and multimedia
applications, telephony, camera, etc.Add to high speed and capacity, and low in cost (per
bit). All this features supports, video streaming, interactive Internet, and other broadband
services. This technology has a great future because it can handle best technologies such as
BDMA. World market willacquires by the 5G technology [5].
The major differences, between present generations and expected 5G techniques as shown
in figure 2 are:
a. High data rates up to 1Gbps.
b. A large range of signal coverage.
c. Reduce the battery consumption.
d. Different and multiple paths for data transmission at the same time.
e. Worldwide wireless web (wwww).
f. More security.
g. Supposed to have no effect on human health.
3
(ICoIT'17)
10thof April, 2017.
Figure 2: Feature of 5G Technology [6]
3. Analysis Methods of 5G technology.
3.1 5G millimeter wave technology.

An increasing need to larger bandwidth arises because the huge growth of large number of
user of mobile. As it is well known the bandwidth is limited in available frequency spectrum,
and which is less the (mm) wave band. So it was discovered wave band as mobile frequency
spectrum by operators to be able to support larger bandwidth. But the problem remains in
the noise, penetrate walls, some objects in buildings, and rain. After taking into consideration
all these factors mm wave can will become stronger and betteras shown in figure 3. Note that
these waves are called 5G millimeter wave [7].
Advantages of 5G millimeter wave:

More number of subscribers can be accommodated due to provides larger bandwidth
More favorable for smaller cell due to less bandwidth in millimeter range.
Coverage is not limited to line of sight (LOS).
4
(ICoIT'17)
10thof April, 2017.
Figure 3: 5G millimeter wave frame and sub frame [7]
Disadvantages of 5G mm wave:
Millimeter wave goes through different losses such as penetration, rain attenuation etc.
Moreover path loss at mm is proportional to square of the frequency.
Loss in mm wave frequencies due to Foliage.
Supports only LOS (Line of Sight) propagation. Hence coverage is limited to LOS.
Loss of energy because of higher at millimeter wave due to more number of radio
frequency modules due to large number of antennas [7].
3.2Mobile Ad-hoc Networks (MANETs).
Mobile Ad Hoc Network (MANET) It is a connecting multiple nodes or terminals with some
in a way wireless communications without the service of any controlling or microcontroller.
These nodes can be able to communicate and transfer information away from the use of any
existing fixed network infrastructure [8]. Mobile computing in the field of computers and
communications it has become a high-level Due to the rapid development of Computing.
(MANET) it is a completely wireless connection, and it's done by nodes at each calculator
or Mobile, and usually take the Dynamic form with a limited bandwidth. The most prominent
challenges in Mobile ad hoc networks is the process of guiding thedata packets from end to
end (source to destination), Where it will be difficult in mobile ad hoc networks because of
the continued movement and the absence of centralized control over all nodes (Mobile
Station) [9] as shown in figure 5. Types of MANET as shown in figure 4:
1. VANETs. Communication between vehicles, also between vehicles and equipment on the
Roadsides can be use VANETs.
2. Mobile nodes and stable Internet-nodes can be linked by Internet Based Mobile Ad-hoc
Networks (I MANET)[10].
5
(ICoIT'17)
10thof April, 2017.
Figure 4: type of networks normal ad hoc [10].
WIFI Access Cellular

Point Backbone Tower
Wireless Ad-Hoc Network
Figure 5: Ad-hoc network [7].
4. BEAM DIVISION MULTIPLE ACCESS technology
The main goal of mobile communication systems is to provide good services to a larger
number of mobile users at lower costs. To provide good services, three are major multiple
access techniques FDMA, TDMA and CDMA used to share the available bandwidth in a
6
(ICoIT'17)
10thof April, 2017.
wireless communication system as shown in figure 6. This techniques can be illustrated as

follows:
1. Frequency Division Multiple Access (FDMA) Each user have a dedicated allocation of
one or several frequency bands in this technique.
2. Time Division Multiple Access (TDMA) - Number of users can share the same frequency
channel by partition the signal into slots of time in this technique.
3. The codes that provides the mobile stations can be separates calls by using code Division
Multiple Access (CDMA) technique.
4. Enhance resource of system frequency can be done through division and allocation by
using Orthogonal Frequency Division Multiple Access (OFDMA).
In future, due to the increasing in number of user (MS) and the volume of data to be
transferred, the capacity of (wireless network communication system) required to increase
[11].
Figure 6: Multiple Access Technology [11]
5. Concept of BDMA Technology:
Due to the increasing subscribers' request to need a new technology which can meet their
needs, such that it should improve the channel efficiency, data rate and provide quality in
service. MIMO is one technology to develop solutions for this situation, it is able to provide
the required performance using multiple transmitter and receiver antennas [10]. BDMA is
7
(ICoIT'17)
10thof April, 2017.
the last technique in which an orthogonal beam is allocated to each mobile station. this
orthogonal beam is allocated to every mobile station (MS) need to communication with base
station (BS), The BDMA technique depend on time division multiple access with slow
frequency hopping (SFH-TDMA) and orthogonal frequency division multiplex
(OFDM)[10]. division of beam can be done through BS based on locations and number of
MS, thereby significantly increases the capacity of the system since BS can be transmit single
or multiple beam at the same time and without interference to provide multiple access of
system [11]. When MS and a BS can be in Line of Sight an (LOS) state BS will transmit
single beam, But When MS are positioned at different angles compared to the BS, and the
BS will transmit beams at different angles to transmit data at the same time. BDMA uses
phased antenna array, beam forming technique to produce dedicated beam and uses multiple
beam for multiple access, in a new space division multiple access technique. When mobile
stations are present at different positions or angles, each one of them are provided with a
beam separately. If the mobile stations are in same location then they use a single beam, by
using the available frequency - time for multiple accessesas shown in figure 7. The change
in position, speed or direction of mobile station means change the number of beams, width
and direction according to movements of mobile station [12].
Figure 7: BDMA Technology [12]
The steps show us to set of connectionas shown in figure 8:
A. In the beginning, the positions of BS and MS unknown to both; the mobile stations must
be discovering its location and moving speeds and the location of base station.
B. The mobile station transmit there moving speed and position to base station, and then BS
determines the width and direction of downlink beam (DL).
8
(ICoIT'17)
10thof April, 2017.
C. Once, the direction and width of the DL is computed by mobile station, and transmits the
uplink beam (UL) in the same direction.
D. Base station receiving the UL, after that the (uplink and downlink beam) should be
updating to start of carrying data [12].
Figure 8: Step for set of connection between BS and MS[12]

6. MIMO-BDMA SYSTEM MODEL
Beam forming is a method used to configure the radiation from the antennas to the desired
shape and the desired direction (mobile phone) Directed radiation empty for non-desirable
trend. It used for direct data transfer and receipt of the signal between the base station and
mobile station. There are basically two approaches:
1. Switched beam antenna: Generate a limited number of radiation patterns, which makes the
RF signal and one is available each beam is possible. It has been compiled using the patterns
of wireless network the simplest approach beam Switched compared with adaptive approach.
This it provides a significant increase in Network capacity compared with the Conventional
multi-directional antenna Systems or existing systems sector.
2. Adaptive Array Antenna: an infinite number of radiation patterns of use of adaptive air
Matrix is one of the areas that promise further improve Wireless systems and provide
improved capability Safety by the capabilities of the site's position.
This is Matrices can be used to reject any interference as shown in figure 9by:
9
(ICoIT'17)
10thof April, 2017.
* altering spatial position of the site through the trend.

* ending measurements and the development and improvement.
* Channel models through the corner of Canal and arrival.
* Sounding measurements.
Figure 9:MIMO-BDMA SYSTEM MODEL[13]

Tracking of mobile phone by this system can be done by constantly directing the (main beam)
toward the user, and the formation of null values in the orientations Interference signal. Ray
turned like Systems, it also includes arrays. Typically, the received signal from each of the
distributed elements of antenna it is multiplied by its weight. These signals it is combined to
produce the output range. This complex is calculated by weights Complex adaptive
algorithm, which is Pre-programmed digital signal processing the unit that manages the
signal Emitted from the BS[3].
In this paper, we assume a channel model, which describes the spread between transmit array
and receive array. Suppose that there are P physical paths between the (BS) and (MS), and
the p1th path of the u user has an attenuation of (^(p,u)) an angle of (^(p,u)) with the
transmit , and an angle of (^(p,u))with the Receive .Then N M physical MIMO channel
matrix in the downlink linked with the pth path of the uth user is given by equationbelow:
,
[ ] =, ((,) /)* (, )
(
,
) (1)
Where d means downlink, , is the physical distance between transmit and receive of
antenna along path p, and is the r carrier wavelength. Added to that, () CN1 which
obtain ||()||2= 1is the subscribeof the (antenna array)., and et() CM1 which obtain
et()_2 = 1 is the base station (BS) of the antenna array. The channel frequency is the
equation below:
[, ] =1
=1
,
exp (2 (,) /)* (, ) ( , ) exp(2 (,) )(2)
Where (,) is the (Spread delay)linked with the (pth path). We suppose that the uplink
physical channel has the same parameters ( , , , , (,) ) as the downlink channel except
for the carrier frequency and the initial phases Hence,uplink channel can be calculate[14] .
10
(ICoIT'17)
10thof April, 2017.
If we assume that users are quite far away from the BS such that the phases 2d ^ (p,u)/r
are uniformly distributed on [0, 2] and due to the uncorrelated scattering, it become
mutually independent. Now, we have:
{, exp(2(,) /)} =0(3)

2 (,)
{, exp(2(,) /)(, exp ( )) } = , (p-p', u-u') (4)

This assumption depend on the BS antennas are located much higher rooftops this means a
lack of distractions in BS. Where ^(p,u) = E {|o^(p,u)|2} is the channel gain of path p [15].
lim () ()= (- ) (5)
For each user, when the sampling of , satisfies the condition
( (, ) (, )) = (N-N ')(6)
We can rewrite the channel matrix in (2) as
, = 1 2 , ,
=1 1=1[ ]N,b1 ( ) ( 1 ) = Wu,
(7)
WhereV=[mt(1), mt(2), . . . , mt(M)] CMM, andWuk=[mr(1,u),mr(2,u), . . . ,

er(N,k)] CNN is aidentity matrix. Note that, as the number of BS antennas b2tends to
infinity,Vbecomes an asymptotically identity matrix. We call [,
]as the beam domain
channel [14], we have the following equation:
[,
]N, b1 (,)(,1)
,
exp(2(,) /) exp(2 (,) )(8)
Where (,)is the receiving angles for set of all paths, and (,1) is the transmit angle for
set of paths.

So,number of columns in channel matrix , is [15]:
2 (,)
Vec(, )= 1
=1 1=1 1{ [, ]nm () ( ) (9)

And calculate the full correlation matrix of,

Vec(, )= =1 =1 { [, ]nm[, ] * ( () r(n,k))( () r(n,k))(10)
We can define the Eigen mode channel coupling matrices (CCM) as [16]:

, ={ , (, ) } (11)
11
(ICoIT'17)
10thof April, 2017.

We find that the Eigen mode CCM , is likewise independent of sub-carriers. According
to [17], the (UL) and (DL) statistical channelstate information (CSI) are usually reverse in
both time-division duplex (TDD) and frequency-division duplex (FDD) systems. Thus, we
can indicate the (UL) and (DL) CCMs as:
1
, = (, ) = (12)
Whose elements can be approximated by [u]Nb1 (,)(,1) , Note that the

superscript u1denotes (UL).The received signal at the uth user in the th OFDM subcarrier
can be rewritten in the form

, = , , + , (13)

Where , CM1 is the signal meant for the uth. The received signal of thuser in the th
sub-carrier can be as follows:

, = , , + =1, , , + u,(14)
The 5G gave us through the use of technology BDMA a lot of advantages, the important one
is the high speed in data transforming as compared with previous generation and the smart
use of send data instead of Wi-Fi technology and losing in power and band, and this leads to
increase in capacity and the number of user who can connect with the system, and keep the
bandwidth from dispersion by allocating beam (packets of data) directly to the subscribers
this allows us to send more data size whether video, audio, etc., this is the most important
part of the idea of BDMA [18]. It should also be noted that BDMA reduce latency
significantly compared to the fourth generation (4G) as well as previous generations. This
make a revolution in communication and data transform. Add to that BDMA solve problem
between cell overlapping and solve performance deterioration problems of subscriber at cell
that occurring in a cellular system. Taking into consideration radiation efficiency of the
antennas can be maximized when radiation pattern of the base station and radiation pattern
of the mobile station are designed to match each other [19], as for the disadvantage of
BDMAcan be summarized as follows:
1- Does not support older devices.

2- At the moment the fifth generation technique has not been applied on the ground.
3- BDMA It needs to be qualified engineers for the purpose of operating and maintenance.
7. Application
The idea of BDMA designed to the next generation in (mobile communication). In

addition to 5G mobile communication network will be faster than 3G and 4G, it also has the
efficiency and ability to revolutionize in many directions such as health care, automotive,
production and energy. It also will make it easier transportation to and from wired - wireless
connections. In this case installation costs will be reduced and flexibility will be increase
[20].
12
(ICoIT'17)
10thof April, 2017.
A. E-health: it can take advantage of this application to calculate and monitor the
patient's blood pressure as well as to know the pulse rate. His breathing by wearing
the devices assigned to the piece. This application enables the patient to reduce costs
by allowing patients to stay out and away from the care facilities [21].
B. Transportation routes: the usefulness of this application to participate in the car with
other cars with information on what comes in ways greater efficiency and a very large
safety.
C. Smart train can be more pleasant as travelers travelling on higher speed trains through
the vast areas and uninhabited, full 5Gnetwork can work.
D. at densely packed crowd, at an event of sport for example, enhanced the speed and
capacity can be Increases entertainment due to the Increases of the virtual reality
close to the truth.
E. It is possible to get a revolution in Industry and manufacturing due to the Fifth-

generation technologies. Where the robots of factories will depend on 5G in many
aspects of the work, this will increase production and reduce the cost[22].
F. In CGCS (complementary ground component) it is ought to reuse frequencies

between (satellite) and (CGC) components under allowable interference level [23].
8- CONCLUSION
In this paper we have surveyed 5G wireless technology phones. Like the 5G BDMA-based
technology that provides the best quality of service. And at the lowest cost to serve a greater
number of connections to the number of users without quality within the limited spectrum
available loss with increased system capacity. There are a lot of enchantment from 1G to 5G
wireless technology. BDMA is the latest technology used in the next generation. Some new
technologies has been added to wireless networks 5G, This makes the 5G wireless network
is more high-end than the 4G network, and more technical. It 5G will cover the world and
offers seamless service as its basic goal Like the 5G BDMA-based technology that provides
the best quality of service. And at the lowest cost to serve a greater number of connections
to the number of users without quality within the limited spectrum available loss with
increased system capacity compare with the. Also 5G will achieve its goals through the
efficient use of spectrum, dynamic allocation of broadband, transmission technology
upgrades and other aspects. Where we can say that a full update for 3G and 4G. Can 5G use
broadband more efficient to improve the transmission speed and the flow of information
within the time unit and have a huge Development on cloud computing, e-commerce, and
the wisdom of life related industries The use of (BDMA technology) in the 5G can improve
the capacity of system, and thus can increases the coverage of service on larger areas and
throughput of system. Compere with 5G millimeter wave technology that supports only LOS
(Line of Sight) propagation. Hence coverage is limited to LOS and loss of energy because of
higher at millimeter wave due to more number of radio frequency modules due to large
number of antennas. In BDMA technique BS transmit different beams with different
directions to MS without interference. Compared with Mobile Ad-hoc Networks (MANETs)
13
(ICoIT'17)
10thof April, 2017.
is the process of guiding the data packets from end to end, where it will be difficult in mobile
ad hoc networks because of the continued movement and the absence of centralized control
over all nodes (Mobile Station).
Reference
[1]Manjurul H. Khan, Barman, "5G- FUTURE GENERATION TECHNOLOGIES OF

WIRELESS COMMUNICATION REVOLUTION 2020, American Journal of
Engineering Research (AJER) e-ISSN: Volume-4, Issue-5, pp-206-215, 2015
[2] Rohini C G, Satyanarayana, " Analysis of Beam Division Multiple Access in Massive
MIMO Communication Systems", National Level PG Project Symposium On Electronics &
Communication, Computer Science, 2016
[3]Hamsalekha. P. S, Puneeth, "Implementation of Adaptive Beam Forming Algorithm for

Smart Antenna", International Journal of Engineering Research & Technology (IJERT) ,
Vol. 2 Issue 9, September 2013
[4] PriyaGoyal, Avtar Singh Buttar, "A Study on 5G Evolution and Revolution International
Journal of Computer Networks and Applications", Volume 2, Issue 2, March April (2015)
[5] Saddam Hossain, "5G Wireless Communication Systems", American Journal of

Engineering Research (AJER), Volume-02, Issue-10, pp-344-353, 2013
[6] Prof. Alberto Bartoli, Prof. FulvioBabich, "Beamforming Techniques forWireless

Communications in Low-Rank Channels: Analytical Models and Synthesis Algorithms",
UNIVERSITA DEGLI, 2008
[7]T. S. Rappaport, " Millimeter Wave Mobile Communications for 5G", Digital Object
Identifier, 2013
[8]Chlamtac, M. Conti, J.J.-N. Liu, Mobile Ad Hoc Networking: Imperatives and

Challenges,, vol. 1, no. 1, pp. 1364, Jul-2003
[9] Dr. SNS Rajalakshmi, R. Marutha Veni1, "Mobile Ad hoc Network" International Journal
of Science and Research (IJSR), India Online, 2013
[10]Belding-Royer,E.M. and C.K. Toh,1999. A review ofcurrent routing protocols for ad-
hoc mobile wirelessnetworks. IEEE Personal Communication magazinepp: 46-55.
[11] Imthiyaz Ali, "BEAM DIVISION MULTIPLE ACCESS (BDMA) FOR 5G",
International Journal of Science and Research (IJSR), 2011
[12] PriyaGoyal, "A Study on 5G Evolution and Revolution", International Journal of

Computer Networks and Applications, Volume 2, Issue 2, 2015
14
(ICoIT'17)
10thof April, 2017.
[13] Dong-ho cho, "Beam Division multiple system and method for mobile communication
system, Espacenet, 2010
[14] D. Tse, P. Viswanath, "Fundamentals of Wireless Communication" New York, NY,

USA: Cambridge Univ. Press, 2005.
[15] M. Viberg, B. Ottersten, and A. Nehorai, Performance analysis of direction finding

with large arrays and finite data IEEE Trans. Signal Process., vol. 43, no. 2, pp. 469477,
Feb. 1995.
[16] K. Higuchi, Y. Kishiyama, Non-orthogonal access withsuccessive interference

cancellation for future radio access,9th IEEE Vehicular Technology Society Asia
PacificWireless. Communication Symposium (IEEE VTS APWCS2012), Kyoto,
Japan.2012.
[17] HeeWook Kim, Unhee Park, Kunseok Kang, Bon-Jun Ku, "Satellite Frequency Reuse
Method for Complementary Ground Components in an Integrated MSS System", ETRI
Daejon, Korea 305-350, 2013
[18] A. M. Sayeed, Deconstructing multi-antenna fading channels, IEEE Trans. Signal

Process. vol. 50, no. 10, pp. 25632579, Oct. 2002
[19] G. Barriac and U. Madhow, Space-time precoding for mean and convariance feedback:
Application to wideband OFDM, IEEE Trans. Commun., vol. 54, no. 1, pp. 96107, Jan.
2006.
[20] HeeWook Kim, Unhee Park, Kunseok Kang, Bon-Jun Ku, "Satellite Frequency Reuse
Method for Complementary Ground Components in an Integrated MSS System", ETRI
Daejon, Korea 305-350, 2013
[21] X. Li, S. Jin, and X. Q. Gao, Multi-user MIMO downlink eigen-mode transmission
over jointly correlated MIMO channels, Sci. China Inf. Sci., vol. 54, no. 10, pp. 21242137,
Oct. 2011.
[22] "ANALYSIS Understanding 5G: Perspectives on future technological advancements in

mobile", GSMA Intelligence, 2014
[23] Ron Davies, "5G network technology'', EPRS | European Parliamentary Research
Service, 2014
15
(ICoIT'17)
10thof April, 2017.
A Comparative Analysis of Texture Methods for Visual

Object Categorization
Hayder Ayad
Department of Computer Science, University of Baghdad, Iraq
Hayder.alsultany@gmail.com.
Loay E. George
Department of Computer Science, University of Baghdad, Iraq
loayedwar57@yahoo.com
Mamoun Jasim Mohammed
Department of Computer Engineering, The Iraqia University, Iraq
mamoun_87@yahoo.com

This paper presents a comparative study to the most common
Article History: texture features analysis methods. In fact, there are two kinds
Received: 20 March 2017 of approaches have been proposed to extract the texture
features for the purpose of object categorization, the former
deals with the intensity pixels which derived the intensities
Published: 10 April 2017 texture and the second method dealing with the edge pixels
which obtained the edge texture. However, to extract and
Keywords: Texture make a comparative analysis to the texture maps there are
analysis, Co-occurrence several approaches have been presented in this research, i.e.,
Matrix, Gradient, Edge Gradient, Co-occurrence matrix, Contrast and Edge
Histogram, SVM, Histogram Descriptor. A real world images dataset denoted
Caltech 101 Dataset. by Caltech 101 dataset has been adopted to evaluate the
proposed texture analysis methods. Mostly, the first 20
classes with 40 images per-class have been chosen to
demonstrate the methods performance. Fundamentally, the
objects of these images have almost isolated for the purpose
of categorization. The experiment results show that the edge
histogram descriptor outperformed the other proposed texture
analysis methods with average accuracy 71.1751.355775
because the edge histogram descriptor is less sensitive to the
noise and variation of pixels intensities which most of the
objects of Caltech 101 dataset profoundly affected.
2. INTRODUCTION
T
he fast growth in the Internet and the available image devices, such as digital
cameras and image scanners has led to a tremendous increase in the management
and organization demands of such digital libraries. Therefore, several kinds of
research have involved and introduced an application for the purpose of object
categorization. In fact, Visual Object Categorization (VOC) is considered one of the
16
(ICoIT'17)
10thof April, 2017.
most challenging applications; hence it took the researchers' attention. The most
common issue facing this application is the semantic gap, the differences between the
low-level features and the richness of human mind concept [3, 7].
Several types of research have introduced various solutions that involve the use of object
ontology or low-level features associated with the machine learning algorithms. With
respect to the low-level features, the researchers introduced three kinds of features:
color, texture, and shape. Mainly, the texture features have been widely utilized and
viewed as a fundamental solution for various applications, such as: remote sensing,
facial recognition, and object categorization [8].
However, the texture is seen as repeated patterns of information with regular intervals.
It also indicates the surface properties and object appearance. The review of the
literature shows that the there are four approaches that have introduced the extract the
texture features. These approaches include, namely, the Structural, Statistical, Model,
and Transform based features extraction [4].
In fact, the efficient methods for texture features extraction rely on the excellent analysis
to the local features. Mostly, some of the texture features extraction methods are based
on the pixel intensities while the other are based on the edge pixels. The Gray level Co-
occurrence Matrix (GLCM) extracts the texture features based on the relationship of the
pixels. There are three ways to extract the texture features of the GLCM method. These
ways are based on pixels neighboring relationship; i.e., first-order statistics, second-
order, and high-order statistics [5, 1].
Besides, two main factors were commonly used measure the texture information: (i) the
roughness and (ii) smoothness of the object surface. Therefore, numerous statistical
equations have been offered to measure the degree of roughness and smoothness of the
object surface. In addition, there are other factors which can add valuable information
to the texture features to measure the brightness and darkness of the texture [10]. In
contrast, to obtain the edge texture features, diverse methods were presented; as a case
in point is the Edge Histogram Descriptor (EHD) of MPEG-7 descriptors. Obviously,
this descriptor has various characteristics which made it well-known and broadly used
by several applications, such as: Content-Based Image Retrieval (CBIR) and Visual
Object Categorization (VOC) [11; 2].
In this paper work, a valuable comparative analysis between the intensity and edge based
texture features had been carried out. For the intensity-based texture features, two
methods are proposed; namely, Gradient, and Co-occurrence Matrix. Besides, for the
edge-based texture features, EHD is adopted to extract the texture features. Test results
indicated that the first 20 categories of Caltech 101 dataset, which was chosen, to
demonstrate the proposed methods performance.
17
(ICoIT'17)
10thof April, 2017.
3. TEXTURE MAP CONSTRUCTION

The texture is a complicated set of patterns, which provides valuable visual information.
However, there are several texture characteristics that have been presented in the literature
to show how the texture of the current regions behaves. Such characteristics include
roughness, smoothness, regularity, density, coarseness, randomness, fineness, uniformity,
etc. [9, 10]. Therefore, in this research paper, different methods are presented to extract the
texture maps. Besides, several mathematical equations that measure the texture features were
manifested in this research. The sections below illustrate the texture maps extraction methods
and the mathematical equations for its analysis.
2.1. Co-occurrence Matrix:
A co-occurrence matrix is considered the second order statistical method that is widely
used in the last two decades to extract the texture features. In fact, this approach takes into
consideration the relationship between the neighboring pixels to generate the Co-occurrence
matrix. To construct this matrix, the occurrence of the neighboring pixels in different
directions (ex: vertical, diagonal (45 and 135) and horizontal) are counted [1, 9]. The order
of the co-occurrence matrix depends on the relationship between the current pixel and its
next neighboring pixels. However, in this research, the relationship between the first and
third neighboring pixels is counted to generate the co-occurrence matrix. The main idea
behind using the first and third neighboring pixels is, in some cases, due to the fact that the
distribution of the gray level is very high. That is; if the first neighboring relationship did not
succeed in capturing the probability distribution of the gray level, the third neighbor pixel
will succeed. To calculate the co-occurrence matrix, the distance and the theta should be pre-
defined (, ). After calculating the co-occurrence matrix, different statistical equations
have been proposed to analyze the texture features. The following figure presents an example
of the calculation of the co-occurrence matrix [7].
j
1 1 0 0
0 1 2
1 1 0 0
0 #(0,0) #(0,1) #(0,2)
0 0 2 2
i 1 #(1,0) #(1,1) #(1,2)
0 0 2 2
2 #(2,0) #(2,1) #(2,2)
(a)
(b)
8 2 0 8 2 2 4 2 1
2 4 0 2 4 0 2 2 0
0 0 4 2 0 4 1 0 2
(1,0 , 180 )

(1,90 , 270 )

(1,45 , 225 )
(c) (d) (e)
4 2 1
2 2 1
1 1 2

(1, 135 , 315 )
(f)
18
(ICoIT'17)
10thof April, 2017.
FIGURE 1: An example of the Co-occurrence Matrix, (a) (44) Sub-image, (b) A

General Form of Co-occurrence Matrix, (c), (d), (e), (f) are co-occurrence matrix results.
In fact, before calculating the co-occurrence matrix, the colors of the image are quantized
into N levels. The aim of performing the quantization step is to ignore the redundant
information and keep the adequate information as the human vision does. Besides, the co-
occurrence matrix will be calculated for each image band (Red, Green, Blue, and Gray). The
purpose of extracting the co-occurrence features from these bands is to capture every possible
information from the entire image. The following equations are applied to analyze the texture
features of the co-occurrence matrix [9].
1- Mean
(,)
= (1)

2- Standard Deviation
1 2
=
=1 ((, ) ) (2)
3- Skewness
1 3
((,))
=1
= (3)
3
4- Angular Second Moment or Energy
2
= (, ) (4)
5- Contrast
= ( )2 (, ) (5)
6- Correlation
{ (,)(,)}
= (6)

Where the are the means values and the are the standard deviation.
7- Entropy
= (, )(, ) (7)
8- Inverse Difference Moment
(,)
= 1+2() (8)
9- Variance
= ( )2 (, ) (9)
; (, )
10- Sum Average
= 2 =2 + () (10)
+ () = ,;+= (, ).
11- Sum Entropy
= 2 =2 + () log(+ ()) (11)
19
(ICoIT'17)
10thof April, 2017.
12- Homogeneity
(,)
= 1+|| (12)
13- Difference Entropy
= 1
=0 () log( ()) (13)
This set of mathematical equations has been used to measure the various characteristics
of the texture maps. These characteristics involve the following: homogeneity, regularity,
complexity, dependency, uniformity, correlations and others, which were constructed using
the co-occurrence matrix. In fact, in this research, two levels of relationships between the
neighboring pixels have used the first and third neighboring pixel. Therefore, the number of
features for each band will be 26 features, and the total number of features will be 104
features extracted from each image.
2.2. Gradient Texture Analysis: A Gradient is another texture map method, which uses the
edge density based on the pixel level contrast. In the literature, several methods have been
introduced to extract the edge map. Such methods can be later employed for different
purposes, such as features extraction, object segmentation, and texture analysis. In fact, the
edge is essentially based on the intensity change of an image. Moreover, good edges are
imperative to higher level processing. The edge detector is paramount in the area of computer
vision. Certainly, different methods were proposed to extract strong edges, including the first
and second order derivatives [6]. Simply, to construct the edge map, the relationship between
the neighboring pixels should be taken into account. These edges will be used later as texture
maps to be analyzed [8]. However, to extract the edges, two methods were used: the first and
the second order derivative.
In this research, the second order derivative is used to construct the texture maps. Similar
to the co-occurrence matrix, the colors of the images are quantized from 256 levels into N-
level. This process will improve the processing time and act as the human vision does.
Additionally, the texture maps will be constructed based on two areas, namely darkness and
brightness. The main purpose of using this principle is to measure the texture into two
different parts of the image. This can in turn give an indication of how the texture is
distributed in the image. The following equation illustrates the Gradient of the second order:
2
= ( + 1) + ( 1) + 2 () (14)
2
2
where represents the Second-order derivative, () represents the current pixel and
2
( + 1 ) and ( 1 ) represent the forward and backward pixels, respectively in (vertical,
horizontal and diagonal).
The most important purpose of using the second order derivative is to capture the strong
edges based on the different directions and values of the three color bands
20
(ICoIT'17)
10thof April, 2017.
Apart from that, various statistical equations are used to measure both the dark ad bright
areas in this research. These statistical equations include the following: the Mean, Standard
Deviation, Skewness, Central Moment with Low Order, Short Run Emphasis, Long Run
Emphasis, Gray Level Non-uniformity, Run Length Non-uniformity, Run Percentage, Low
Gray Level Run Emphasis, High Gray Level Run Emphasis, Short Run Low Gray-Level
Emphasis, Short Run High Gray-Level Emphasis, Long Run Low Gray-Level Emphasis,
Long Run High Gray-Level Emphasis, Entropy and Energy. The following section
demonstrates this set of equations [9].
1. Central Moment
p = ni=0(x )p probability(x ) (15)
Where represents the moment of order, represents the expectation function,
represents the mean, represents the number of samples, and represents the input
sample.
2. Short Run Emphasis
(,)

=1 =1 ( )
2
= (16)

=1 =1 (,)
3. Long Run Emphasis

2

=1 =1 (,)
= (17)

=1 =1 (,)
4. Gray Level Non-uniformity

=1 [=1 (,)]2
= (18)

=1 =1 (,)
5. Run Length Non-uniformity

=1 [
=1 (,)]2
= (19)

=1 =1 (,)
6. Run Percentage
1
=
=1 =1 (, ) (20)
Where n represent the number of samples.
7. Low Gray Level Run Emphasis
(,)

=1 =1 ( 2 )

= (21)

=1 =1 (,)
8. High Gray Level Run Emphasis

2

=1 =1 (,)
= (22)

=1 =1 (,)
9. Short Run Low Gray Level Emphasis

(,)

=1 =1 ( 2 2 )

= (23)

=1 =1 (,)
10. Short Run High Gray Level Emphasis
21
(ICoIT'17)
10thof April, 2017.
2 (,)

=1 =1 ( )
2
= (24)

=1 =1 (,)
11. Long Run Low Gray Level Emphasis

2 (,)

=1 =1 ( )
2
= (25)
=1 =1 (,)
12. Long Run High Gray Level Emphasis

=1 [
=1 2 2 (,)]2
= (26)

=1 =1 (,)
The number of features extracted from each channel of an image is 34 features.

Additionally, the total number of features obtained from the entire region of interest for the
four color channels is 136 features.
2.3. Edge Histogram Descriptor: During the last decades, several techniques were
introduced to serve as texture feature extraction methods. These methods were introduced
for different purposes. For instance, some of them were used for medical images and some
for entirely texture image while others have been used for real world images [7, 2]. However,
the perception of the human eyes is more sensitive to the edges of the image. As general
speaking, the edges are divided into two parts based on the intensity or texture. The MPEG-
7 descriptor was introduced a texture descriptor based on the local distribution of the edge
histogram. This method was proved to be very useful in extracting the local texture features
with real world images [11].
Generally, this descriptor was presented to describe the frequency distribution of the edges
in different five angles (horizontal, vertical, two diagonal and non-directional). The analysis
of the edge distribution makes this descriptor efficient even with the non-homogenous texture
[11]. There are three levels of processing to get the local distribution of the texture edge of
an image. First of all, the image is partitioned into 16 blocks, which are called sub-images.
Next, the edge histogram filters are applied (ex.: horizontal, vertical, diagonal 45, diagonal
135 and non-directional) on each sub-image to extract the five texture edges. Moreover, the
last step is to make normalization. Figure 2 gives an illustration of the sub-images and image
blocks of the EHD descriptor.
FIGURE. 2: Image partitioning process of EHD, (a) Dividing the Image into 16 Sub-
images, (b) Dividing the Sub-Images Blocks.
22
(ICoIT'17)
10thof April, 2017.
4. PROPOSED TEXTURE FEATURE EXTRACTION METHODS

As it was mentioned before, three methods were used to extract the texture features from the
region of interest. These methods include: the co-occurrence matrix, the gradient of the
second order derivative and the edge histogram descriptor. For the co-occurrence matrix and
the gradient of the second order derivative, the image channels are quantized from 256 levels
into N levels. In fact, the best N quantization level for the co-occurrence matrix the gradient
of the second order derivative and EHD is 16, 12 and 16, respectively. As it is stated earlier
in section 2, the main gains of using quantization are to increase the system performance and
to focus on the most dominating and adequate information as the human vision does. As the
quantization further help apply the co-occurrence matrix and the gradient of the second order
derivative on all image channels. The power of using all channels is to capture any potential
information about the texture of the region of interest.
Besides, for the EHD, there are different sizes of image partitioning as have been shown in
the literature. The most common one is (44), which means 16 sub-images will be
constructed. After that, for each sub-image, five filters that represent the edge directions with
the size of (22) were applied to extract the edge texture features. Certainly, after convolving
each block of each sub-image the maximum value among the five edge types, it will then be
compared with the edge strength threshold. If it is greater than the threshold, the block will
have sufficient edges and the histogram will be built [11]. Figure 3 shows the five edge filters.
(a) (b) (c)
(d) (e)
FIGURE 3: Edge Histogram Descriptor Filters, (a) Vertical Edge Filter, (b) Horizontal
Edge Filter, (c) Diagonal 45 Edge Filter, (d) Diagonal 135 Edge Filter, (e) Non-Directional
Edge Filter.
The last step, after building the histogram, is based on the strength edges. The histogram
distribution is normalized and is then quantized for a binary representation. To conclude,
this operation produces 80 bins of edge distributions that represent the entire image [2].
Figure 4 gives an illustration of the proposed texture feature extraction diagram.
After extracting the features by the proposed texture feature extraction methods, these
features will be proceeded by the Support Vector Machine (SVM) classifier to get the final
23
(ICoIT'17)
10thof April, 2017.
classification results. However, in this research, a libSVM with a Radial Basis Function
(RBF) has been utilized to achieve the final classification accuracy.
5. RESULTS AND DISCUSSION

The first 20 classes of Caltech 101 dataset have been used to demonstrate the performance
of the proposed texture extraction methods. To verify the strength of the proposed methods,
ten random runs have been performed. To train and test the proposed methods, fifty percent
of each class images have been used to train the classifier while the rest have been used to
test the classifier. Table 1 and Figure 5 give full details to the achieved results of the proposed
methods.
FIGURE 4: The proposed texture analysis method prototype
24
(ICoIT'17)
10thof April, 2017.
Table 1: Classification Accuracy Results of the Proposed Texture Feature Extraction

Methods Based on Caltech 101 Dataset
Co-occurrence Gradient of the Second Order Edge Histogram

Runs
Matrix Derivative Descriptor
1 43.25 45.75 69
2 42.25 45 69.25
3 44.75 49 71
4 40 45.75 72.75
5 41.25 50.25 69.75
6 41.5 45.25 71.25
7 43.25 49.5 67.75
8 41.5 50.75 69.5
9 42 47.75 71.25
10 44 49.75 70.25
Average % 42.375 47.875 70.175
STD 1.356696 2.133805 1.355775
Based on the achieved results in Table 1 and Figure 5, it has become clear that EHD has
outperformed the other state-of-the-art methods with average accuracy that is equal to 70.175
1.355775. Clearly, the main reason that makes EHD overcomes these descriptors is because
EHD is based on edge texture when describing the texture distribution of an object. On the
other hand, the state-of-the-art method relies on the intensity pixels which are highly affected
by the significant intensity variations. In other words, this work shows that the texture based
edges are more discriminative in describing the objects of the real world images even with
complicated objects and scenes.
From the other side, the standard deviation shows that all proposed methods are not much
stable with random training and testing. This parameter gives an indication that each class
may include more than one template to represent it.
Besides, the t-test statistical was used to verify the significant results of the proposed
methods. Obviously, the t-test shows that EHD is extremely and statistically significant when
it is compared with the co-occurrence matrix and Gradient of the second order derivative.
However, as it was reported in the literature, the proposed texture extraction methods have
been successfully applied to different kinds of images such as medical images, artificial
images, and industrial images. Based on the review of the literature, these methods had
shown distinctive results when they compared with other feature extraction methods such as
color and shape. In this paper work, these methods were undertaken to real world images to
verify their strength and show how much their features are distinctive for the purpose of
visual object categorization. Apparently, the texture classification with real world images
considered one of the most challenges task because of the board variation of the objects
within the same class, complex objects and scenes, and high interference between object
foreground and background. The experimental results showed that the proposed texture
features extraction methods did not give high classification results compared to the results of
the medical images, artificial images, and industrial images. On the other hand, the EHD
showed more accurate results when it compared with the Gradient of the second order and
Co-occurrence matrix because it based on edges in extracting the texture features. To
25
(ICoIT'17)
10thof April, 2017.
conclude, using the edges in obtaining the texture features can be considered more valuable
than using the intensity pixels for the real world images because the edges are more resistance
to the images effects and produce distinctive features.
6. CONCLUSION
Categorization and retrieving real world images are among the most channeling tasks in the
computer vision community. This is because of the great complicated objects and scenes.
However, in this research, a comparative study was carried out based on the real world image
dataset, called Caltech 101 dataset. The primary goal of this comparative study is to verify
the strength and weakness of the two main approaches in extracting the texture features,
which rely on edges and intensities. The results of the experiment reveal that the edge
histogram descriptor outperforms the other state-of-the-art methods. This is because EHD is
essentially based on edges in extracting the texture features that are much valuable with real
world images. Besides, the state-of-the-art methods, which are primarily based on the
intensities of pixels, are highly affected by the board variations of intensities. Lastly, the
standard deviation reveals that the proposed methods are not much stable with random
training and testing runs. This means that there is more than one template to represent each
class of Caltech 101 dataset. To expose how many templates each class contains, the
clustering technique can be appropriated for this purpose, a good suggestion for future work.
7. ACKNOWLEDGMENT
The authors would like to thank Allah for his guidance and inspiration.
8. REFERENCES
[1] Avramovi, A., evo, I. and Reljin I. 2015. Comparative analysis of texture
classification based low and high order local features. 23rd Telecommunication Forum
Telfor (TELFOR). Pp: 799 802.
[2] Ayad. H., Abdullah, S.N.H.S. and Abdullah, A. 2012. Visual Object Categorization
Based on Orientation Descriptor. Sixth Asia international conference on mathematical
modeling and computer simulation (ASM 2012). Bali, Indonesia, 29-31.
[3] Ayad, H., Abdulameer, M.H. George, L.E and Hassan, N.F. 2015. Descriptor Trends in
Texture Classification for Material Recognition. Research Journal of Applied Sciences,
Engineering and Technology. Vol. 10, Issue 10, Pp: 1089-1101.
[4] Belsare, A. D., Mushrif, M. M., Pangarkar, M. A. and Meshram, N. 2015. Classification
of Breast Cancer Histopathology Images using Texture Feature Analysis. IEEE Region
10 Conference (TENCON). Pp: 1-5.
[5] Jabid. T., Mohammadt, T., Ahsan, T., Abdullah-AI-Wadud M. and Chae, O. 2011. An
Edge-Texture based Moving Object Detection for Video Content Based
Application.14th International Conference on Computer and Information Technology.
Pp: 112 116.
26
(ICoIT'17)
10thof April, 2017.
[6] Jain, R., Kasturi, R. and Schunck, B.G. 1995."Machine vision." McGraw-Hill, New York,
NY, USA.
[7] Liua, Y., Zhanga,D., Lua, G. and Mab, W-Y. 2007. A survey of content-based image
retrieval with high-level semantics. Pattern Recognition, Vol. 40, Pp: 262 282,
[8] Nadernejad, E. 2008. Edge Detection Techniques: Evaluations and Comparisons.
Applied Mathematical Sciences, Vol. 2, No.31, Pp: 1507 1520.
[9] Salam, H.I. A. 2008. Content-Based Image Retrieval System Using Fuzzy Rule. Master
thesis. Collogue of Science, Al-Nahrain University. Iraq.
[10] Satpathy, A., Jiang, X., and Eng., H-L. 2014. LBP-Based Edge-Texture Features for
Object Recognition. IEEE TRANSACTIONS ON IMAGE PROCESSING, Vol. 23.
Pp: 1953 1964.
[11] Won, C.S., Park, D. K. and Park, S-J. 2002. Efficient Use of MPEG-7 Edge Histogram
Descriptor. ETRI Journal. Vol. 24, Issue 1.
27
(ICoIT'17)
10thof April, 2017.
A Developed Realistic Urban Road Traffic in Erbil City

Using Bi-directionally Coupled Simulations
Prof. Dr. Saad Talib Hasoon
College of Information Technology, University of Babylon - Iraq
saad.aljebori@gmail.com
Lecturer Muhammed A. Mahdi
College of Information Technology, University of Babylon - Iraq
mohamedubed@yahoo.com

Internet of Vehicles (IoV) is increasingly brought to light last years.
Article History:
It has an impact to improve the transportation system in smart
Received: 15 March 2017
cities. It is a big umbrella of VANET. IoV has contributed in
Accepted: 1 April 2017 changing the transportation system since the emergence of smart
Published: 10 April 2017 vehicles. Due to Internet service availability data can smoothly be
shared among vehicles. There are three scenarios to utilize the IoV
Keywords: VANETs,
services; these are urban, rural and highway scenarios. One of the
IoVs, IVC, SUMO,
VEINS, and Bi- main challenges in IoVs is how to simulate Inter Vehicles
directionally Coupled Communication (IVC) in a realistic simulation. This paper will
Simulations.
present a proposed realistic urban traffic simulation in Erbil city
using bi-directionally coupled simulations. A real map of the city
has been involved to simulate the traffic of vehicles; while the
Simulation of Uraban MObility (SUMO) is used to simulate the
road traffics, its output would be used as input to the VEhicle In
Network Simulation (VEINS) framework to simulate the network
under OMNeT++ environment. On the other hand, data is shared
between vehicles and roadsides to test the simulation. The results
of the simulation have been analyzed in the last stage.
1. INTRODUCTION
N
Nowadays Internet of Vehicles (IoV) has positively changed the transportation
systems. Indeed, Intelligent vehicles that have CPU and transceiver device to
communicate with each other are the main components of that system [1]. These
intelligent vehicles help in providing safe and smart environment. IoV, is subclass of Internet of
Things (IoT) by which transportation system will be more intelligent. IoVs and VANETs are
28
(ICoIT'17)
10thof April, 2017.
associated terms. While VANETs are used in numerous applications such as automobiles speed
monitoring, traffic jam avoidance, best route finding, and outside to vehicles communication [2].
VANET is subclass of IoV, where VANET is local network while vehicles in IoV need to global ID.
Internet services might be required in IoV[3]. A large number of vehicles are included in IoVs and
some roadside stations that could be deployed in an ad hoc or cellular mode. There are three
communication scenarios in IoVs. These are Vehicles to vehicles, vehicles to roadsides and roadsides
to roadsides [4].
However, vehicles are severely constrained in terms of mobility, direction and dynamic topology.
Therefore, realistic simulation needs to be considered to reduce the non-realism. Numerous methods
have been proposed to achieve some of those goals through different strategies. Each strategy can be
used to achieve some efficiencies depending on the application for which the VANET is designed.
The big challenge is how to simulate realistic vehicular networks because of their constrains [5]. In
this event, virtual reality can also be used in increasing the realism of the simulations [6]. Several
simulations are available to simulate IoVs or VANETs. Nonetheless, the realistic is not considered
in all. This paper will focus on developing urban scenario for Erbil city. This scenario has been
proposed, built, promoted, analyzed and discussed in three sequenced stages. The First stage is to
prepare a real map which is downloaded from OpenStreetMap website. The second stage will be
assigned to design a realistic traffic simulation using SUM0. The last stage is related to link SUMO
with VEINS under OMNeT++ environment for building the network. The rest structure of this paper
is divided into sixth sections in addition to the 1st section. The second section is to mention the related
works that have been done so far. The third section is to discuss the VANET simulations. The
proposed system will be investigated in the fourth section. The fifth section will be about discussing
the results. Conclusion will be the last section.
2. Related Works
Several mobility models are available to model traffic of vehicles. However, not all of these models
take in account specific characteristics of traffic urban scenario. As a result the simulation could be
unrealistic [7]. Some of them are unidirectional simulations and others are bidirectional simulations.
The latter means that there are two simulations are coupled to achieve the realism. Many
unidirectional simulators have been designed for vehicular network [8]. NS2 is used to simulate
network of vehicles [9]. However traffic simulation can be achieved through another simulation.
GloMoSim is another simulator that has ability to simulate the mobility[10]. It is designed to simulate
wired and wireless networks. VanetMobSim is widely used to simulate vehicular networks[11]. It
29
(ICoIT'17)
10thof April, 2017.
represents an extension of the CanuMobiSim. NetLogo can be also used to simulate the behavior of
vehicles [12]. Authors in [13] suggest a road model approach using NetLogo.
On the other hand, bi-directionally coupled simulations have been also proposed to combine two
simulations. First one is to simulate road traffic whereas another one is to simulate the network. In
[14], SUMO is suggested as a traffic simulation. It is coupled with NS-3 to simulate the network. In
[15], authors simulate the road traffic using VanetMobSim simulation and NS-2 as a network
simulation. VISSIM is derived from the German sentence "Verkehr In Stdten - SIMulationsmodell"
(that means "Traffic in cities - simulation model"). It is coupled with NS-2 as a realistic simulation
for VANETs [16].
2.1 VANET Simulation

Several tools of Traffic Flow Simulation (TFS) are available. SUMO, VISSIM, Paramics and
AIMSUM are examples of road traffic simulations. However, no one can simulate the communication
of VANETs. Similarly, the tools of Communication Network Simulation (CNS) are also available
(eg. NS2, OMNeT++, Qualnet etc), but they are not able to reflect the interactions between vehicles
behaviors and vehicles communication [16]. So there is a need to interactively link TFS with CNS.
There are two ways to achieve this task. The first approach is to feed the output of road traffic
simulation to the network simulation; while the second approach is to operate both simulations in
parallel. The latter is seemed to be more realistic due to the live interactions.
2.1.1 Micro Simulations

Modeling the mobility of vehicles represents a big challenge. Estimating or obtaining a model for the
real movement trace of vehicles is not a simple process. Even though it is possible to obtain an
approximation to this real movement trace from the real world, it is still not possible to simulate all
other scenarios of VANETs. Changing one parameter and keeping others may be not reasonable for
all tasks. In this context, researchers have paid their attention to design realistic road traffic
simulations that take in account vehicles behaviors [7].
The road traffic simulations can be classified into three classes according to the granularity in which
the road traffic will be simulated. These classes are macroscopic, mesoscopic and microscopic [7].
The first two classes have not an offer to simulate the single vehicle behavior and their interactions
with other vehicles, whereas the microscopic model can do it. SUMO is microscopic model which is
widely used in road traffic simulations. It can send sequence of commands and vehicles positions to
OMNET++ module. The latter can easily react to create new nodes, to delete nodes that reached to
the destinations and to move nodes depending on their behaviors in SUMO.
30
(ICoIT'17)
10thof April, 2017.
2.1.2 Network Simulations

It has been argued that simulating the network configurations is recommended before nodes are
deployed in real world. Indeed, due to the cost and danger environments it is feasible to use
simulations. Besides, evaluating the behavior of new protocols before real deployment could be more
feasible. Several network simulations are available for simulating the network communications.
Mostly network protocols are tested using discrete event simulations. Some of them are open source
such as NS2, OMNET++, Jist etc., whole others are commercial such as OPNET. This project will
use OMNET++ and its VEINS framework.
2.1.3 Bi-directionally Coupled Simulations

For any event in VANET, for example waiting time of vehicles, accident information or obstacle
information, needs to make certain interaction between road traffic simulation and network
simulation. For this reason bi-directionally coupled simulations have been developed. Two tasks can
run concurrently. One is related to the road traffic simulation, where its output will be used as input
for the network simulation. The latter will create a node corresponding to each vehicle in road traffic
simulation. Both simulations exchange their information such as speed or position of vehicles.
3. Methodology & Implementation
This section will describe the implementation of the realistic simulation. The simulation will be done
in two phases. The first phase is to simulate the road traffic in Erbil city (The capital of Iraqs
Kurdistan) using SUMO simulator. The second phase is to use the network simulation VEINS
framework, which is an OMNET++ framework. Both simulators will be operated in parallel. The
proposed bi-directional coupled simulation system in this study is illustrated in figure (1).
Start
Download a real map Create ned and ini files in Omnet++
Convert the map into xml file Build the veins project
Create route, trip, and net files
Linking both simulations
Start both simulations

Figure 1 Flow chart of the proposed
bi-directional coupled simulations End
31
(ICoIT'17)
10thof April, 2017.
3.1 Importing the Real Map
Even though it is possible to create network files manually, these file could be unrealistic. So
obtaining realistic simulation requires a real map. The openstreetmap web site provides such this real
map as .osm file [17]. This map is converted into net.xml file using SUMO using netconvert
command.. Figure (2) shows the Erbil city map in (a) and the network file in (b).
a. Erbil city map .osm file b. Network map net.xml file
Figure 2 Erbil city map
The network file defines all edges that link the start points and the end points of roads.
3.2 Road traffic Simulation (SUMO)
SUMO is an open source traffic simulator. Microscopic and mobility model are used in SUMO. It
has been developed by the German Aerospace Center (Institute of Transportation Systems) [7]. Each
vehicle has a unique path defining its source and its destination. There are four main files should be
created in SUMO. These are network file to define the road network, trip file to define the start and
end point for each vehicle, route file to define vehicles and their routes, and configuration file to
control all the mentioned previously files. As the network file is mentioned in the previous section,
the rest of files will be discussed in the next sections.
- Creating trip file: As it is discussed previously the network file defines all edges in the
network. However, vehicles paths are still not defined. For this reason trip file is responsible
to define a path for each vehicle. As it is seen trip file has parameters like route id, departure
vehicle id, and starting and ending lane.
- Creating route file: The route file defines the information of the route a vehicle would travel
through. Since each vehicle should have the route and it may have many lanes, it is not
possible to define this file manually. So SUMO provides a facility to generate this file from
the network file and trip file which are defined previously.
- Creating configuration file: Creating a configuration file is a necessary step to link all the
created files.
3.3 Network Simulation (VEINS)
There are many network available simulators. Some of them are open source whereas others are
commercial. Due to extensive cost of testing the network in real world, network simulation is used.
Most of the previous projects use the open source simulator OMNET++ and its framework VEINS
32
(ICoIT'17)
10thof April, 2017.
to simulate the vehicles network. It contains two main types of files. They are .ned files and .ini file.
The former is to describe the network whereas the latter is to include the configuration of the network.
The output of SUMO will be used as input to VEINS in parallel [18].
3.4 Linking SUMO and VEINS
SUMO contains Traffic Control Inteface (TraCI) module for OMNET++. This module contains a
small daemon, sumo-lunched, which makes running of coupled simulations easier and simpler. The
sumo-lunched daemon is listening for each incoming request. For this process TCP connection is
needed. In this connection SUMO will operate as server and OMNET as client. To achieve such that
connection opening port is also required.
At this stage, both simulators are ready to start in parallel as illustrated in figure (3).
Figure 3 SUMO and VEINS in parallel
4. Results and Analysis
1000 vehicles are used in the road traffic simulation (SUMO). However, 200 vehicles along 4 km
road are traced in the network simulation (VEINS). 1026.4 s is elapsed time for the network
simulation. In VEINS, 40448 events have occurred ranging from packet sending, waiting time and
changing positions. Figures (4 &5) show the start and end time for each vehicle.
Time
Vehicles
Figure 4 Start time of vehicles

Time
Vehicles
Figure 5 end time of vehicles

33
(ICoIT'17)
10thof April, 2017.
Busy time for each vehicle is collected and shown in figure (6).
Time/ sec
Vehicles
Figure 6 busy time of vehicles
Each vehicle changes its behavior according to the behavior of others and traffic conditions. So
each vehicle may change its speed. Where Max speed is setup to 60km/h in this project.
As it could happen in any network, some packets may be lost. If there is one packet sent by every
vehicle, there will be 199 packets should be received by each vehicle. According to these results,
decreasing the lost packets is recommended to increase the reliability of the system. So figure (7)
shows the number of lost packets in the network.
Packets
Vehicle ID
Figure 7 lost packets of vehicles
5. Conclusion
Internet of vehicles IoVs are networks of distributed autonomous vehicles and roadside stations that
can cooperate together for sharing physical or traffic conditions. It has been regarded that the
transportation systems have a significant impact in many countries. On the other side, using advance
technology to improve intelligent traffic has been available. This encourages researchers to pay their
attention to develop the transportation systems. As long as simulation of road traffic system is a
considerable issue due to high mobility that characterizes this system. So there is a need to realistic
simulation to achieve some realism. This paper presents an example of realistic simulation to
investigate the road traffic in Erbil city by combining the road traffic simulator (SUMO) and the
network simulator (VEINS). The behavior of vehicles can be enhanced according to some results of
bi-directionally coupled simulations. This project will also contribute in investigating the important
cases and challenges of the transportation system. It will also contribute in decreasing the cost and
the time of studying such those cases. In fact, the real world experiments could be risky, costly and
time consuming. Moreover, this project may help in developing Erbil city. Extending roads and
predicting challenges can be investigated before the costly real world experiments. Besides, in term
of future work, the road traffic simulations can be brought to life by using Virtual reality tools. These
tools may give more realism to the simulations.
34
(ICoIT'17)
10thof April, 2017.
References
1. Botta, A., et al., Integration of Cloud computing and Internet of Things: A survey. Future
Generation Computer Systems, 2016. 56: p. 684-700.
2. Avinash Devare, et al., A Survey on Internet of Things for Smart Vehicles. IJRSET, 2016.
5(2).
3. Keertikumar, M., M. Shubham, and R.M. Banakar. Evolution of IoT in smart vehicles: An
overview. in 2015 International Conference on Green Computing and Internet of Things
(ICGCIoT). 2015.
4. Rawat, D.B. and G. Yan, Infrastructures in vehicular communications: Status, challenges
and perspectives. IGI Global, 2010.
5. Al-Sultan, S., et al., A comprehensive survey on vehicular Ad Hoc network. Journal of
Network and Computer Applications, 2014. 37: p. 380-392.
6. Rahimian, P., et al. Using a virtual environment to study the impact of sending traffic alerts
to texting pedestrians. in Virtual Reality (VR), 2016 IEEE. 2016. IEEE.
7. Sommer, C., R. German, and F. Dressler, Bidirectionally coupled network and road traffic
simulation for improved IVC analysis. IEEE Transactions on Mobile Computing, 2011.
10(1): p. 3-15.
8. Martinez, F.J., et al., A survey and comparative study of simulators for vehicular ad hoc
networks (VANETs). Wireless Communications and Mobile Computing, 2011. 11(7): p.
813-828.
9. Sebastian, N.V. and T. Jeyaprakash, Appraising Vehicular ADHOC Networks Routing
Protocols Using NS2. 2014.
10. Hassan, A., VANET simulation. 2009, Hgskolan i Halmstad/Sektionen fr
Informationsvetenskap, Data-och Elektroteknik (IDE).
11. Alwakeel, A.M., Implementations Of The DTM, DADCQ And SLAB VANET Broadcast
Protocols For The Ns-3 Simulator. 2016, FLORIDA ATLANTIC UNIVERSITY.
12. Shanshan, W. and Z. Chunxiao. NetLogo Based Model for VANET Behaviors Dynamic
Research. in Intelligent System Design and Engineering Applications (ISDEA), 2013 Third
International Conference on. 2013. IEEE.
13. Hasson, S.T. and Z.Y. Hasan, Simulating Road Modeling Approachs in Vanet
Environment Using Net Logo. Research Journal of Applied Sciences, 2016. 11(10): p.
1130-1136.
14. Su, Y., H. Cai, and J. Shi. An improved realistic mobility model and mechanism for VANET
based on SUMO and NS3 collaborative simulations. in Parallel and Distributed Systems
(ICPADS), 2014 20th IEEE International Conference on. 2014. IEEE.
15. Poonia, R.C. and D. Bhargava, A Review of Coupling Simulator for Vehicular Ad-hoc
Networks. 2016.
16. Sun, J., Y. Yang, and K. Li, Integrated coupling of road traffic and network simulation for
realistic emulation of connected vehicle applications. Simulation, 2016. 92(5): p. 447-457.
17. Openstreetmap. [cited 2017 7th Feb. ]; Available from: www.openstreetmap.org.
18. VEINS. [cited 2017 8th Feb. ]; Available from: http://veins.car2x.org/documentation/.
35
(ICoIT'17)
10thof April, 2017.
A Hybrid Neur-Fuzzy Optimization Method for PV

Steady State Improvement in Maximum Power Point
Tracking Controller
Ashty M. Aaref
Dept. of Software Engineering, College of Technical Engineering, University of Northen, Kirkuk, Iraq
ashty_06@yahoo.com
Ali M. Humada
Electricity Production Directorate of Salahaldeen, Ministry of Electricity, 34007 Baiji,
Iraqalimhm82@gmail.com

This paper presents ahybrid Neural-Fuzzy optimization
Article History: method for the purpose of assuringthe steady state case of the
Received: 20 March 2017 Maximum Power Point (MPP) for a Photovoatic (PV) system.
Published: 10 April 2017 Also, it works to decrease the oscillation occurring around
MPP. The proposed method implemented via using an
Keywords: Maximum improved intelligent control method, neural fuzzy networks
Power Point Tracking (Nueural-Fuzzy). The results showed that the PV system with
(MPPT), PV system, Maximum Power Point Tracking (MPPT) always tracks the
Oscillation of MPPT, peak powerpoint of the PV module under various operating
Neural-Fuzzy control conditions. Afterwards, these results are validated by the
method. P&O MPPT comparison with the conventional and most used methods
method (P&O) and showed that the proposed MPPT method has
better steady state and therefore less oscillation around the
MPP than the P&O. It is also shown that, the increase in the
output power due to usingthe MPPT is about 48.2% for a clear
weather conditions and 34.8% for a shadow conditions.
1. INTRODUCTION
T
he PV solar cells are an evolving technology. The efficiencies of the solar cells
in the late 2015 reached 34.7%(Ali M Humada, Hojabri, Hamada, Samsuri, &
Ahmed, 2016a). Therefore, more focusing efforts installed via the world, its
investments increased by 300% at the end of 2015 compared to
2012(Abdelsalam, Massoud, Ahmed, & Enjeti, 2011). On the other hand, high oil prices,
pollution and political pressured to reduce carbon dioxide emissions are a few of the major
reasons that drive governments to consider renewable energy sources. The increasing
demand of electric energy drives nations to consider renewable and sustainable energies.
Several inverter topologies are being introduced to minimize losses and cost as well as
provide a better performance(Ali Mahmood Humada et al.). With the demand for more
reliable solar maximum power point tracking (MPPT), further work should be considered
regarding the control aspects of the inverter to overcome the drawback of the traditional hill-
climbing techniques. The PV solar cell model a challenging application field since the
36
(ICoIT'17)
10thof April, 2017.
changes in the solar irradiance and the temperature are not known. This nonlinear and highly
complex system of a grid-connected PV cells can be stimulated to control engineers.
Various adaptive control methods, both considering the possible variations (intelligent
control) (De Brito, Galotto Jr, Sampaio, de Azevedo e Melo, & Canesin, 2013; Dorofte,
Borup, & Blaabjerg, 2005; Elgendy, Zahawi, & Atkinson, 2012)and without considering the
possible variations (hill climbing) have been studied (Ishaque & Salam, 2013; Ji et al., 2011;
Jiang, Maskell, & Patra, 2013). PV cell inputs are highly indeterminate; it is not possible to
model the system mathematically. Therefore, there is a need for implementing control
method can consider the following issues,the inability of modelling the mathematical system
accurately, operate under a wide operating range, adaptation to unforeseen system changing
parameters and conditions, insensitivity to nonlinearity,simplicity of implementation and
design, and finally decrease the oscillation around the maximum power point.
Intelligent control systems such as artificial neural networks (ANNs), fuzzy logic (FL)
as well as meta-heuristic optimization methods meet the characteristics listed above. The
intelligent control has presented some promising results. Commonly, researchers would use
only one intelligent control technique, ANNs, FL, or evolutionary computing. The
combination of ANN and FL has proven to be effective in several applications. The ANN
compensates for the lack of learning capability and the limited adaptability of the FL. The
FL, on the other hand, is a great way to model linguistic information(Zadeh, 1996).
Combining the two methodologies provide a hybrid structure that is capable of dealing with
systems involving different types of knowledge. Therefore, the proposed
methodimplemented to solve the problems of the hill climbing methods such as, accuracy of
tracking and decreasing the oscillation around the maximum power point, to provide finally
an acceptable result compared with the most used tracking methods (P&O).
2. EFFECT OF THE PARTIALLY SHADING CONDITION ON PV SYSTEM

PV cells have an extreme sensitivity to partial shading. In cases where a cell or a small
modules part in a series string becomes shaded, instead of making a contribution to the
power output, the shaded cells get to absorb the power coming from the other cells that are
in the string. The power that is absorbed thusly is then converted into heat; this contributes
to hot spots that can bring damage to the cells. A majority of the commercial modules
employs bypass diodes across a series of cells towards overcoming this effect (Ali M
Humada, Hojabri, Hamada, Samsuri, & Ahmed, 2016b). The shaded cells are thus bypassed.
Consequently, the only thing that is lost is the power of the shaded cell series. The graphs
below illustrate the PV modules characteristics under half- and full-shading conditions.
Using a small PV system shown in Figure 1, the PV modules IV and PV curves with and
without bypass are shown in Figures 2 and 3, respectively. Figure 2 presents the IV curves
for each PV module as well as for the combined PV string with and without bypass diodes.
37
(ICoIT'17)
10thof April, 2017.
Figure 1: Two PV modules connected in series, when one module shaded and another operate with normal condition.
Figure 2: The IV characteristic curve of a PV string, two module (Module 1 and Module
2) connected in series. In partial shadowing condition, the radiance on every
PV module as follows: Module1=400 W/m2 and Module 2 = 1000 W/m2. In
the normal condition: M1 = M2= 1000 W/m2.
A different irradiance value (250 W/m2, 750 W/m2) is received by each PV module.
The PV curves are given in Figure 3, which likewise shows that, in the PV curve for the
module string, there is only one MPP that does not have bypass diodes (Ali M Humada,
Hojabri, Sulaiman, Hamada, & Ahmed, 2016). Therefore, without bypass diodes, number of
MPPs remain the same, similar to the case when there is a uniform irradiance (at 250 W/m2)
on these two modules.
A few researchers (Bidram, Davoudi, & Balog, 2012; El-Dein, Kazerani, & Salama, 2013;
Storey, Wilson, & Bagnall, 2014) made suggestions using a bypass diode for every solar
cell towards reducing the consequence of the partial shading within a PV module (i.e. power
loss), but this would bring additional complexity and cost to the module.
3. MPPT UNDER PARTIALLY SHADED CONDITIONS
As demonstrated above, towards protecting the PV panel from hot spots caused by the
module which acts as a load and consuming power, the addition of bypass diodes to
commercial PV modules is undertaken. For the PV panel protection, two bypass diodes (i.e.
Diode 1 and Diode 2) are added to the PV system; this is done through their parallel
38
(ICoIT'17)
10thof April, 2017.
connection with the PV modules (Figure 1). Additionally, towards preventing the reverse
current from the loader unbalance the current ow from other parallel PV strings, the
blocking diode is typically connected in the series PV string. Because of their series
connection for the PV string that has bypass diodes, the current that passes through these two
modules of different irradiance is equal. As such, this assemblys IV curve (in green dotted
line) can be derived through the addition of the voltages across them, while, at the same time,
the current value constant is kept constant (Figure 2). From these two modules IV curve,
there is evidence showing the formation of the two knee points, which then generates two
local maxima (Global 1 and Global 2) in the PV curve generated therein (Figure 3).
150
G3
Normal conditions (uniform)
Module 2
With baypass diode
Without baypass diode
P2
Module 1
100
Power (Watt)
P1
50 G1 G2
0
0 5 10 15 20 25 30
Voltage (Volt)
Figure 3: The PV characteristic curve of a string with the same connection and weather
conditions in Figure 1. However, P1 and P2 represent the PV modules in
normal condition and partial shading condition, respectively.
4. PROPOSED NEURAL-FUZZY SYSTEM MODEL

The hybrid NF system acts as one entity that utilizes the parallelism of the ANN and the
functionality of FL. The architecture of the hybrid NF system is represented in Figure 4.
The system implemented is four-layer ANFIS architecture with the hidden layers
performing the tasks of the fuzzy inference system. The NF system has 2 inputs and 1 output.
The first layer fuzzifies the two crisp inputs, the second layer has the fuzzy rules and applies
the antecedent IF part, the third layer is the consequent THEN part of the logic and the fourth
layer defuzzifies the fuzzy output from layer 3. A mathematical model is not needed to
represent the system. The major drawback on the other hand is once the parameters of the
system are set, it acts like a blackbox. The NF controller can either be trained offline or
online. The offline training requires prior knowledge of the input and the output while the
online training is adaptive.
39
(ICoIT'17)
10thof April, 2017.
Figure 4. Hybrid Neuro-fuzzy logic structure [24]
The system implements a feed-forward architecture where the flow of information moves
in one direction (input to output). The back propagation learning algorithm is used to train
the weights of the hidden layers. The algorithm solves an optimization problem based on the
radiant descent law, which involves the minimization of the cumulative error.
1 q
E(k) =
2
i 1
[t (k ) o(k )]2 (1)
K is the number of training patterns, which is in this case

is not predefined since the training will be done online. Vector w denotes the network weight
vector with components corresponding to the interconnection weights. The cost function that
needs to be minimized is Eqn. (2).

1 q
minE(k) = min i 1
[t (k ) o(k )]2 (2)
2
w in eqn. (3) will be the delta rule for updating rules allowing for minimization of (2).
e(k ) e(k )
w = - (3)
w w
is the gradient of the error E(k) with respect to w. is the learning rate which is fixed
throughout the simulation. The learning rate is initialed with a small value to avoid
convergence problems. Usually a momentum is added to provide fast convergence and fewer
oscillations. The momentum does not account for in the EMTP simulations. Therefore the
weigh updates will be according to 4. The first step in the training is initializing the weights
and the threshold.
The threshold of all hidden layers is initialized to 0.5 and the weights are initialized to
0.25. The weights are initialized to a small random number, hence why the 0.25 is chosen.
40
(ICoIT'17)
10thof April, 2017.
Since the training is done online, the input pattern obtained from the voltage and power
profile of the PV model. The output of the NF system is compared to the power output of the
PV model and the error is calculated for each sampling period. The error is then propagated
backward to each layer using eqn. (3), and the weights are updated. The maximum tolerable
error is set to be 1% tentatively.
5.RESULTS AND DISCUSIONS

The collected data of proposed PV system, registered from January to December 2016, the
global and diffuse solar radiation were recorded. During the midday (122PM), the solar
radiation values were recorded as high. In the early (7AM9AM) and late (57PM) daytime,
a small solar radiation value was recorded. In this tropical area, Bangi has a reasonable clear
sky for most days (at noon time), as depicted in Figure 5.
Figure 5. Daily irradiance level in Bangi area for January 2016.

The power produced from the solar PV power plant was measured after MPPT
controller to assure reliability of the proposed method. Two shadow conditions were
selected to test the credibility of the hybrid ANN-Fuzzy controller. The output curves for
the two shadow conditions presented in Figure 6.
41
(ICoIT'17)
10thof April, 2017.
Figure 6. PV characteristics of the PV array operating under shadow conditions.
The ability of tracking the MPP point throughthe above presented shadow condition was
tested by using the proposed method. The results of the tracking presented in Figures 7 and
8.
Figure 7. Result of MPP tracking by using the proposed method to compare with ideal
MPP.
42
(ICoIT'17)
10thof April, 2017.
Figure 7 presents the tracking process by using the proposed method, while Figure 8presents
the results by using the comparative method, MPPT based on P&O algorithm. By
comparison, we can see that the proposed method, Figure 8 tracked the maximum power
point before the time 1 second, however, in the comparative method, P&O method, it was
only able to track the MPP at 1.25 second. In addition, the over shot value of the proposed
method was less than the comparative method by around 50%. Therefore the proposed
method could be supposed as a novel method when compared with an efficient method like
P&O method.
____Ideal
____P&O
Figure 8: Result of MPP tracking by using the P&O algorithm to compare with ideal MPP.
CONCLUSION
A hybrid ANN-Fuzzy optimization method implemented in this study to track the maximum
power point of a PV system within shadow conditions. In this method, applied a hybrid
algorithmto assure the ability to recover the effects of shadow conditions in PV cells, which
isthe main issue in the PV installation.The proposed method has assuredits abilityof attaining
the steady state case and fast tracking time, compared with the conventional P&O method.
Also,the results showed that the proposed method could track the MPP at any condition and
day time with fast time response and low oscillation compare with the P&O method.
REFERENCES
Abdelsalam, A. K., Massoud, A. M., Ahmed, S., & Enjeti, P. (2011). High-performance adaptive
perturb and observe MPPT technique for photovoltaic-based microgrids. Power
Electronics, IEEE Transactions on, 26(4), 1010-1021.
Bidram, A., Davoudi, A., & Balog, R. S. (2012). Control and circuit techniques to mitigate partial
shading effects in photovoltaic arrays. Photovoltaics, IEEE Journal of, 2(4), 532-546.
De Brito, M. A. G., Galotto Jr, L., Sampaio, L. P., de Azevedo e Melo, G., & Canesin, C. A. (2013).
Evaluation of the main MPPT techniques for photovoltaic applications. Industrial
Electronics, IEEE Transactions on, 60(3), 1156-1167.
43
(ICoIT'17)
10thof April, 2017.
Dorofte, C., Borup, U., & Blaabjerg, F. (2005). A combined two-method MPPT control scheme for
grid-connected photovoltaic systems. Paper presented at the Power electronics and
applications, 2005 European conference on.
El-Dein, M. S., Kazerani, M., & Salama, M. (2013). Optimal photovoltaic array reconfiguration to
reduce partial shading losses. Sustainable Energy, IEEE transactions on, 4(1), 145-153.
Elgendy, M., Zahawi, B., & Atkinson, D. J. (2012). Assessment of perturb and observe MPPT
algorithm implementation techniques for PV pumping applications. Sustainable Energy,
IEEE transactions on, 3(1), 21-33.
Humada, A. M., Hojabri, M., Hamada, H. M., Samsuri, F. B., & Ahmed, M. N. (2016a). Performance
evaluation of two PV technologies (c-Si and CIS) for building integrated photovoltaic based
on tropical climate condition: A case study in Malaysia. Energy and Buildings.
Humada, A. M., Hojabri, M., Mohamed, M. B., AL-Duliamy, M. N. A., Sulaiman, M. H. B., &
Akorede, M. F. Reconfiguration Method Based on DC/DC Central Converter within
Different Mismatch Conditions.
Humada, A. M., Hojabri, M., Sulaiman, M. H. B., Hamada, H. M., & Ahmed, M. N. (2016).
Photovoltaic Grid-Connected Modeling and Characterization Based on Experimental
Results. PloS one, 11(4), e0152766.
Ishaque, K., & Salam, Z. (2013). A review of maximum power point tracking techniques of PV
system for uniform insolation and partial shading condition. Renewable and Sustainable
Energy Reviews, 19, 475-488.
Ji, Y.-H., Jung, D.-Y., Kim, J.-G., Kim, J.-H., Lee, T.-W., & Won, C.-Y. (2011). A real maximum power
point tracking method for mismatching compensation in PV array under partially shaded
conditions. Power Electronics, IEEE Transactions on, 26(4), 1001-1009.
Jiang, L. L., Maskell, D. L., & Patra, J. C. (2013). A novel ant colony optimization-based maximum
power point tracking for photovoltaic systems under partially shaded conditions. Energy
and Buildings, 58, 227-236.
Storey, J., Wilson, P. R., & Bagnall, D. (2014). The optimized-string dynamic photovoltaic array.
Power Electronics, IEEE Transactions on, 29(4), 1768-1776.
Zadeh, L. A. (1996). Fuzzy logic= computing with words. IEEE transactions on fuzzy systems, 4(2),
103-111.
44
(ICoIT'17)
10thof April, 2017.
A Multi Secret Sharing Approach for Vulnerability

Identification in Social Media
Dr. Saman Mirza Abdullah
Software Engineering, Koya University - Iraq
saman.mirza@koyauniversity.org
http://sites.koyauniversity.org/saman-mirza/

Online Social Networks (OSN)s become a suitable channel
Article History: for attackers and threats to break systems security and
Received: 19 March 2017 disclose privacies. In the few years, the demand on keeping
Published: 10 April 2017 OSN users safe from attackers was the trend of researchers.
However, it was not a challenge-less process. The most
Keywords: Social Media, significant challenge was the awareness of the OSN users
OSN, Vulnerability, about their vulnerability level. Works have been achieved to
Shamir Access Sharing. identify relations between OSN user vulnerability and the
policies that used by attackers to penetrate systems, and then
providing solutions, which were in the form of awareness that
difficult for OSN users understand them. Different
techniques, such as statistical and datamining tools, have been
provided utilized to provide solutions for identifying
vulnerabilities, however, target formation was still in
primitive stages. This work presents a secrete sharing
approach for computing the vulnerability percent of OSN
users. The approach computes the rate of the vulnerability of
an OSN user based on a polynomial equation.
1. INTRODUCTION
F rom different countries and having a variety interest and perspective, millions of
OSN user are doing huge actives every day. Activities, such as uploading
emotional information, downloading applications, and sharing or disclosing
private information are attractive to attackers. Such activities make the job of intruders and
system breakers very easy to find gaps and vulnerabilities among users and systems. Existing
of such types of OSN user activities are changing the trend direction of attackers policies
from misusing the systems vulnerabilities to misuse users vulnerabilities. Such attackers
trend encouraged researchers to focus mostly on OSN behaviors and attackers fingerprint
analysis[1].
Most researchers provided security solutions in the form of advice and awareness so that
OSN users can follow them for keeping their systems clean. Solutions that presented in the
text format are included many security and privacy related terminologies [2-4]. To
understand and follow these solutions, readers need to have good knowledge and skills about
systems and security issues, which most OSN users are not familiar with them. Moreover,
explaining a problem with texts cant give users a measurable indicator for their vulnerability
level. Therefore, OSN users need simple and understandable solutions with some measurable
45
(ICoIT'17)
10thof April, 2017.
indicators about their gaps and vulnerabilities. At the same time, it is necessary to teach OSN
users the behaviors that make them disclose to attackers. Also, users should know which
misused behaviors they have and at what vulnerability level they are.
This work focusing more on identifying the level and the rate of vulnerability of an OSN
user. The rest of the work is going like this. Section-2 provides the some works that achieved
in this field to show how researchers provided their solutions to OSN users, and expected
gap that not solved yet. Section-3, provides the concept of the secret sharing approach and
the point where crosses the concept of OSN vulnerability identification. Section-4 and later
show the process of building the model from the scratch, and some results.
2. Literature Review
In the recent years, the use of OSNs has increased dramatically. People find OSNs very
attractive as they offer many services, such as virtual relations with persons having similar
hobbies and backgrounds. Facebook, for example, currently has more than 1.39 billion active
members. Due to the portability of the current devices, location and time limitations are not
challenging the OSN users anymore to connect OSN websites. Each user has allowed doing
whatever activities he/she likes over there. Moreover, every day OSN websites are providing
new services and facilities for their users. With such availability, Internet users can stay
connecting OSN website more and more times, and can do wider and wider range of
activities. For example, users on FB doing around 3.21 billion activities per day. Of course,
not all activities are clean in the viewpoint of security and privacy standards[5].
The variety of users, activities, and even OSN websites facilitate the work of attackers to
find a number of gaps to systems and penetrating more number of users and systems.
Attackers and intruders now are learned to get a ride over systems through OSN users
vulnerability rather than systems vulnerability. Such trend changed the direction of
researching toward analyzing the users behavior more than operating systems behavior. In
such studies, the aims were not for improving the systems, however, they aimed to improve
the users behaviors. Therefore, most works provided security and privacy solutions to OSN
users [6, 7]. However, only a few of OSN users can understand these solutions perfectly.
To simply the solutions for users, most works classified the attacks and threats on OSN
users. They even sub-classify them into security related and privacy related works [6, 8-12].
The studies usually provided an overview of the types of OSN attacks. They explained the
possibility of those threats to users; furthermore, they provided recommendations and
suggestions as solutions for OSN users to protect their privacy and security. A study
presented by researchers[7] and they perfectly defined four groups of threats and attacks. In
that study, the authors offered many commercial and scientific solutions for OSN users. In
other studies [11, 13], privacy, security, and users behavior were discussed. The works
focused on one type of user vulnerability, namely Third Party Applications (TPAs). The
authors of both studies highlighted the unique security and privacy design challenges posed
by the core functionalities of OSNs, and some opportunities for utilizing social network
theory to mitigate these design conflicts and provide possible solutions to limit the
information disclosure. For both cases, it is difficult for users to understand and use those
models and theories, as they are not sufficiently clear, and users would need plenty of time
to learn them.
46
(ICoIT'17)
10thof April, 2017.
On the other hand, users need to know their vulnerability level while they are active on
OSNs. This could be done through studying the policies of the OSN based attacks and the
vulnerabilities that opened by OSN users for those attackers. The problem is how to calculate
the vulnerabilities so that a model can decide a user is disclosed or not. A linear model has
been designed by [5] and attached to Artificial Neural Network to compute the rate of
vulnerability. This work presents a Mutli-Secrete Sharing (MSS) approach to reveal the
vulnerability level of a user.
3. Shamirs based Identification Model

The concept of vulnerability is disclosing privacy and doing abnormal behaviors by OSN
users while they are interacting with systems. Within each disclose and abnormal behavior,
a secret will be provided or a channel will be opened for attackers or threats to penetrate
systems. The aim that needs to obtain is to compute the overall vulnerability rate for an OSN
user based on their disclosing and abnormally behaving rates.
This work uses the Shamirs (n, t) secret-sharing scheme. In this scheme, n shares of
original secret are created, and to retrieve or disclose the secret, an attacker needs to have at
least t number of these shares, which means any t-1 or fewer number of shares are not
allowed the attacker to reconstruct the original secrete. This is called Shamirs Threshold
Scheme. For the vulnerability perspective, an OSN user has n level of security (or
vulnerability) for a certain type of the abnormally behaviors. An attacker needs to reach a
certain level, which should be equal to t to get advantage of the vulnerability for getting-over
the system. Only on that level, the disclosed secret or opened vulnerability is useful for the
attacker. However, there is not only one secret (vulnerability) or one attack.
An OSN user may not open only a type of vulnerability (secrete), it may open more than
one channel. On the other hand, an attacker has a chance to get over a system through the
availability of only one from a set of vulnerabilities. Therefore, there is a possibility of
existing more than one channel (vulnerability), and only one channel is enough for an
attacker to get-over a system. From other side, there is a possibility of using one vulnerability
by more than one attack to get systems.
Assume V = {v1 , v2 vm } is the set of all available vulnerabilities, and the Va : Va V
is the subset of vulnerabilities that a specific attack can misuse one or some of them to
overcome a system. For both (V and Va ), the same Shamirs (t, n) threshold should be applied
on each v vulnerability. To achieve this aim, this wok proposes a multi-secret sharing
approach.
Through the proposed approach of this work, secrete V will be divided over m participants
(channels), in which each v is including all types of levels of a vulnerability (a secret)
{S1 , S2 , . Sl }c that an attacker should find t of them so that it become able to penetrate a
system.
3.1. System Model

The proposed model by this work consist of three main parts. The first part is the OSN
users whom do normal and abnormal activities and behaviors over OSN platforms. The
second part is the type of abnormal activities that considered as vulnerabilities by system.
The last part are the threats and attackers that misusing these activities (vulnerabilities) to
47
(ICoIT'17)
10thof April, 2017.
penetrate systems. Table 1 shows mapping between the required model and the proposed
scheme terminologies.
TABLE 1: Model and Scheme mapping

Shamirs Secret Defined
OSN identification Model
Sharing Symbol
Vulnerability Original message V
Types of vulnerabilitys Share of message v
Level of each v Content of v n
Attack policies Number of necessary v t
Attacker A set of t number of v Reconstruct V
The work of these three parts of the OSN user vulnerability identification regarding the
Table 1 is going like this. An OSN user has V vulnerability, which is considered as an original
message in Shamirs Secret Sharing Scheme. When the user does some activities on OSN
platform, some vulnerabilities (v) are disclosed and become public for attackers. The level
of publicity of each vulnerability depend on the frequency rate of doing that vulnerability
which rates by the users themselves. To reconstruct the original message V by an attacker, a
t number of v elements inside the V should be found.
3.2. Models Algorithm

The process of this work can be illustrating with two different algorithms. The first
algorithm is announcing the vulnerabilities by OSN users, and the second algorithm is
reconstructing the vulnerabilities by the attacker.
A. Announcement Algorithm
1. The set of all available vulnerabilities V will be initialized.
2. For any vi inside V, there is a corresponding weigh wi shows the degree of risks
that caused by the vi .
3. An OSN user dose some activities that announces all or a part of vulnerabilities of
the set V, crating Va .
4. Equation (1) is using to compute the final rate of the vulnerability for an OSN
users caused by a Va number of vulnerabilities.
V(v, w) = v1 w1 + v2 w22 + + vt wtt (1)

B. Collecting Algorithm
1. An attacker tries to scan an OSN user activities to collect the t number of

vulnerabilities that needs for attacking or penetrating a system.
2. When the number of vulnerabilities reach to some predefined t number, the system
should give an alarm that reaches a predefined level of vulnerability and indicate
the attackers that might possible to attack the system.
48
(ICoIT'17)
10thof April, 2017.
3.3. Dataset
This work is evaluating the OSN user behaviors. To start with, the work needs a set of
dataset that collecting the user behaviors. This process needs to distribute a questionnaire
sheets among OSN users and ask them about their activities they do. This work depends on
a dataset that used by [5]. The set covered 18 attack types distributed over 32 policies and
vulnerabilities. The type of attacks that covered by that are Spams, XXS and CSRF attacks,
Phishing, Internet Fraud, Clickjacking, De-anonymous, Face Recognize, SocialBot, Identity
Clone, Inference Attack, Information Leakage, Location Leakage, Socware, Online
Predators, and Cyberbullying. There are many behaviors that bring risks to users beside
systems, such behaviors called Risky Behaviors. With all these attacks and risks, there are
many unwanted software called malicious software (malwares). Possibility of becoming
vulnerable to malwares increases with increase of doing many risky behaviors and activities.
The dataset has built a nice relation between each type of attack and the vulnerabilities of the
OSN users. Table 2 explains that each attack is associating with at least more than five
vulnerabilities. The table shows the maximum number of vulnerability that an attack can
associate with them is 17.
TABLE 2: Permeance Evaluation

Attacks
A10
A11
A12
A13
A14
A15
A16
A17
A18
A1
A2
A3
A4
A5
A6
A7
A8
A9
Vulnerabilitie
Number of
12
11
11
12
12
11
25
17
11
9
7
s
There are five levels (from 1 to 5) of the risk that a threat may attack or penetrate a system.
A user at level 1 to 2 in not considered as vulnerability. However, users with levels of 3 to 5
degree are considered as vulnerable. The dataset consists of 705 observations collected in
two different countries.
4. Scheme Analysis
The proposed scheme should show a good permeance, and this section analysis this
performance in the viewpoint of the responding of the scheme to cases that given in the used
dataset. To analysis the scheme, this work will consider a case from the dataset as example
of evaluation OSN user behaviors. Table 3 shows the situation of a user. The table shows
that the behavior of this user is open for 16 vulnerabilities. For example, v1, v2 are considered
as vulnerable, while v3 is not considered as risky.
TABLE 3: An OSN user case

v10
v11
v12
v13
v14
v15
v16
v17
v18
v19
v20
v21
v22
v23
v24
v25
v26
v27
v28
v29
v30
v1
v2
v3
v4
v5
v6
v7
v8
v9
3 4 2 3 2 3 3 3 1 2 2 4 1 2 3 5 4 2 2 1 3 5 3 2 5 2 3 3 3 1
49
(ICoIT'17)
10thof April, 2017.
Through both parts of the algorithm, it could be easy to identify which vulnerability(s) a
user may have, and what is the rate, and finally awareness is already exist for those
vulnerabilities. Table 4 shows the main differences between the performance if this work
with two other recent works.
TABLE 3: Permeance Evaluation
Revealing Calculation
Rate of
Schemes and approaches Awareness Type of of
Vulnerability
attacks vulnerability
[5] No Yes No Linearly
[4] Yes No No NA
This work Yes Yes Yes Polynomial
5. CONCLUSION
OSN user vulnerability becomes a direct target for attackers to penetrate system rather
than systems vulnerability. Attackers may get advantage for more than one vulnerability to
get over the control of a system. Studies showed that three measures are important for an
OSN user; (1) awareness (2) rate of vulnerability of a user (3) type of attack that brings
threats for a use. The out performance of this work is showing that no work can include all
these three measures in one work. Moreover, the linearity of computing the rate of the work
is primitive, because an impact of single vulnerability on the risk degree should be increase
nonlinearly (not linearly) if second vulnerability is found be a threat.
6. ACKNOWLEDGEMENTS
This work wants to thank both researchers of the work [5] to provide the dataset of that
collected by them to be used by this work.
7. REFERENCES
1. Contributors:, et al. Security Issues and Recommendations for Online Social

Networks. 2007. 1, 36.
2. Franchi, E., A. Poggi, and M. Tomaiuolo, Information and password attacks on

social networks: An argument for cryptography. Journal of Information Technology
Research (JITR), 2015. 8(1): p. 25-42.
3. Savage, D., et al., Anomaly detection in online social networks. Social Networks,
2014. 39: p. 62-70.
4. Hassanzadeh, R., Anomaly detection in online social networks: using data-mining

techniques and fuzzy logic. 2014.
50
(ICoIT'17)
10thof April, 2017.
5. Abubaker, F.R. and P.S. Boluk. An Intelligent Model for Vulnerability Analysis of
Social Media User. in Future Internet of Things and Cloud Workshops (FiCloudW),
IEEE International Conference on. 2016. IEEE.
6. Sadeghian, A., M. Zamani, and B. Shanmugam. Security threats in online social

networks. in Informatics and Creative Multimedia (ICICM), 2013 International
Conference on. 2013. IEEE.
7. Fire, M., R. Goldschmidt, and Y. Elovici, Online Social Networks: Threats and
Solutions. 2013.
8. Fire, M., R. Goldschmidt, and Y. Elovici, Online Social Networks: Threats and
Solutions. Communications Surveys & Tutorials, IEEE, 2014. 16(4): p. 2019-2036.
9. Gao, H., et al., Security issues in online social networks. Internet Computing, IEEE,
2011. 15(4): p. 56-63.
10. Daniel, W., et al., PRIVACY ISSUES IN ONLINE SOCIAL NETWORKS: USER
BEHAVIORS AND THIRD-PARTY APPLICATIONS. 2014.
11. Zhang, C., et al., Privacy and security for online social networks: challenges and
opportunities. Network, IEEE, 2010. 24(4): p. 13-18.
12. Guha, S., K. Tang, and P. Francis. NOYB: Privacy in online social networks. in
Proceedings of the first workshop on Online social networks. 2008. ACM.
13. Daniel, W., et al., PRIVACY ISSUES IN ONLINE SOCIAL NETWORKS: USER
BEHAVIORS AND THIRD-PARTY APPLICATIONS.
51
(ICoIT'17)
10thof April, 2017.
A PAPER SURVEY ON THE IMPACT OF (IT) ON THE QUALITY

OF FINANCIAL DATA
(CASE STUDY OF MUNICIPALITIES AND TOURISM MINISTRY
IN KURDISTAN REGION IRAQ)
Firas N. Mardan
Accounting and Finance, College of Administration and Economics, Lebanese French University - Iraq
Firas.n@lfu.edu.krd
Farah Qasim Ahmed
Information Technology, Lebanese French University - Iraq
frhalyousuf@lfu.edu.krd

Information technology is a major resource to supply suitable
Article History: information to help enhance it on the process of rational and
Received: 20 March 2017 managerial decision-making. Information Technologies also
contribute to the administration ability to do its functions (for
example, planning, control and decision-making), and
Published: 10 April 2017 monitoring the governmental revenue, control the public
funds to ensure the safety of its spending and make sure that
the decision and administrative procedures subject to the
Keywords: Information supervision of the office of financial supervision are done
Technology, IT, according to the operative legalization, then keeping the
Information System, E-
public funds and manage it well. By doing that, the reports
accounting System, IS
will be presented with full results supreme and legislative
authorities in the government, in order to have a strong and
fit monitoring and follow-up system. To achieve the aim of this
work, the questionnaire has been distributed to the Municipal
offices that have a financial activity and have a clear picture about
the governmental accounting information system. To achieve the
aim of this work, the questionnaire has been distributed to the
municipal offices that have a financial activity and have a clear
picture about the governmental accounting information system.
This study reaches that E-accounting information system in the
ministry of municipalities and tourism has basic ingredients that
contribute to achieving the medium aims of the system. Where the
municipal offices have electronic devices and accounting programs
but these systems are different from an office to the other because
of their activities and the experiences for the employees.
52
(ICoIT'17)
10thof April, 2017.
1. INTRODUCTION
T
here are a lot of governmental organizations in Kurdistan Region that are active
in computer technologies and make use of them in their systems. By doing this,
organizations use this valuable data for information technology because of its
efficiency and productivity. At present, information technology is among most indispensable
services in the daily life and enhancing general issues, disclosure, transparency and work
efficiently to help improve the use of public resources of the government and protecting the
public funds from wastage through a comprehensive and independent oversight to help future
planning. Information technology tools lead to reduce cost, time and effort, and enhance the
efficiency of the work, and other issues that positively reflected on the general performance
of the government.
Because of the importance of this topic, this paper had been prepared to know the
accounting information system and its impact on the accounting system performance in
Kurdistan Region to do the role of accounting system in an efficiency way in the financial
and administrative control, so the governmental units can execute their role and auditing on
the institutions activities according to the allocations that specified in the budget to reduce
the financial corruption. Studying the electronic accounting information system considered
as an important subjects in various of business sectors, and especially in governmental
sectors, so that the world will direct to use and implement the electronic system in different
economic sectors, the thing that makes the governmental sector to improve their accounting
systems and use the information technology and trends towards the e-government, to take
the advantages of the benefits for the electronic systems which are the speed, accuracy, and
provide the suitable information to manage the governmental financial administrative. Where
the electronic systems contribute to providing the time to finish an extra processing and
achieve the accuracy in completion of the transactions. Information technology provides the
suitable information to prepare the public budget to execute it and control the execution. In
addition, IT contribute in to manage the general policies for the government. To achieve the
general benefits from the electronic systems the basic components should be provided, these
components contribute to achieving the benefits from the e-accounting information system.
By this information, this paper will study and evaluate the governmental accounting system
in Ministry of Municipalities and Tourism in Kurdistan Region, this ministry especially
controls the revenue, and generally control the general expenses by using a database built by
different software applications (such as Ingres, Oracle, PHP, SQL e.tc.).
2. Problem Statement:
This work is trying to answer the following:

I. Does the Electronic Accounting Information Systems in Kurdistan Region/ministry of
Municipalities and Tourism have the basic components that contribute to achieving its
objectives?
II. How does the Electronic Accounting Information Systems in Kurdistan Region/ministry of
Municipalities and Tourism meet the financial management needs?
53
(ICoIT'17)
10thof April, 2017.
3. E-Accounting Information System:
This system is for integrating the information for municipal offices, and also it is a software and
administrative system that uses the power of the computer to support and gather all the activity of
municipal offices and control the financial administration using the database.
Table 1 below will show the features and the benefits of E-Accounting Information System
Table 1 the benefits of E-Accounting IS
Feature Benefit
Including DBMS to
Reliability of accessing coordinate and accuracy
the information the information, and
enhance the reports.
Bring the data to the data
Avoid the extra process center avoiding the data
entry repeating.
Save time by the way that
the reports for the
Reduce time and costs
organization could be
enhanced.
4. The aim of the study:
This study aims at:

I. Recognizing the reality of Electronic Accounting Information Systems in Ministry of
Municipalities and Tourism according to the availability of the basic components for
electronic information systems.
II. Extending the ability of Electronic Accounting Information Systems to meet the financial
management needs to get the information about preparation of the general budget, control
over its implementation, and control the general funds that follow the ministry.
5. Hypothesis:
This study aims to test the following hypothesis:

(H1): E-Accounting Systems in the ministry of Municipalities and Tourism have the basic
components to achieve the systems aims.
(H2): E-Accounting Systems in the ministry of Municipalities and Tourism provides the quality of
the E-accounting system.
(H3): E-Accounting Systems in the ministry of Municipalities and Tourism meet the financial
management needs.
Electronic Accounting Information System depends on many components that contribute to
achieving its aims, the components are:
54
(ICoIT'17)
10thof April, 2017.
I. Equipment and devices: to achieve the aims of the system, it is necessary to take the
advantages of high-efficiency and advanced devices that contribute to the success of these
systems.
II. Automated Software: information systems cannot achieve the aims of the system without the
software that contribute to providing the information that is suitable for software. The
software is a series of instructions that make the computer has the ability to translate and
process them, this software are written in a specific language.
III. Regulations and policies: any E-accounting system is derived from a manual system,
scientific approach, and some specific instructions. In the governmental system,
governmental accounting system based on instructions issued by financial management that
help to design the E-Systems, these E-Systems matching the legislation and financial
instructions. Governmental transactions depend on some specific statements, like payable
and receivable statements that considered as main statements under a government system.
The payable statement presents the statements that paying the governmental expenses that
determined in the state budget. While receivable statement represents a specific model that
collect the money according to the instructions and financial systems that related to revenue
collecting.
IV. Administrators and managing of information technology: for the success of governmental
accounting information system in governmental institutes, the management of information
technology should exist to run and maintain the software, also to develop the system and
update it in line with the user needs. In addition to training the users to use these systems.
V. Access controls: governmental database consists of financial and non-financial information,
this information are vulnerable to hacked and miss-use, like editing or deleting, these
procedures lead to damage to society and loss of citizens' rights, so the controls should be
taken to reduce the damages like hacking and miss-uses.
6. Population of Study:
The number of the municipal offices used in the ministry of Municipalities and Tourism were 48
offices. Out of which 21 offices have been chosen because these offices are officially using the
governmental accounting information system. The questionnaires distributed to the management
department, system user, and to one (1) of IT department workers. 67 questionnaires were distributed,
but only 67 were returned, and just 85 questionnaires were thoroughly analyzed. To complete the
questionnaire the use of a five (5) Likert Scale was applied to three (variables) each respectively.
7. Analysis and Results:
This chapter begins with the preparation the data analysis. Following this, the analytic
procedure used in the study is introduced and its assumptions are tested. Then, the main thrust
of the analysis which was carried out is presented. The analysis presented covers the total
sample.
7.1 The demography details of respondence include:

- Gender
- Age group
- Level of education
55
(ICoIT'17)
10thof April, 2017.
- Department
Table 2 Gender
Frequency Percent Valid Percent Cumulative
Percent
Valid male 37 63.8 63.8 63.8
female 21 36.2 36.2 100.0
Total 58 100.0 100.0
Table 2 above shows the gender of the respondent that (63.8%) of the respondent are male
and the rest of (100%) Which is (36.2%) are female.
Table 3 Age group

Frequency Percent Valid Percent Cumulative Percent
Valid 18-25 8 13.8 13.8 13.8
26-35 19 32.8 32.8 46.6
36-45 25 43.1 43.1 89.7
46-60 6 10.3 10.3 100.0
Total 58 100.0 100.0
Table 3 above shows the age of the respondent. We can see (13.8%) of respondents are
between (18-25) years old, (32.8%) of respondents are between (26-35) years old, (43.1%)
of respondents are between (36-45) years old, and (10.3%) of respondents are between (46-
60) years old.
Table 4 Level of education

Valid Secondary
7 12.1 12.1 12.1
school
Diploma 17 29.3 29.3 41.4
Undergraduate 31 53.4 53.4 94.8
Master's 3 5.2 5.2 100.0
Total 58 100.0 100.0
Table 4 and above shows the education level of the respondent. We can see that (12.1%)
from the respondents have (Secondary school) holders, (29.3%) from the respondents have
(Diploma) degree holders, (53.4%) from the respondents have (Undergraduate) degree
holders, (5.2%) from the respondents have (Master's) degree holders.
56
(ICoIT'17)
10thof April, 2017.
Table 5 Department
Valid Accounting 9 15.5 15.5 15.5
Budgeting 21 36.2 36.2 51.7
Ownership 28 48.3 48.3 100.0
Total 58 100.0 100.0
Table 5 above shows the Department of the respondent. We can see that (15.5%) from the
respondents are (Accountant Department), and (36.2%) from the respondents are (Budgeting
Department), and (48.3.2%) from the respondents are (Ownership Department).
7.2 Scale: ALL VARIABLES
Table 6 Reliability Statistics
Cronbach's Alpha N of Items
.886 3
The reliability is =0.886
7.3 Correlations
Table 7 Correlations
V1 V2 V3
V1 Pearson Correlation 1 .906** .724**
Sig. (2-tailed) .000 .000
N 58 58 58
V2 Pearson Correlation .906** 1 .527**
Sig. (2-tailed) .000 .000
N 58 58 58
V3 Pearson Correlation .724** .527** 1
Sig. (2-tailed) .000 .000
N 58 58 58
Table 7 above, indicates to the correlation between the basic components to achieve the systems
aims and the quality of the E-accounting system, (Person correlation=0.677) it is a very strong
positive relationship, and statistically significant because its less than (5%). The correlation
between have the basic components to achieve the systems aims and the financial management
needs, (Person correlation=0. 724) it is a strong positive relationship, and statistically
significant because it is less than (5%). The correlation between the quality of the E-accounting
system and the financial management needs (Person correlation=0. 527) it is a positive
relationship, and statistically significant because it is less than (5%).
57
(ICoIT'17)
10thof April, 2017.
8. Conclusions:
This work discovered that:

The e-accounting information system in Ministry of Municipalities and Tourism has
basic ingredients that contribute to achieving the medium aims of the system.
Municipal offices have electronic devices and accountant programs but these
materials are different between each of these offices according to their activities and
the experiences of its employees.
The e-accounting information system in Ministry of Municipalities and Tourism has
a medium quality in E-accounting information systems, where these systems have
accuracy, speed, efficiency, effectiveness, flexibility and reliability of the processed
data.
The e-accounting information system in Ministry of Municipalities and Tourism
meets the medium needs of the data users and the financial management. This system
provides the information to the different financial management, regulators and aid
agencies.
9. Recommendations:
This work recommends the following:

- Working in this system and develop it to include all management levels in all
municipal offices.
- Working in this system and develop it to make it more effective through the workers.
10. References:
- Celebi, N. (2010). Public high school teachers opinions on school administrators
supervision duty in Turkey. Cypriot Journal of Educational Sciences, 5(3), 212-231.
- Elliott, R. K. 1998. Who are we as a profession. Moreover, what must we become?
Journal of Accountancy (February): 81.85.
- Gogan, J., L. M. Applegate, and R. Nolan. 1995. KPMG Peat Marwick: The shadow
partner. Harvard Business School Teaching Note 5-196-066. (Dec. 1). Cambridge,
MA: Harvard Business School.
- Reardon, J., R. Hasty, and B. Coe. 1996. The effect of information technology on
productivity in retailing. Journal Of Retailing 72 (4).
- Shafer, S. M., and T. A. Byrd. 2000. A framework for measuring the efficiency of
organizational investments in Information technology using data envelopment
analysis. Omega 28: 125.141.
- Smith, S. 1997. The smart way to invest in computers. Journal of Accountancy
(May): 63.65.
58
(ICoIT'17)
10thof April, 2017.
- Wang, C. H., R. D. Gopal, and S. Zionts. 1997. Use of data envelopment analysis in
assessing information Technology impact on firm performance. Annals of Operations
Research 73: 191.213.
59
(ICoIT'17)
10thof April, 2017.
A PAPER SURVEY ON THE IMPLEMENTATION OF THE

PARALLEL FDTD ON MULTIPROCESSORS USING MPI
Adamu Abubakar Isah
Computer Engineering, Cyprus International University North Cyprus
albattuta88@gmail.com
Asst. Prof. Dr. Oyku Akaydin
Assoc. Prof. Dr. Mehmet Kusaf
Computer Engr and Elect Elect Department, Cyprus International University North Cyprus
oakaydin@ciu.edu.tr, mkusaf@ciu.edu.tr

The research work explains a cost-effective, high-
Article History: performance computing platform for the parallel
Received: 20 March 2017 implementation of the FDTD algorithm on PC clusters using
the message-passing interface (MPI) library, which is a local
area network system consisting of multiple interconnected
Published: 10 April 2017 personal computers (PCs), and is already widely employed
for parallel computing.
Keywords: Distributed,
shared, memory,
computer architectures
and parallel computing.
1. INTRODUCTION
The Finite-difference time-domain is a numerical analysis technique used for
modeling computational electrodynamics (finding approximate solutions to the associated
system of differential equations). Finite- difference time
domain (FDTD) solves the electromagnetic wave equation in the time domain using finite-
difference approximations and is today one of the most popular technique for the solution of
electromagnetic problems [5].
1.1 PURPOSE OF STUDY
1. Performing tasks on remote computers for distributed jobs to take advantage of multi-
core workstations (FDTD, varFDTD solvers) to overcome the restriction size of
multiple cores in order to have an implementation with good efficiency
characteristics and low execution time.
2. To understand FDTD mathematically, as well as some basic FDTD algorithm,
because it seems to have a high and efficient parallelization.
3. To implement parallelizing in order to reduce memory use and improve the time for
execution . The 6 partial differential equations of the FDTD numerical algorithm
60
(ICoIT'17)
10thof April, 2017.
1.2 APPLICATION AND USES OF FDTD

In electromagnetic simulation in radar cross-section calculations
Used in microwave ovens
Antenna design analysis
Electromagnetic compatibility analysis
Medical research such as breast cancer detection
Computer chips and circuit design
Physics-based signal processing and imaging
2. LITERATURE REVIEW
2.1 PARALLEL FDTD
Generally Multiprocessors can be either: Shared memory, Distributed memory or
Distributed-shared memory.We need to parallelize FDTD to satisfy features like load
balance, memory usage balance, minimize communication overhead, and reduce sequential
bottlenecks and scalability [3].
Ey
Ex Hz
Ex
Ey
Ez Ez
Hy
Ez Ez
Hx
(i , j , k )
Ex Ey
Ex
Ey
Y
X
Figure 1: The Yee Cell [4].
61
(ICoIT'17)
10thof April, 2017.
3. METHODOLOGY
Thus, a large number of memory and computational time will be consumed when big
dimension and complex structure object is considered, which is beyond the capability of
single computer [12]. The research computation of this study is based on FDTD and MPI,
using C and its correctness will be tested by comparing different results respectively [14].
3.1 PARALLEL PROGRAMMING
In simple terms its a type of computation that allows various calculations or execution of
tasks simultaneously[15]. Parallel program consists of multi-tasks running on multiple
processors waiting to be compiled or set of instructions executed by a processor [10]. Taking
current observations of weather and processing these data with computer models to forecast
the future state of weather [9]. Used to study wealth of ocean using multiprocessors having
large computational power with low power requirements. Parallel processing is used for
modeling of economy of a nation/world. High interaction between processors along with
Shared memory modules can communicate calculations to distribute them across as many
processors as you can get simulations to complete faster, improve accuracy, or simulate
bigger physical systems [1] [8].
2.2 PARALLEL PROGRAMMING TECHNIQUES GPU
They include MPI which is explained throughout these piece of paper, VALU vector
arithmetic logic unit, AVX advanced vector extensions, CUDA compute unified device arch,
PVM parallel virtual machines and NUMA non-uniform memory access. [11].
3.2.1 GPU
Used primarily for 3-D applications. It is a single-chip processor that creates lighting effects
and transforms objects every time a 3D scene is redrawn. These are mathematically-intensive
tasks, which otherwise, would put quite a strain on the CPU. Lifting this burden from the
CPU frees up Cycles that can be used for other tasks[11].
3.2.2 CUDA
CUDA is a parallel computing platform and programming model invented by NVIDIA [13].
It enables dramatic increases in computing performance by harnessing the power of the
graphics processing unit (GPU) [6].
3.2.3. PVM
Is a software package that permits heterogeneous collection of Unix and/or Windows
computers hooked together by a network to be used as a single large parallel computer. [7].
[10]
4. DISCUSSION OF RESULTS AND COMPARISONS
The study uses MPI to develop a widely used standard for writing message passing programs.
The interface establishes a practical, portable, efficient and flexible standard, specifies a
library in a language in-dependent form and provides a bind between C and FORTRAN. It
has gained wide acceptance in parallel computing ranging from Massively Parallel systems
to network of computers and workstations [1]. [2].
62
(ICoIT'17)
10thof April, 2017.
1.50E-07
1.00E-07
5.00E-08
0.00E+00
0 100 200 300 400 500 600
-5.00E-08
-1.00E-07
-1.50E-07
Figure 2: the comparison results between execution single computer and multiple computers.
8.00E-08
6.00E-08
4.00E-08
2.00E-08
0.00E+00
-2.00E-08 0 100 200 300 400 500 600
-4.00E-08
-6.00E-08
-8.00E-08
Figure 3: the compilation and memory usage of single computer and multiple computers.
Table 1: Comparison table on the methods

SCHEME T TOTAL TIME CPU TIME MEMORY
(Secs) (MB)
FDTD 25*10-12 5 19 2.06
10 24 16.5
15 35 4.0
63
(ICoIT'17)
10thof April, 2017.
4.1 SUMMARY
The proposed system depends on progressively applying forward and differencing in time
for the conduction of execution and compilations, the is more effective for substantial
Courant numbers. This work presents another structure that allows less machine memory
storage, and more importantly CPU time, for instance in the computational area is much
smaller than the wavelength of the methods. Numerical simulation observations
demonstrates that the new technique is extremely proficient, and the results are exceptionally
well with FDTD method as explained more in the table.
5. RECOMMENDATIONS
The (FDTD) system for settling the full-wave Maxwell's comparisons has been as of late
stretched out to give exact and numerically steady operation for time steps surpassing each
other as far as possible. One such class of issues is the investigation of fast execution time
which interconnects the techniques that are frequently required for the exact determination
of a typical electromagnetic wave phenomena.
6. CONCLUSION
It has been observed that the parallel implementation of the proposed algorithm provides a
significant reduction in simulation time needed to update field components as well as the
communication time needed to perform the interprocessor communication, when partitioning
the computational domain over many processors, especially for small domains, the efficiency
of the parallelization will reach a limitation due to increase in the communication time
between the processors.
ACKNOWLEDGEMENT
All praises and thanks to Allah S.W.T who created man and thought him what he knew not.
It is by His grace that I have been able to attain this point in my life. I am indeed most grateful
to my parents Alhaji Abubakar Isah and Hajiya Fatima and my ever capable lovely
Habibateey Fatima Idris whose constant love, support and guidance have been my source of
strength all these years. I wish to express my profound gratitude to my supervisors Asst Prof.
Dr. Oyku Akaydin and Assoc. Prof. Dr. Mehmet Kuaf an erudite academics, for their time,
patience, advice and correction which have contributed vastly to the completion of this work.
I cannot forget to acknowledge the support I received from my beloved brothers and sisters;
who stood by me throughout the stormy years and gave me the courage that I very much
need to pursue my studies. Jazakallah and I love you all.
REERENCES
1. O. Ramadan, O. Akaydin, M. Salamah, and A.Y.Oztoprak, Parallel Implementation of
the Wave-Equation Finite-Difference Time-Domain Method Using the Message Passing
Interface, Lecture Notes on Computer Science (LNCS), Vol. 3280, 2004, pp. 810-818
2. K.S. Yee, Numerical solution of initial boundary value problems involving Maxwells
equations in isotropic media, IEEE Transaction on Antennas and Propagation, Vol. 14,
1966, pp. 302307
64
(ICoIT'17)
10thof April, 2017.
3. Mehmet Kusaf and Abdullah Y. Oztoprak : Multi split-step unconditionally stable finite
difference time domain methods August 2009
4. I. R. Capoglu, A. Taflove, V. Backman, "A Frequency-Domain Near-Field-to-Far-Field
Transform for Planar Layered Media", IEEE Transactions on Antennas and Propagation,
vol. AP-60, no. 4, pp. 1878-1885, April 2012.
5. M. Avinash Reddy, et al., Gain and Bandwidth Enhancement of a MSP Antenna Over
Multiple-layer Dielectric Substrate, International Journal of Advances in Computer,
Vol. 3, pp. 55-57, March 2014.
6. W. Yu, X. Yang , Y. Liu , R. Mittra, and A. Muto, Advanced FDTD Method:
Acceleration, Parallelization and Engineering Applications, Artech House, Norwood,
Mass, USA, 2011.
7. K. Yee, Numerical solution of initial boundary value problems involving Maxwells
equations in isotropic media, IEEE Transactions on Antennas and Propagation, vol. 14,
no. 5, pp. 302307, 1966.
8. A. Taflove and S. Hagness, Computational Electrodynamics the Finite-Difference Time
Domain Method, Artech House, Norwood, Mass, USA, 2005
9. L. Catarinucci, S. Guglielmi, R. Colella, L. Tarricone,Compact switched-beam antennas
enabling novel power-efficient wireless sensor networksIEEE Sensors Journal, 14 (9),
pp. 3252-3259, 2014
10. Fenghua,Z. & Zhizhang, C. (2001). Numerical Dispersion Analysis of the
Unconditionally Stable 3-D ADI-FDTD Method: IEEE Transactions on microwave
theory and technique, vol.49, no.5.
11. F.D. Hastings and others, "application of the perfectly matched layer (pml) absorbing
boundary-condition to elastic-wave propagation", The Journal of the Acoustical Society
of America, 100(5), 1996, pp. 3061-3069
12. Fu, W., & Tan, E. (2004). Development of split-step FDTD method with higher-order
spatial accuracy: IEEE Electronics letter, vol.40, no.20.
13. F.Zhang, Z Chen and J. ZhangTowards the development of a three dimensional
unconditionally stable finite difference time domain method, IEEE Trans. microwave
theory Tech, Vol 48, no 9, 2000.
14. Guilin, S. (2004). Development & Evaluation of Novel Finite-Difference Time-Domain
methods for solving Maxwells equations (PhD dissertation) Canada, Concordia
University, Montreal, Quebec, Canada.
14.1 Harari, E. Turkel Accurate finite difference methods for time-harmonic wave
propagation J. Comput. Phys., 119 (1995), pp. 252270.
15. Jaakko, S. & Theodoros, D. (2000). Reduction of Numerical Dispersion in FDTD method
through artificial Anisotropy: IEEE Transactions on microwave Theory and Techniques,
vol.48, no.4, April 2000.
65
(ICoIT'17)
10thof April, 2017.

Ahssan Ahmed Mohammed
Department of software, College IT, University of Babylon - Iraq
ahssan_ahsan@yahoo.com
Dr.Abdulkareem O. Ibadi
Department Computer Engineering, Baghdad College for Economic Sciences University - Iraq
dr.ibadi@yahoo.com

In this work, A proposed block cipher algorithm is
Article History: designed with multiple of 32 bits as a length of block
Received: 20 March 2017 performed in heterogonous multiple block cipher dependent
Published: 10 April 2017 on the application field , Multiple of 48 bits as a key length
Keywords: Encryption, are entered automatically used in variable permutation,
Decryption, Block addition function and xor operation. A proposed algorithm
cipher, Keys, SubKey, performed in single round to giving fast and high security
Map Function, Wave algorithm used in different Purposes. Another proposed
Function, Permutation. functions for different purpose such as Balance Function
between balance fixed number and initial permutation and
inverse initial permutation, Maps function as lookup table
(map2,map3,map4 and map5). Wave Function as a nonlinear
function of 8-bits of input and same for the output. This
algorithm can be designed in several blocks length depends
on the application field, increase speed, higher complexity
and for reduction of cost.
.
1. INTRODUCTION
E
ncryption Operation is the process of converting a original data called plaintext
into ciphertext using secret key, and can be convert back into the original data
with the same secret key. There are several kinds of data encryption schemes
which form the fundamental of security. [1]
Encryption Operation schemes are generally based on either block cipher or stream cipher.
Historically for providing confidentiality should be Encryption focus on symmetric
encryption. It is only in the last several decades that other considerations, such as Integrity,
Authentication and Digital Signature have been included in cryptology theory and it practice.
The security of the data basically depends on two laborer first one is Confidentiality and
second one is Authentication. Achieving confidentiality means the data is encrypting bulk
digital data used block cipher. Single round of encryption offers inadequate security but
multiple rounds offer increasing security but the side effect is time consumed not always, it
is dependent on the strength and security of encryption functions used .[2]
The proposed algorithm is different of another by used block length multiple of 32
bits as a homogenous block cipher (4-8-12-16 bytes) with multiple of 48 bits of key and 4
nonlinearity Maps(Map2, Map3,Map4 Map5) all of them processing in one round and new
subkey generation algorithm also been designed new function for generated permutation
66
(ICoIT'17)
10thof April, 2017.
array used as initial permutation and inverse initial permutation then reversible variable
permutation dependent on secret key value as well Wave function is a nonlinear function.
Cryptography is one of many aspects of security, It is science and art of writing.
Egyptian Hieroglyphs in an inscription is first used of cryptography in 1900 B.C. The
primary functions of cryptography are confidentiality, Privacy, Authentication, Integrity,
Non-repudiation and Key exchange.[3]
The three types of Cryptography algorithms are :
1- Secret Key Cryptography or symmetric encryption : Used a single key for both
encryption and decryption, Primarily used for confidentiality and privacy.
2- Public Key Cryptography or asymmetric encryption: Used two keys, one of them
for encryption and another for decryption, mostly used for key exchange, non-
repudiation and authentication .
3- Hash Function: Used a mathematical transformation to encrypt of information,
providing a digital fingerprint. mostly used for message integrity.
Secret key cryptography algorithms in general categorized as being either stream
cipher or block cipher algorithms will focus on this aspect in our proposal. Stream ciphers
operated on a one bit at a time and key is continuously changed, means the Encryption
Operation of each digit depends on the current state of the cipher engine, Stream cipher is
also call state cipher. Generally, single bit/bites are used as single digits to avoid concerns of
security . Error propagation is a problem of the Stream cipher as a bit corrupted in
transmission will result of corrupted set of bits at the receiving side.[4]
Block Cipher Encrypts/Decrypts one block from data at a time used the same key on
each block with fixed block/key length. Encryption key scheme is determines the order in
which Transportation, Substitution and other mathematical functions are applied on each
block. Block cipher algorithms used two operations are Confusion and Diffusion to encrypt
Plaintext Block into Ciphertext Block. The aim of confusion is to make the relationship
among the encryption key and the ciphertext as complex as possible. Ideally, every bit in the
key should influence in every bit of the Ciphertext Block. By contrast, Diffusion operation
propagates the influence of each bit in the Plaintext Block over several bites in the Ciphertext
Block, making the cipher few oversensitive to statistical attacks.[6]
The actuality that stream cipher data Encryption/Decryption single bit at a time means
that they are particularly well appropriate to real time hardware applications, like image ,
video and audio applications. Stream cipher algorithm is practical weaker and less efficient
than block cipher algorithm for software applications and it minimal frequently used in that
sphere. Block cipher algorithm is easier to implement in Software application because the
message encrypted in blocks length that software already uses. The Encryption key is
predominantly the same size as the block length. Hence block ciphers achieve higher
diffusion and error propagation than stream ciphers, although it is worth noting that the
algorithm determines the amount of diffusion.[4][5]
Block ciphers algorithms usually require large memory, because they perform on larger
data chunks and predominating have carry over from preceding blocks, whilst Stream ciphers
algorithms perform on only a fewest bits or byte at a time they have relatively less memory
requirements and therefore cheaper to implement in limited scenarios like firmware,
embedded devices and special hardware.[5]
Stream ciphers algorithms are more difficult to perform appropriate, and apt to
weaknesses based on usage, since the fundaments are similar to one time pad, the key stream
cipher has more stringent requirements.[4]
67
(ICoIT'17)
10thof April, 2017.
On the other hand, Block ciphers algorithms encrypt a whole block at a time and
moreover have feedback modes which are most recommended, they are over sensitive to
noise in transmission, means if you mess up one data part, all the remnant is maybe
unrecoverable. while stream ciphers algorithms bytes are individually encrypted with no
connection to other data chunks in most ciphers/modes, and typically have support for
interruptions on the line.[5]
Stream ciphers algorithms do not provide security authentication or integrity
protection, but some block ciphers algorithms can provide integrity protection and
confidentiality depending on mode.[4][5]
The reason for choosing Block ciphers in this proposed because more useful when the
data amount is pre known such as a data file, fields, or request protocols.
2. A PROPOSED BLOCK CIPHER

A proposed block cipher algorithm is a based on Non-Feistel Block Cipher algorithms.
The block diagram of it is shown in figure(1) and the algorithm is given in the following
eight steps
Algorithm:
INPUT : block of length 32 bits, key of length 48 bits.
OUTPUT: block of length 32 bits.
1. Compute 32 bits after initial permutation by balance function = P block Xor
balance constant block.
2. Compute addition operation on balance block and subkey1 to yield P sum block
(each 8 bits of balance block addition to a single 8 bits of subkey1).
3. Perform Map Function on P-sum block to convert each 8 bits to covalence bits.
4. Perform variable permutation on P-Map block based on bits location of key.
5. Perform Reversible Wave function on variable permutation block.
6. Shifting and Rotation operations for each 8 bits of wave block.
7. Compute Xor operation on each 8 bits of block with different 8 bits of subkey2.
7. Swap among each 8 bits of block to produce the C block of length 32 bits.
8. Next 32 bits.
68
(ICoIT'17)
10thof April, 2017.
Figure: 1 Illustrate Encryption & Decryption algorithm structure
The main idea of a proposed block cipher algorithm is begin by input block of
plaintext and the output is block of ciphertext after executed in number of different steps.
The algorithm has a multiple of 32-bits block size and multiple of 48 bits key length.
It is a 7-step built based on non Feistel cipher algorithm. The blocks are constructed by
converting a characters into 32 bits by using ASCII and also entered key by converting a 6-
character into 48 bits by using ASCII for each character. The figure 2&3 illustrate encryption
and decryption steps structure.
2.1. BALANCE FUNCTION

Two main functions are operate in Balance function they are : Initial permutation and
XOR function, The input/output of the Balance function is a block of length 32 bits, the input
block permutated in permutation generation algorithm to generate initial permutation (IP) as
in table 1 and inverse permutation (IP-1) as in table 2 for first function then Xor-ed with a
block of balance constant of the same length. The block diagram of it is shown in figure 2.
Table 1 : Initial Permutation IP
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
4 11 24 13 9 16 3 17 8 14 2 20 19 0 26 15
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
10 7 18 27 12 23 6 25 31 28 1 21 30 5 29 22
69
(ICoIT'17)
10thof April, 2017.
Table 2 : Inverse Initial Permutation IP-1

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
13 26 10 6 0 29 22 17 8 4 16 1 20 3 9 15
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
5 7 18 12 11 27 31 21 2 23 14 19 25 30 28 24
2.1.1 PERMUTATION GENERATION ALGORITHM

Input : list of number schedule arranged from 0 to n-1
Output: array of random numbers between 0 to n-1 without repetition
1. Loop from 0 to n-1 .
2. Using predefined random function in vb.net (Random) working storage area.
3.Using list operation to move location of random number generated to the output array
4. Remove the location and reduce list length by one .
5. Next.
Figure 2: Illustrate Balance Function structure
2.2. ADDITION FUNCTION

Addition operation is second function of a proposed block cipher algorithm. The output
block of Balance function will dividing into set of 8-bits each one will addition with single
8-bits of subkey1 block as performed by used permutation and selection for the secret key as
shown in table 3.
Figure 3: Illustrate Addition Function structure
Table 3 : Permutation and Selection of Subkey1
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
47 31 22 38 29 1 10 45 41 36 16 12 37 28 17 2
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
9 15 32 3 39 19 6 18 7 14 26 8 24 40 34 33
70
(ICoIT'17)
10thof April, 2017.
2.3. MAP FUNCTION

Map function is performed for a byte of input block, A proposed block cipher used 4
maps each one has a different number. Different permutation is present after each Map to
help for bits propagation and diffusion increased with high complexity.
2.3.1 Map Function generation
Map generated by converted the input byte into complement number excepted the
bits has location equal to the random value generate between 0 -7. A proposed block cipher
used 4 maps each one with different number of excepted bits (Map2 means two bits has
generated and so on for Map3, Map4 and Map5). Maps function structure is shown in figure
4. Table 4 is example of Maps function.
Table 4 : Example of Maps Function

Input Random Values Output
Map2 10001110 2-4 01011001
Map3 10010011 3-4-7 01110101
Map4 10011000 0-1-6-7 10100100
Map5 10011101 0-1-2-3-4 10011010
Figure 4 : Maps Function structure
2.4. REVERSIABLE VARIABLE PERMUTATION

Variable permutation is a reversible function, it relying to secret key values. Variable
permutation performed for each byte of block dependent on location of secret key bits. The
variable permutation is implement by three levels as shown in figure 3. Level one check the
secret key and make swapping between two neighboring bits when the read bit is one. Level
two and three perform swapping between widely bits as shown in following structure to
guarantee permute from first to last bit.
71
(ICoIT'17)
10thof April, 2017.
Figure 5 : Reversible Variable Permutation(VP) structure
2.5. WAVE FUNCTION

The wave function is a nonlinear function designed to be of the same size inputs
and outputs which help to increase the confusion and diffusion of output result because
whenever a zero is located in a sequence the rest of the sequence will be converted[7]. A
proposed block cipher algorithm used the Wave function with byte of input and same for output.
If X is the input set and Y is the output set so Y can write as Y=WF(x) where,
Y1=X11
Y2=X2X11
Y3= X1X2X31
Yn=X1X2.Xn-1Xn1
Figure 6 : Reversible Wave Function structure
Another feature of the Wave Function is that never results in the same sequence
and the output will be repeated after 2n trial.
2.6. SR FUNCTION
A Shifting and Rotation function is propagation function for each byte of block. A
proposed block cipher used four shifting and rotation function (S&R1 means shifting each
byte by one and so on for S&R2 , S&R3 and S&R4).
2.7. XF FUNCTION
XF function is xor operation between input block and subkey2, The input of the XF
function is a block of length 32 bits Xor-ed with a block of Subkey2 of the same length
generated by permutation and selection operation for secret key. In the end the output will
72
(ICoIT'17)
10thof April, 2017.
swap among each 8 bits and saving the ciphertext block in the binary file as shown in the
main structure. Figure 6 illustrate XF function structure or single 8 bits
Figure 6 : Single XF function structure
2.8. KEY MANAGMENT

A proposed block cipher algorithm used multiple of 48 bits as a secret key by entered
automatically dependent on application field such as password , data, image, audio, video,
satellite communication its. Secret key used in variable permutation to generate permutation
dependent on each bit as discussed in 4.4 . Subkeys are generated by scheduling of secret
key in two operations as selection and permutation . The input of the algorithm is the secret
key (6 character) witch are converted into 48 bits using ASCII code, A subkey1 and subkey2
are generated by selection and permutation of 32 bits from the 48 bits of secret key as shown
in a proposed algorithm structure.
3. EVALUATION
There are two types of block cipher Feistel and Non-Feistel cipher, a feistel cipher
algorithm is a block cipher construction as a symmetric build. This came on the label of a
German inventor Horst Feistel. A Wide proportion of block ciphers use the Feistel network,
such as the (DES) Data Encryption Standard, Blowfish, KASUMI, Simon and etc. The
advantage of the Feistel Network is very similar for the encryption & decryption operations
it is identical in some cases and requiring only moving from up to down ,From top to bottom
or a reversal of the key schedule. Feistel ciphers and similar algorithms are product ciphers,
and repeated operations for number of rounds.[5]
A non Feistel cipher is the sound type of block cipher so a Non-Feistel block cipher
algorithm is a chain of linked mathematical operations performed in block cipher, it is
network with a block of Plaintext and Key as inputs and applies several alternating
operations to produce the ciphertext block such as AES , Square, 3-Way, SHARK , SAFER
, PRESENT, and , Kuznyechik .
Although the Feistel block cipher that used S-box (like Data Encryption Standard) is fully
similar to Non-Feistel block cipher, but there are some differences that make either Feistel
or Non-Feistel more applicable in assured situations. For a given amount of Diffusion and
Confusion, an Non-Feistel block cipher has extra inherent parallelism[11], and so given a
CPU with a more number of execution units can be computed faster than a Feistel block
cipher.[12] CPUs with less execution units such as most smart cards cannot take advantage
of this inherent parallelism. Also Non-Feistel ciphers require s-box to be invertible (to
perform decryption). Feistel inner functions have no such restriction and can be constructed
as one way functions.
The advantage of a proposed algorithm is that, since its external and internal properties
are change from function to function, it may be extremely difficult to find any kind of
characteristic that propagates well through all of the different kinds of functions that appear
73
(ICoIT'17)
10thof April, 2017.
in the cipher. However, implementing of a proposed algorithm and analyzing it is often much
more complicated. Hardware implementations are generally cheaper, and software
implementations smaller and easier to correctly code.
A proposed block cipher algorithm used a balance function, These structure provide
efficient implementations for mutable of the 32 bits of plaintext block by initial permutation,
and used xor-ed operation with fixed balance binary number for solving problem when all
plaintext block (0) or (1).
A Map function is substitutes a byte (the input of map function) by another byte ( Map
Function output). This Substitution should be one byte to one byte, to ensure inevitability for
decryption of a proposed algorithm. The good Function goal is changing one input bit will
change about half of the output bits.
The Wave Function is a reversible nonlinear function which helps to increase the
confusion of the resulting output besides, it helps to increase the diffusion because whenever
a zero is located in a sequence the rest of the sequence will be converted. Another feature is
that never results in the same sequence and the output will be repeated after 2n trial.
Addition operation with subkey1, Variable permutation dependent on key locations
values, Shifting and Rotation operations of byte plaintext block, Xor-ed with subkey2 and
Bytes Swapping operation are helps for increasing alteration among each byte.
Diffusion is important concept in the design of robust of block cipher algorithm, It is
operation between plaintext block and ciphertext block means if changed one bit of plaintext
block then several bits of ciphertext block should be changed and Vice versa. In another
interpretation, This means the frequency statistics of bit in the plaintext block are diffused
over several bits in the ciphertext block and statistical attack need more meaningful to do a
ciphertext block. Table 5,6 shown the range of diffusion between plaintext block and
ciphertext block in the block cipher algorithm proposed.
Block length Key length Diffusion Range

32 48n 12-18
64 48n 27-35
96 48n 40-57
128 48n 61-68
Table 5: shown the range of diffusion with different of block size .
Key 000000000000000000000000000000000000000000000000 Zero One Diffusion

Plaintext 00000000000000000000000000000000 32 0 -
Ciphertext 00111011100111111111101001011110 10 22
Plain 00000000000000000000000000000001 31 1
Cipher 01011001100010010011000011011100 18 14 12
Plain 00000000000000000000000000000011 30 2
Cipher 00110001100110111011110100011110 14 18 12
Plain 00000000000000000000000000000111 29 3
Cipher 11010010101110101011110011101010 13 19 13
Plain 00000000000000000000000000001111 28 4
Cipher 11011110010011011001001010110110 14 18 17
Plain 00000000000000000000000000011111 27 5
Cipher 00010011011001010010001010110011 18 14 12
74
(ICoIT'17)
10thof April, 2017.
Plain 00000000000000000000000000111111 26 6
Cipher 10110100101001001011101000111011 15 17 13
Table 6: shown the range of diffusion with same of block size .
4. CONCLUSION
A proposed block cipher algorithm designed to create a strong algorithm that runs in
a reasonable time on presently available hardware, that is rationally easy to implement,
etc. As such, a proposed algorithm design may allow us to design faster and better block
cipher algorithm than we could if we limited ourselves to number of round or Non-Feistel
network structures.
1- We concluded the relationship between the blocks length and hardware , It can
perform multiple of 32 bits as block length as a homogenous algorithm to reduce the
time of execution
2- A number of rounds should be less to speed increasing.
3- Multiple of 48 bits as a secret key should be entered automatically and it dependent
on application field such as password , data, image, audio, video, satellite
communication its.
4- Permutation operation should be after each map function to propagate bits in each
block and increase confusion and diffusion operations.
5. REFERENCES
References must be listed in the order they were cited (numerical order). The references
must not be in alphabetical order eg:
1- Suzaki, T., Minematsu, K., Morioka, S., Kobayashi, E.: Twine: A lightweight block
cipher for multiple platforms. In Selected Areas in Cryptography - SAC 2012. LNCS,
vol. 7707, pp. 339-354. Springer, Heidelberg (2013).
2- Wu, W., Zhang, L.: Lblock: A lightweight block cipher. In Applied Cryptography
and Network Security - ACNS 2011. LNCS vol. 6715, pp. 327-344. Springer,
Heidelberg (2011).
3- van Tilborg, Henk C. A.; Jajodia, Sushil, eds. (2011). Encyclopedia of Cryptography
and Security. Springer. ISBN 978-1-4419-5905-8., p. 455.
4- Junod, Pascal & Canteaut, Anne (2011). Advanced Linear Cryptanalysis of Block
and Stream Ciphers. IOS Press. p. 2. ISBN 9781607508441.
5- Chakraborty, D. & Rodriguez-Henriquez F. (2014). "Block Cipher Modes of
Operation from a Hardware Implementation Perspective". In Ko, etin
K. Cryptographic Engineering. Springer. p. 321. ISBN 9780387718163.
6- Lars R. Knudsen and John Erik Mathiassen, On the Role of Key Schedules in Attacks
on Iterated Ciphers, ESORICS 2014, pp322334
7- Dr.Abdulkareem O. Ibadi, Special Security e-Mail System , Ph.D. dissertation,
University of Technology , June 2007
8- Ben Morris, Phillip Rogaway, Till Stegers. "How to Encipher Messages on a Small
Domain". CRYPTO 2010.
75
(ICoIT'17)
10thof April, 2017.
9- Menezes, Alfred J., Oorschot, Paul C. van; Vanstone, Scott A. (2011). Handbook of
Applied Cryptography (Fifth ed.). p. 251.
10- Christof Paar, Jan Pelzl, "The Data Encryption Standard (DES) and Alternatives",
free online lectures on Chapter 3 of "Understanding Cryptography, A Textbook for
Students and Practitioners". Springer, 2012.
11- "Principles and Performance of Cryptographic Algorithms" by Bart Preneel, Vincent
Rijmen, and Antoon Bosselaers.
12- "The Skein Hash Function Family" 2010 by Niels Ferguson, Stefan Lucks, Bruce
Schneier, Doug Whiting, Mihir Bellare, Tadayoshi Kohno, Jon Callas, Jesse Walker
page 40.
13- Sinha, S., Arya, C. (2012). Algebraic Construction and Cryptographic Properties of
Rijndael Substitution Box. Defence Science Journal, 62(1).
14- Paar, C., Pelzl, J. (2010). Understanding Cryptography: A Textbook for Students and
Practitioners. New York City, New York: Springer Publishing.
15- Grocholewska-Czurylo, A. (2011). Cryptographic properties of modified AES-like
S-boxes.
16- Annales UMCS Informatica, AI XI(2), 37-48. DOI: 10.2478/v10065-011-0009-4.
17- Easttom, C. (2015). Modern Cryptography. New York City, New York: McGraw
Hill.
76
(ICoIT'17)
10thof April, 2017.
A Proposed Public Key Encryption Based on Hybrid Chaotic Maps

Prof. Dr. Ahmed T. Sadeeq
Department of Computer Science, University of Technology - Iraq
drahmaed_tark@yahoo.com
Assist Prof. Dr. Alaa K. Farhan
Department of Computer Science, University of Technology Iraq
Dralaa_cs@yahoo.com
Shaimaa A. Hassan
Institute of Medical Technology, Middle Technical University- Iraq
aminamoj@yahoo.com

Cryptography has long history by providing a way to
Article History: store sensitive information or transmit it across insecure
Received: 19 March 2017 networks (i.e. the Internet) so that it cannot be read by
Accepted: 1 April 2017 anyone except the intended recipient. Because data
transmitted over the network is vulnerable to attack, it
Published: 10 April 2017
must be considering the security aspect of it. In this
paper, a public key encryption algorithm based on
Keywords: Public Key hybrid chaotic maps is proposed. The proposed
Encryption, Hybrid algorithm uses a mixing of three dimensional Logistic
Chaotic Map, 3D
map, three dimensional Arnold Cat map, two
Logistic Map, 3D
dimensional Rotation Equation and Chebyshev map to
Arnold Cat Map,
Rotation Equation.
set random values to this algorithm and to generate
privet and public keys that are used to encrypt and
decrypt data. The experimental results show that the
generated keys have the characteristics of truly random
numbers and pass most of statistical and NIST tests.
1. INTRODUCTION
F orthousands of years cryptography and encryption have been used for secure
communication. Cryptography is the science and the study of secret writing
techniques [1].
Asymmetric encryption also known as public-key encryption is one form of
cryptosystem in which different keys are used to perform encryption and decryption one is a
public key and the other is a private key. Public Key encryption can be used for
authentication, confidentiality, or both [2].
The major advantage of using the public key encryption is that they avoid the need of
sharing any shared key between the transmitter and the receiver via a secure channel. In a
shared key encryption scheme the transmission of a private key via a secure channel can
cause a completely danger to the encryption scheme [3].
77
(ICoIT'17)
10thof April, 2017.
Chaos theory is a field of mathematics. It is the study of complex, nonlinear and dynamic
systems. This field was pioneered by Lorenz in1963. [4]. Nonlinear equation is deterministic
that can generates random or chaotic behavior over time [5]. Dynamic system is very
sensitive to initial condition and its response known as a Butterfly Effect, in which small
change in initial state has led to unpredictable change in the final state [6]. The chaotic
behavior has been widely studied and many generators have already been used to generate
sequences of pseudo random numbers [7].
2. RELATED WORK
Many researchers uses chaotic map in public key encryption. In [8] a public-key encryption
algorithm based on iteration of one-dimensional Chebyshev chaotic maps and two-
dimensional of torus automorphisms chaotic map was proposed. This encryption schemes
are both secure and practical, and can be used also for digital signature.
In [9] a new cryptosystem by using iterations of an expansion chaotic map was proposed.
The expansion map is modified logistic map. The proposed cryptosystem is a symmetric-key
cryptography has characteristic of public-key cryptography, that uses three kinds of keys, a
public key, a private key and a common private key.
In [10] an expansion of the public key encryption based on chebyshev polynomials with
a modest hash function was proposed. This model can be applied on multilevel inputs types
such as images and videos.
3. PUBLIC KEY ENCRYPTION BASED ON CHEBYSHEV MAP

A public-key encryption algorithm based on chaotic maps is first proposed by Alfred in 1996.
Chaotic map used in this algorithm is Chebyshev map which shown in Equation (1). The
algorithms described here use a remarkable property called semi group property as shown in
Equation (2).
Tn(x) = 2xTn1(x) Tn2(x) (1)
Where n 2, x [1, 1], T0(x) = 1 and T1(x) = x.
Tr(Ts(x)) = Trs(x) (2)
Public key encryption based on Chebyshev map consists of three algorithms. The first one
is the Key generation algorithm (shown in Algorithm (1)) that generates public and private
keys. The second one is the encryption algorithm (shown in Algorithm (2)), while the third
algorithm is the decryption algorithm (shown in Algorithm (3)) [10].
Algorithm (1): Key Generation.
Output: Pubic key (x,Ts(x)), Private key s.
Begin
Generates a large integer s;
Selects a random number x in the interval [-1, 1] and computes Ts(x);
Sets the public key to (x, Ts(x)) and the private key to s;
End.
Algorithm (2): Message Encryption.

Input: public key (x, Ts(x)).
Output: Cipher message C.
Begin
Obtains authentic public key (x, Ts(x));
Represents the message as a number M in the interval [-1,1];
Generates a large integer r;
78
(ICoIT'17)
10thof April, 2017.
Computes Tr(x), Trs(x) = Tr(Ts(x)) and X =M Trs(x);

The cipher text C = (Tr(x), X);
End.
Algorithm (3): Message Decryption.
Input: Private Key s.
Output: Message M.
Begin
Uses the private key s to compute Tsr(x) = Ts(Tr(x));
Recovers M by computing M = X/Tsr(x);
End.
4. THE PROPOSED ALGORITHMS OF PUBLIC KEY ENCRYPTION BASED ON

HYBRID CHAOTIC MAPS
This method is depends on chaotic maps to generate private and public keys to encrypt and
decrypt data using public key encryption. So, the main purpose of this improvement is to
strength the randomization by using hybrid chaotic maps.
The proposed algorithm uses hybrid chaotic maps in setting the main parameters,
generating public and private keys and also in encrypting and decrypting messages. To
provide a good environment for this map to work, numbers between 0 and 255 are divided
to interval between 0 and 1. The Fraction_code algorithm shown in Algorithm (4) takes
integer values between 0 and 255 and converts them to 256 real ranges between 0 and 1 to
support the values resulted from the chaotic maps.
The main parameters (s,r,x) are set by using a combination of hybrid multidimensional
chaotic maps. The chaotic maps that are used as key generator consist of a combination of
three dimensional Logistic map shown in Equation (3) [11], Rotation Equation map shown
in Equation (4)[12,13] and three dimensional Arnold Cat map shown in Equation (5) [14,15].
The block diagram of this combination is shown in Figure (1).
+1 = (1 ) + 2 + 3
+1 = (1 ) + 2 + 3
+1 = (1 ) + 2 + 2 (3)
Here Equation (3) exhibits the chaotic behavior for 3.53<<3.81, 0<< 0.022, 0<<0.015,
and can take the value between [0,1].
+1 = ( + ) cos + sin
+1 = sin + cos
= 0.5 (2 + 4 + 42 ) (4)
Where the parameters are = 2 and a = 2.8, with initial conditions x0 = 0.5, y0 = 1.0.
x 1 a 0 x
[y] = [b ab + 1 0] [y] mod 1 (5)
z c d 1 z
After the initial values are set, each one of those maps generates real values that are
converted to binary by using a threshold value as shown in Algorithm (5). The threshold
value that is used in the present study is equal to 0.5. So, the resulted binary digits from those
maps are eight digits (three digits from Logistic and Cat maps and two digits for Rotation
map). Those digits are merged together and the result is converted to decimal. This decimal
value is used to loop Chebyshev map shown in Equation (6) [16].
x(n + 1) = Tk(xn) = cos(k arccos(xn)) (6)
Where xn [1, 1] and k [2,).Restricted in the interval [1, 1] [10].
79
(ICoIT'17)
10thof April, 2017.
The final result is converted to binary number. Then, a normalization process is made to
that binary number. This process is described in Algorithm (6). The first loop is used to set
the parameter s which is a large integer that represents the private key. The second loop is
used to set the parameter r, while the third loop is used to set random x which it is between
0 and 1. In key generation step shown in Algorithm (7), the s and x parameters are used with
a Rotation equation to compute the public key. The r, x and s parameters are used with a first
order Chebyshev map and another with a Rotation equation to encrypt and decrypt message.
The encryption algorithm is shown in Algorithm (8).
In decryption steps shown in Algorithm (9) the same chaotic maps with the same initial
values are used to generate the random values. Those values are converted to binary using
Algorithm (5) and an inverse Fraction_code is used in decryption process to convert the result
to ASSCII form as shown in Algorithm (10).
Algorithm (4): Fraction_code
Input: integer values between 1 and 256;
Output: Fraction _code as array of ranges between 0 and 1;
Begin
For i=0 to 255
Represent the value of i as a real value in range [i, i+1) with three fraction digits;
Put the result of above step in Fract_array[i];
End For
Fraction_code = Fract_array;
End.
Algorithm (5): Convert The Result of 3-D Chaotic Map to Binary Using Threshold Value.
Input: Result of 3-D chaotic R.
Output: Binary-digits.
Begin
For j=1 to number of R
If Rj >= 0.5 Then
Binary-digits= Binary-digits & 1
Else
Binary-digits= Binary-digits & 0
End If
End For
End.
Algorithm (6): Normalization of Binary Values.
Input: Real Value R.
Output: Normalized Binary Value BN.
Begin
B= Convert R to binary;
If B contains exponent part and fraction part Then
BN=merge the fraction with the exponent after eliminate '.'
Else If B in exponential (Floating point) format Then
BN= B* 10^number of digit after exponential part
Else
BN=B
End If
End.
80
(ICoIT'17)
10thof April, 2017.
Hybrid 3-D Convert to binary

Initial chaotic maps using threshold R
3-D
value value Result
s of If digit
>= 0.5 concat
chaoti Rotation Equation R
c then Ri=
maps Ri & 1
3-D Cat Map else Ri= R
Apply
Convert Result to Take the last
Chebyshev map
decimal (D) result of
D times
Chebyshev
Normalization
If B contains exponent then
Convert to eliminate Dot Else If B in floating
binary (B) point format then multiply it by 10
^ the number of digit after
exponent part
Key = the first 8

bit 0f B
FIGURE 1: The Combination of the 3-D chaotic maps.

Algorithm (7): Public and private Key Generation
Input: vector of initial chaotic map values.
Output: Public key, Private key;
Begin
For i=1 to 3
Apply 3-D Logistic map shown in Equation (3);
Apply Rotation Equation shown in Equation (4);
Apply 3-D Arnold Cat map shown in Equation (5);
Convert the result of the three maps to binary using threshold = 0.5;
Bit= concat the result of the three maps;
D= convert Bit to decimal;
For j= 1 to D
Apply first order Chebyshev map shown in Equation (1);
End For
B=take the last result of Chebyshev map and convert it to binary;
Normalize B using Algorithm (6);
If i=1 Then
S=take only first 8 bit of B and convert it to decimal;
Else
If i=2 Then
R= Take only first 8 bit of B and convert it to decimal;
Else
T= Take 30 bit from B and convert it to decimal;
X= convert T to real value between [0,1] using Fraction_code;
End If
81
(ICoIT'17)
10thof April, 2017.
End For
Compute Rotation Equation (RRs(x));
Public key=(x,RRs);
Privet key=s;
End.
Algorithm (8): Public Key Encryption
Input: Message M, Initial values of Rotation and Chebyshev.
Output: Cipher.
Begin
M= select a message and convert it to ranges between [0,1] using Fraction-code Algorithm;
Compute Lr(x) by using Chebcheve map shown in Equetion (1);
Compute RRrs(x) using Rotation Equation shown in Equation (4);
X=M.RRrs(x);
Cipher= (Lr(x),X);
End For
End.
Algorithm (9): Public Key Decryption.
Input: Cipher.
Output: Palin.
Begin
Compute Rotation Equation RR(sr);
M= X/ RR(sr);
Plain= Inverse Fraction_Code;
End.
Algorithm (10): Inverse Fraction_code
Input: Fraction _code as array of ranges between 0 and 1;
Output: integer values between 1 and 256;
Begin
For i=0 to 255
Take the real value and find the equivalent integer value in Fraction_code array;
Put the rult of above step in Inverse Fract_array[i];
End For
Inverse Fraction_code =Inverse Fract_array;
End.
5. EXPERIMENTAL RESULTS
In this section the proposed algorithm is tested. Those tests includes testing the key generated
from the proposed algorithm. The five basic statistical tests and NIST tests are used to
examine the randomness of the generated key
The basic tests are used to determine whether the sequence possess specific characteristics
of truly random numbers. However, if the sequence passes all tests there is no guarantee that
it was produced by a random bit generator [16].The generated keys from the proposed
algorithm are examined using those tests. The results show that the generated bit streams
have the characteristics of truly random numbers as shown in Table (1). In this table, the
different length of keys are statistically tested and pass most of tests.
82
(ICoIT'17)
10thof April, 2017.
TABLE 1: The results of Statistical tests.
Key Value X2 Chi Square

Test Name
Size (X) Degree Value (X) Result
Frequency 0.5 0.05 1 3.8415 Pass
Serial 0.88817 0.05 2 5.9915 Pass
32
Poker 0.5 0.05 1 5.9915 Pass
Run 0 0.05 -2 3.8415 Pass
Frequency 0.25 0.05 1 3.8415 Pass
Serial 1.0675 0.05 2 5.9915 Pass
64
Poker 11.25 0.05 3 9.4877 Not Pass
Run 2.125 0.05 0 5.9915 Pass
Frequency 2 0.05 1 3.8415 Pass
Serial 2.4331 0.05 2 5.9915 Pass
128
Poker 7.5238 0.05 7 15.5073 Pass
Run 2.8101 0.05 2 9.4877 Pass
Frequency 8.2656 0.05 1 3.8415 Not Pass
Serial 7.7148 0.05 2 5.9915 Not Pass
256
Poker 13.5411 0.05 7 15.5073 Pass
Run 7.4057 0.05 2 15.5073 Pass
NIST tests consist of sixteen tests that are used to test the randomness of sequences. The
SUCCESS means that the sequence is acceptable and it has enough randomness, while the
FAILURE means that the sequence is not acceptable because of non-randomness.There are
two tests which are Random Excursion, and Random Excursion Variant that do not give
results each running because they only give results when the number of cycles greater than
500 as shown in Equation (7) [17].
j< max (0.005n,500) (7)
Where J indicates the total number of the cycles in the sequence. Another test which is Linear
Complexity test is also not indicates P-value because some bits are discarded. Table (2)
shows the result of the NIST tests. The generated sequence from the proposed algorithm
passes most of those tests.
TABLE 2: The results of NIST tests.
Number Number Number of Percentage of

Test Name
of tests of Success Fail Success tests
Approximate Entropy 181 181 0 100%
Block Frequency 181 181 0 100%
Cumulative Sum (Forward) 362 362 0 100%
Fast Fourier Transform 181 181 0 100%
Frequency 181 180 1 99%
Lempel Ziv Compression 181 181 0 100%
Longest Run of Ones 181 181 0 100%
Non Periodic Template 26788 21429 5359 79%
Overlapping Template of all Ones 181 181 0 100%
83
(ICoIT'17)
10thof April, 2017.
Rank 181 181 0 100%
Run 181 178 3 98%
Serial 362 357 5 98%
Universal 0 0 0 Not number
6. CONCLUSIONS
From the presented study it has been concluded that, using hybrid chaotic maps in public key
encryption instead of one map strengths the system and makes it more robust against the
attackers. Also, the results of hybrid maps have the characteristics of truly random numbers
because they pass most of statistical and NIST test which makes it difficult for the intruder
to crack the original data.
References
[1] Nicholas G. McDonald, "Past, Present, and Future Methods of Cryptography and Data
Encryption", A Research Review, 2010.
[2] William Stallings, "Cryptography and Network Security Principles and Practice", Fifth Edition,
Prentice Hall, 2011.
[3] Adel A., Amr H. , Hany H., "Survey Report on Chaos Based Public-key Cryptosystem" ,
International Journal of Emerging Technology and Advanced (ISSN 2250-2459, ISO 9001:2008
Certified Journal, Volume 3, Issue 12, December 2013).
[4] David Levy, "Chaos Theory and Strategy: Theory, Application. And Managerial Implications",
Massachusetts Strategic Management Journal, Vol. 15, 167-178, U.S.A (1994).
[5] L. Douglas Kiel, Euel W. Elliott, "Chaos Theory in the Social Sciences: Foundations and
Applications", The University of Michigan Press, 1996.
[6] tienne Ghys, "The Butterfly Effect", 12th International Congress on Mathematical Education,
2012.
[7] Kharel, Rupak, "Design and Implementation of Secure Chaotic Communication Systems",
Doctoral thesis, Northumbria University, 2011.
[8] Ljupco Kocarev, Marjan Sterjev, Attila Fekete, Gabor Vattay, "Public-key encryption with
chaos", American Institute of Physics, Vol. 14, No. 4, 1078-1082, 2004.
[9] Shuichi Aono, Yoshifumi Nishio, "A Cryptosystem Based on Iterations of Chaotic Map", IEICE
Technical Report, Vol.107, No.87, 2007.
[10] K. Prasadh, K. Ramar, R. Gnana jeyaraman, "Public key cryptosystems based on chaotic
Chebyshev polynomials", Journal of Engineering and Technology Research, Vol.1 (7), 122-128,
2009.
[11] Pawan N. Khade, and Manish Narnaware, "3D Chaotic Functions for Image Encryption",
IJCSI International Journal of Computer Science Issues, Vol. 9, Issue 3, No 1, 2012 ISSN 1694-
0814.
[12] Skiadas, "Mathematical models of Chaos. In Chaos Applications in Telecommunications",
Stavroulakis, P., Ed.; CRC Press: Boca Raton, FL, USA, 2006; pp. 383413.
[13] Skiadas, " In Chaotic Modelling and Simulation: Analysis of Chaotic Models, Attractors and
Forms", CRC Press: Boca Raton, FL, USA, 2008.
[14] Daniel-Ioan Curiac and Constantin Volosencu, "Path Planning Algorithm based on Arnold
Cat Map for Surveillance UAVs", Defence Science Journal, Vol. 65, No. 6, November 2015, pp.
483-488, DOI : 10.14429/dsj.65.8483 ,2015, DESIDOC
[15] Tang, W.K. & Liu, Y., "Formation of High-Dimensional Chaotic Maps and Their Uses in
Cryptography. Chaos Based Cryptography", Springer Berlin Heidelberg, 2011, pp. 99-136.
84
(ICoIT'17)
10thof April, 2017.
[16] Mohammed Saleh, "Proposed Block Cipher Algorithm with Cloud Computing Based on Key
Generation", M.Sc. Thesis, University of Technology, 2013.
[17] Maytham M. Hammood, Kenji Yoshigoe and Ali M. Sagheer, "Enhancing Security and
Speed of RC4". International Journal of Computing and Network Technology ISSN 2210-1519
Int. J. Com. Net. Teach. 3, No. 2 (May 2015).
85
(ICoIT'17)
10thof April, 2017.

Ahssan Ahmed Mohammed
Department of software, College IT, University of Babylon - Iraq
ahssan_ahsan@yahoo.com
Dr.Abdulkareem O. Ibadi
Department Computer Engineering, Baghdad College for Economic Sciences University - Iraq
dr.ibadi@yahoo.com
ARTICALE INFO ABSTRACT

In this work, A proposed block cipher algorithm is designed
Article History: with multiple of 32 bits as a length of block performed in
Received: 13 March 2017 heterogonous multiple block cipher dependent on the
Accepted: 1 April 2017 application field , Multiple of 48 bits as a key length are
Published: 10 April 2017 entered automatically used in variable permutation, addition
function and xor operation. A proposed algorithm performed
Keywords: Encryption, in single round to giving fast and high security algorithm used
Decryption, Block in different Purposes. Another proposed functions for
cipher, Keys, SubKey,
different purpose such as Balance Function between balance
Map Function, Wave
Function, Permutation. fixed number and initial permutation and inverse initial
permutation, Maps function as lookup table
(map2,map3,map4 and map5). Wave Function as a nonlinear
function of 8-bits of input and same for the output. This
algorithm can be designed in several blocks length depends
on the application field, increase speed, higher complexity
and for reduction of cost.
.
1. INTRODUCTION
E
ncryption Operation is the process of converting a original data called plaintext
into ciphertext using secret key, and can be convert back into the original data
with the same secret key. There are several kinds of data encryption schemes
which form the fundamental of security. [1]
Encryption Operation schemes are generally based on either block cipher or stream cipher.
Historically for providing confidentiality should be Encryption focus on symmetric
encryption. It is only in the last several decades that other considerations, such as Integrity,
Authentication and Digital Signature have been included in cryptology theory and it practice.
The security of the data basically depends on two laborer first one is Confidentiality and
second one is Authentication. Achieving confidentiality means the data is encrypting bulk
digital data used block cipher. Single round of encryption offers inadequate security but
multiple rounds offer increasing security but the side effect is time consumed not always, it
is dependent on the strength and security of encryption functions used .[2]
86
(ICoIT'17)
10thof April, 2017.
The proposed algorithm is different of another by used block length multiple of 32

bits as a homogenous block cipher (4-8-12-16 bytes) with multiple of 48 bits of key and 4
nonlinearity Maps(Map2, Map3,Map4 Map5) all of them processing in one round and new
subkey generation algorithm also been designed new function for generated permutation
array used as initial permutation and inverse initial permutation then reversible variable
permutation dependent on secret key value as well Wave function is a nonlinear function.
Cryptography is one of many aspects of security, It is science and art of writing.
Egyptian Hieroglyphs in an inscription is first used of cryptography in 1900 B.C. The
primary functions of cryptography are confidentiality, Privacy, Authentication, Integrity,
Non-repudiation and Key exchange.[3]
The three types of Cryptography algorithms are :
4- Secret Key Cryptography or symmetric encryption : Used a single key for both
encryption and decryption, Primarily used for confidentiality and privacy.
5- Public Key Cryptography or asymmetric encryption: Used two keys, one of them
for encryption and another for decryption, mostly used for key exchange, non-
repudiation and authentication .
6- Hash Function: Used a mathematical transformation to encrypt of information,
providing a digital fingerprint. mostly used for message integrity.
Secret key cryptography algorithms in general categorized as being either stream
cipher or block cipher algorithms will focus on this aspect in our proposal. Stream ciphers
operated on a one bit at a time and key is continuously changed, means the Encryption
Operation of each digit depends on the current state of the cipher engine, Stream cipher is
also call state cipher. Generally, single bit/bites are used as single digits to avoid concerns of
security . Error propagation is a problem of the Stream cipher as a bit corrupted in
transmission will result of corrupted set of bits at the receiving side.[4]
Block Cipher Encrypts/Decrypts one block from data at a time used the same key on
each block with fixed block/key length. Encryption key scheme is determines the order in
which Transportation, Substitution and other mathematical functions are applied on each
block. Block cipher algorithms used two operations are Confusion and Diffusion to encrypt
Plaintext Block into Ciphertext Block. The aim of confusion is to make the relationship
among the encryption key and the ciphertext as complex as possible. Ideally, every bit in the
key should influence in every bit of the Ciphertext Block. By contrast, Diffusion operation
propagates the influence of each bit in the Plaintext Block over several bites in the Ciphertext
Block, making the cipher few oversensitive to statistical attacks.[6]
The actuality that stream cipher data Encryption/Decryption single bit at a time means
that they are particularly well appropriate to real time hardware applications, like image ,
video and audio applications. Stream cipher algorithm is practical weaker and less efficient
than block cipher algorithm for software applications and it minimal frequently used in that
sphere. Block cipher algorithm is easier to implement in Software application because the
message encrypted in blocks length that software already uses. The Encryption key is
predominantly the same size as the block length. Hence block ciphers achieve higher
87
(ICoIT'17)
10thof April, 2017.
diffusion and error propagation than stream ciphers, although it is worth noting that the
algorithm determines the amount of diffusion.[4][5]
Block ciphers algorithms usually require large memory, because they perform on larger
data chunks and predominating have carry over from preceding blocks, whilst Stream ciphers
algorithms perform on only a fewest bits or byte at a time they have relatively less memory
requirements and therefore cheaper to implement in limited scenarios like firmware,
embedded devices and special hardware.[5]
Stream ciphers algorithms are more difficult to perform appropriate, and apt to
weaknesses based on usage, since the fundaments are similar to one time pad, the key stream
cipher has more stringent requirements.[4]
On the other hand, Block ciphers algorithms encrypt a whole block at a time and
moreover have feedback modes which are most recommended, they are over sensitive to
noise in transmission, means if you mess up one data part, all the remnant is maybe
unrecoverable. while stream ciphers algorithms bytes are individually encrypted with no
connection to other data chunks in most ciphers/modes, and typically have support for
interruptions on the line.[5]
Stream ciphers algorithms do not provide security authentication or integrity
protection, but some block ciphers algorithms can provide integrity protection and
confidentiality depending on mode.[4][5]
The reason for choosing Block ciphers in this proposed because more useful when the
data amount is pre known such as a data file, fields, or request protocols.
2. A PROPOSED BLOCK CIPHER

A proposed block cipher algorithm is a based on Non-Feistel Block Cipher algorithms.
The block diagram of it is shown in figure(1) and the algorithm is given in the following
eight steps
Algorithm:
INPUT : block of length 32 bits, key of length 48 bits.
OUTPUT: block of length 32 bits.
8. Compute 32 bits after initial permutation by balance function = P block Xor
balance constant block.
9. Compute addition operation on balance block and subkey1 to yield P sum block
(each 8 bits of balance block addition to a single 8 bits of subkey1).
10. Perform Map Function on P-sum block to convert each 8 bits to covalence bits.
11. Perform variable permutation on P-Map block based on bits location of key.
12. Perform Reversible Wave function on variable permutation block.
13. Shifting and Rotation operations for each 8 bits of wave block.
14. Compute Xor operation on each 8 bits of block with different 8 bits of subkey2.
7. Swap among each 8 bits of block to produce the C block of length 32 bits.
8. Next 32 bits.
88
(ICoIT'17)
10thof April, 2017.
Figure: 1 Illustrate Encryption & Decryption algorithm structure
The main idea of a proposed block cipher algorithm is begin by input block of
plaintext and the output is block of ciphertext after executed in number of different steps.
The algorithm has a multiple of 32-bits block size and multiple of 48 bits key length.
It is a 7-step built based on non Feistel cipher algorithm. The blocks are constructed by
converting a characters into 32 bits by using ASCII and also entered key by converting a 6-
character into 48 bits by using ASCII for each character. The figure 2&3 illustrate encryption
and decryption steps structure.
2.1. BALANCE FUNCTION

Two main functions are operate in Balance function they are : Initial permutation and
XOR function, The input/output of the Balance function is a block of length 32 bits, the input
block permutated in permutation generation algorithm to generate initial permutation (IP) as
in table 1 and inverse permutation (IP-1) as in table 2 for first function then Xor-ed with a
block of balance constant of the same length. The block diagram of it is shown in figure 2
89
(ICoIT'17)
10thof April, 2017.
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
4 11 24 13 9 16 3 17 8 14 2 20 19 0 26 15
16 17 18 19 2021 22 23 24 25 26 27 28 29 30 31
10 7 18 27 1223 6 25 31 28 1 21 30 5 29 22
Table 1 : Initial Permutation IP
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
13 26 10 6 0 29 22 17 8 4 16 1 20 3 9 15
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
5 7 18 12 11 27 31 21 2 23 14 19 25 30 28 24
Table 2 : Inverse Initial Permutation IP-1
2.1.1 PERMUTATION GENERATION ALGORITHM
Input : list of number schedule arranged from 0 to n-1
Output: array of random numbers between 0 to n-1 without repetition
1. Loop from 0 to n-1 .
2. Using predefined random function in vb.net (Random) working storage area.
3.Using list operation to move location of random number generated to the output array
4. Remove the location and reduce list length by one .
5. Next.
Figure 2: Illustrate Balance Function structure
2.2. ADDITION FUNCTION

Addition operation is second function of a proposed block cipher algorithm. The output
block of Balance function will dividing into set of 8-bits each one will addition with single
8-bits of subkey1 block as performed by used permutation and selection for the secret key as
shown in table 3.
Figure 3: Illustrate Addition Function structure
90
(ICoIT'17)
10thof April, 2017.
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
47 31 22 38 29 1 10 45 41 36 16 12 37 28 17 2
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
9 15 32 3 39 19 6 18 7 14 26 8 24 40 34 33
Table 3 : Permutation and Selection of Subkey1
2.3. MAP FUNCTION

Map function is performed for a byte of input block, A proposed block cipher used 4
maps each one has a different number. Different permutation is present after each Map to
help for bits propagation and diffusion increased with high complexity.
2.3.1 Map Function generation
Map generated by converted the input byte into complement number excepted the
bits has location equal to the random value generate between 0 -7. A proposed block cipher
used 4 maps each one with different number of excepted bits (Map2 means two bits has
generated and so on for Map3, Map4 and Map5). Maps function structure is shown in figure
4. Table 4 is example of Maps function.
Input Random Values Output
Map2 10001110 2-4 01011001
Map3 10010011 3-4-7 01110101
Map4 10011000 0-1-6-7 10100100
Map5 10011101 0-1-2-3-4 10011010
Table 4 : Example of Maps Function
Figure 4 : Maps Function structure
91
(ICoIT'17)
10thof April, 2017.
2.4. REVERSIABLE VARIABLE PERMUTATION

Variable permutation is a reversible function, it relying to secret key values. Variable
permutation performed for each byte of block dependent on location of secret key bits. The
variable permutation is implement by three levels as shown in figure 3. Level one check the
secret key and make swapping between two neighboring bits when the read bit is one. Level
two and three perform swapping between widely bits as shown in following structure to
guarantee permute from first to last bit.
Figure 5 : Reversible Variable Permutation(VP) structure
2.5. WAVE FUNCTION

The wave function is a nonlinear function designed to be of the same size inputs
and outputs which help to increase the confusion and diffusion of output result because
whenever a zero is located in a sequence the rest of the sequence will be converted[7]. A
proposed block cipher algorithm used the Wave function with byte of input and same for output.
If X is the input set and Y is the output set so Y can write as Y=WF(x) where,
Y1=X11
Y2=X2X11
Y3= X1X2X31
Yn=X1X2.Xn-1Xn1
Figure 6 : Reversible Wave Function structure
Another feature of the Wave Function is that never results in the same sequence
and the output will be repeated after 2n trial.
92
(ICoIT'17)
10thof April, 2017.
2.6. SR FUNCTION
A Shifting and Rotation function is propagation function for each byte of block. A
proposed block cipher used four shifting and rotation function (S&R1 means shifting each
byte by one and so on for S&R2 , S&R3 and S&R4).
2.7. XF FUNCTION
XF function is xor operation between input block and subkey2, The input of the XF
function is a block of length 32 bits Xor-ed with a block of Subkey2 of the same length
generated by permutation and selection operation for secret key. In the end the output will
swap among each 8 bits and saving the ciphertext block in the binary file as shown in the
main structure. Figure 6 illustrate XF function structure or single 8 bits
Figure 6 : Single XF function structure
2.8. KEY MANAGMENT

A proposed block cipher algorithm used multiple of 48 bits as a secret key by entered
automatically dependent on application field such as password , data, image, audio, video,
satellite communication its. Secret key used in variable permutation to generate permutation
dependent on each bit as discussed in 4.4 . Subkeys are generated by scheduling of secret
key in two operations as selection and permutation . The input of the algorithm is the secret
key (6 character) witch are converted into 48 bits using ASCII code, A subkey1 and subkey2
are generated by selection and permutation of 32 bits from the 48 bits of secret key as shown
in a proposed algorithm structure.
3. EVALUATION
There are two types of block cipher Feistel and Non-Feistel cipher, a feistel cipher
algorithm is a block cipher construction as a symmetric build. This came on the label of a
German inventor Horst Feistel. A Wide proportion of block ciphers use the Feistel network,
such as the (DES) Data Encryption Standard, Blowfish, KASUMI, Simon and etc. The
advantage of the Feistel Network is very similar for the encryption & decryption operations
it is identical in some cases and requiring only moving from up to down ,From top to bottom
or a reversal of the key schedule. Feistel ciphers and similar algorithms are product ciphers,
and repeated operations for number of rounds.[5]
A non Feistel cipher is the sound type of block cipher so a Non-Feistel block cipher
algorithm is a chain of linked mathematical operations performed in block cipher, it is
network with a block of Plaintext and Key as inputs and applies several alternating
operations to produce the ciphertext block such as AES , Square, 3-Way, SHARK , SAFER
, PRESENT, and , Kuznyechik .
Although the Feistel block cipher that used S-box (like Data Encryption Standard) is fully
similar to Non-Feistel block cipher, but there are some differences that make either Feistel
or Non-Feistel more applicable in assured situations. For a given amount of Diffusion and
93
(ICoIT'17)
10thof April, 2017.
Confusion, an Non-Feistel block cipher has extra inherent parallelism[11], and so given a
CPU with a more number of execution units can be computed faster than a Feistel block
cipher.[12] CPUs with less execution units such as most smart cards cannot take advantage
of this inherent parallelism. Also Non-Feistel ciphers require s-box to be invertible (to
perform decryption). Feistel inner functions have no such restriction and can be constructed
as one way functions.
The advantage of a proposed algorithm is that, since its external and internal properties
are change from function to function, it may be extremely difficult to find any kind of
characteristic that propagates well through all of the different kinds of functions that appear
in the cipher. However, implementing of a proposed algorithm and analyzing it is often much
more complicated. Hardware implementations are generally cheaper, and software
implementations smaller and easier to correctly code.
A proposed block cipher algorithm used a balance function, These structure provide
efficient implementations for mutable of the 32 bits of plaintext block by initial permutation,
and used xor-ed operation with fixed balance binary number for solving problem when all
plaintext block (0) or (1).
A Map function is substitutes a byte (the input of map function) by another byte ( Map
Function output). This Substitution should be one byte to one byte, to ensure inevitability for
decryption of a proposed algorithm. The good Function goal is changing one input bit will
change about half of the output bits.
The Wave Function is a reversible nonlinear function which helps to increase the
confusion of the resulting output besides, it helps to increase the diffusion because whenever
a zero is located in a sequence the rest of the sequence will be converted. Another feature is
that never results in the same sequence and the output will be repeated after 2n trial.
Addition operation with subkey1, Variable permutation dependent on key locations
values, Shifting and Rotation operations of byte plaintext block, Xor-ed with subkey2 and
Bytes Swapping operation are helps for increasing alteration among each byte.
Diffusion is important concept in the design of robust of block cipher algorithm, It is
operation between plaintext block and ciphertext block means if changed one bit of plaintext
block then several bits of ciphertext block should be changed and Vice versa. In another
interpretation, This means the frequency statistics of bit in the plaintext block are diffused
over several bits in the ciphertext block and statistical attack need more meaningful to do a
ciphertext block. Table 5,6 shown the range of diffusion between plaintext block and
ciphertext block in the block cipher algorithm proposed.
Block length Key length Diffusion Range

32 48n 12-18
64 48n 27-35
96 48n 40-57
128 48n 61-68
Table 5: shown the range of diffusion with different of block size .
94
(ICoIT'17)
10thof April, 2017.
Key 000000000000000000000000000000000000000000000000 Zero One Diffusion

Plaintext 00000000000000000000000000000000 32 0 -
Ciphertext 00111011100111111111101001011110 10 22
Plain 00000000000000000000000000000001 31 1
Cipher 01011001100010010011000011011100 18 14 12
Plain 00000000000000000000000000000011 30 2
Cipher 00110001100110111011110100011110 14 18 12
Plain 00000000000000000000000000000111 29 3
Cipher 11010010101110101011110011101010 13 19 13
Plain 00000000000000000000000000001111 28 4
Cipher 11011110010011011001001010110110 14 18 17
Plain 00000000000000000000000000011111 27 5
Cipher 00010011011001010010001010110011 18 14 12
Plain 00000000000000000000000000111111 26 6
Cipher 10110100101001001011101000111011 15 17 13
Table 6: shown the range of diffusion with same of block size .
4. CONCLUSION
A proposed block cipher algorithm designed to create a strong algorithm that runs in
a reasonable time on presently available hardware, that is rationally easy to implement,
etc. As such, a proposed algorithm design may allow us to design faster and better block
cipher algorithm than we could if we limited ourselves to number of round or Non-Feistel
network structures.
5- We concluded the relationship between the blocks length and hardware , It can
perform multiple of 32 bits as block length as a homogenous algorithm to reduce the
time of execution
6- A number of rounds should be less to speed increasing.
7- Multiple of 48 bits as a secret key should be entered automatically and it dependent
on application field such as password , data, image, audio, video, satellite
communication its.
8- Permutation operation should be after each map function to propagate bits in each
block and increase confusion and diffusion operations.
5. REFERENCES
References must be listed in the order they were cited (numerical order). The references
must not be in alphabetical order eg:
18- Suzaki, T., Minematsu, K., Morioka, S., Kobayashi, E.: Twine: A lightweight block
cipher for multiple platforms. In Selected Areas in Cryptography - SAC 2012. LNCS,
vol. 7707, pp. 339-354. Springer, Heidelberg (2013).
95
(ICoIT'17)
10thof April, 2017.
19- Wu, W., Zhang, L.: Lblock: A lightweight block cipher. In Applied Cryptography
and Network Security - ACNS 2011. LNCS vol. 6715, pp. 327-344. Springer,
Heidelberg (2011).
20- van Tilborg, Henk C. A.; Jajodia, Sushil, eds. (2011). Encyclopedia of Cryptography
and Security. Springer. ISBN 978-1-4419-5905-8., p. 455.
21- Junod, Pascal & Canteaut, Anne (2011). Advanced Linear Cryptanalysis of Block
and Stream Ciphers. IOS Press. p. 2. ISBN 9781607508441.
22- Chakraborty, D. & Rodriguez-Henriquez F. (2014). "Block Cipher Modes of
Operation from a Hardware Implementation Perspective". In Ko, etin
K. Cryptographic Engineering. Springer. p. 321. ISBN 9780387718163.
23- Lars R. Knudsen and John Erik Mathiassen, On the Role of Key Schedules in Attacks
on Iterated Ciphers, ESORICS 2014, pp322334
24- Dr.Abdulkareem O. Ibadi, Special Security e-Mail System , Ph.D. dissertation,
University of Technology , June 2007
25- Ben Morris, Phillip Rogaway, Till Stegers. "How to Encipher Messages on a Small
Domain". CRYPTO 2010.
26- Menezes, Alfred J., Oorschot, Paul C. van; Vanstone, Scott A. (2011). Handbook of
Applied Cryptography (Fifth ed.). p. 251.
27- Christof Paar, Jan Pelzl, "The Data Encryption Standard (DES) and Alternatives",
free online lectures on Chapter 3 of "Understanding Cryptography, A Textbook for
Students and Practitioners". Springer, 2012.
28- "Principles and Performance of Cryptographic Algorithms" by Bart Preneel, Vincent
Rijmen, and Antoon Bosselaers.
29- "The Skein Hash Function Family" 2010 by Niels Ferguson, Stefan Lucks, Bruce
Schneier, Doug Whiting, Mihir Bellare, Tadayoshi Kohno, Jon Callas, Jesse Walker
page 40.
30- Sinha, S., Arya, C. (2012). Algebraic Construction and Cryptographic Properties of
Rijndael Substitution Box. Defence Science Journal, 62(1).
31- Paar, C., Pelzl, J. (2010). Understanding Cryptography: A Textbook for Students and
Practitioners. New York City, New York: Springer Publishing.
32- Grocholewska-Czurylo, A. (2011). Cryptographic properties of modified AES-like
S-boxes.
33- Annales UMCS Informatica, AI XI(2), 37-48. DOI: 10.2478/v10065-011-0009-4.
34- Easttom, C. (2015). Modern Cryptography. New York City, New York: McGraw
Hill.
96
(ICoIT'17)
10thof April, 2017.
A Survey On Unsupervised Evaluation Criteria For Image

Clustering Validation
Akar Taher
Department of Software Engineering, Faculty of Engineering, Koya University University Park,
Kurdistan Region of Iraq.
Akar.taher@koyauniversity.org

The evaluation of clustering results is the most difficult and
Article History: frustrating part of cluster analysis. The challenge is to validate
Received: 13 March 2017 the obtained results without any apriori information. Validity
indexes are widely used approach for evaluation of clustering
Accepted: 1 April 2017 results. These approaches can use three criteria: i) external
Published: 10 April 2017 (also called supervised) criteria: this type is based on
comparing the obtained results with a previously known
result (frequently called ground truth) and compute the
Keywords: Evaluation, similarity, ii) internal criteria (also called unsupervised)
Criteria, Validation criteria: estimate the quality of the result using internal
Unsupervised, information of the data alone, and iii) relative criteria: this
Supervised, Clustering. means multiple usages of one of the two above types
ofdifferent results and see which is better than the other.
Therefore we can say: depending on the information available
and the problem type, different types of indexes might be used
for cluster validation. Sometimes due to the complexity of the
datasets, one validity index is not sufficient to evaluate the
quality of the obtained results, and then a combination of two
or more index should be used. In this paper, a basic general
review on evaluation criteria is first given and then the focus
is spotted on unsupervised criteria as they are much more
useful, thanks to their objective functionality.
1. INTRODUCTION
T
he evaluation of clustering [1][3] results is an unavoidable process [4][7]
which is used to quantify the performance of clustering algorithms. The quality
evaluation of a clustering result is an active area of research and many criteria
are being developed regularly. Unfortunately, the evaluation of a clustering
result always contains some elements of subjectivity and the criteria do not always give
satisfactory evaluation [8]. For this reason, it is impossible to define a universal criterion to
evaluate the results produced by all the existing criteria. However, a number of criteria exist
and are repeatedly used by many researchers to compare clustering results [9], [10]. Since
there are a large number of possible clustering results for the same dataset, the objective is
97
(ICoIT'17)
10thof April, 2017.
to assess whether any of these results is better than another. So to correctly evaluate and
validate the results, it can be necessary to use multiple evaluation criteria like in [11].
In this paper, the generalities of the evaluation criteria are provided, and then focus is on
the unsupervised ones.
This following of this paper is organized as: the second section will describe evaluation
criteria types, then followed by unsupervised evaluation criteria in the third section, and
fourth section will give the conclusions.
2. EVALUATION CRITERIA TYPES

Several types of evaluation methods have been proposed in the literature [7], [12]. They
are classified into three main groups. The first group contains unsupervised criteria that use
only internal information of the data such as the distance between objects. These criteria are
also called internal quality measures. The second group contains supervised criteria that
calculate the degree of correspondence between the clustering produced by the algorithm and
a known data partitioning. These criteria are also known as external quality measures. The
last group is called relative criteria; this type of evaluation allows comparing the results
obtained from the same algorithm. These measures are simply the use of internal or external
criteria to evaluate multiple results produced by the same algorithm and to choose the best
one among them. In this paper only internal quality criteria are reviewed.
3. UNSUPERVISED EVALUATION CRITERIA

Unsupervised evaluation criteria [9] are based on internal information and do not need
any a priori knowledge. This type of criteria generally computes statistical measures such as
the standard deviation or the disparity of the classes. These measures are often based on the
simplest definition of partitioning which says that objects from the same class should be as
close as possible, and that objects from two distinct classes should be as far apart as possible
[13] (see Figure 1) . To assess whether a clustering result complies with this intuitive
definition, the distances between the class centers and the class objects are calculated. These
unsupervised measures assess the compactness and the separateness of the classes. The
evaluation of the quality of a cluster is not formally defined, so there are many different
criteria, which estimate the quality of the results differently. Some of these criteria can be
directly used as the objective function of a clustering algorithm. However others are very
time-consuming, and therefore intended to be calculated after the application of the algorithm
for the final evaluation process.
One of the most basic and intuitive criteria able to quantify the quality of a clustering
result is the within-class uniformity. The simplest way to calculate this uniformity is the sum
of the squared errors (SSE) which is calculated as follows in Equation (1):
98
(ICoIT'17)
10thof April, 2017.
Within-class disparity measure
Between-class disparity measure
Class 1 members
Class 2 members
Class 1 center Class 1

Class 2 center
Class 2
Figure 1: illustration of within-class homogeneity and the between-class disparity
NC
SSE I R d x g (Ci )
2
(1)
i 1 xci
where g (Ci ) is the center of the class Ci and d is a distance measure.
Weszka and Rosenfeld [14] proposed such a criterion with thresholding that measures
the effect of noise to evaluate some thresholded images. Based on the same idea of within-
class uniformity, Levine and Nazif [15] also defined a criterion that calculates the uniformity
of a class as follows:
2
1 NC xCi gl x xCi gl x
LEV1 I R 1 (2)
maxxCi gl x minxCi gl x
2
N i 1
where
I R is the partitioning result of the image I into NC classes C C1,..., CNC ,
N is the number of pixels of the image I,
gl x is the gray level of pixel x in the image I.
A standardized uniformity measure was proposed by Sezgin and Sankur [16] that is based
on the Cochran homogeneity measurement [17]. However, this method requires a threshold
that is often arbitrarily selected, thus limiting the usage of this criterion. Another criterion to
measure the within-class uniformity was developed by Pal and Pal [18]. It is based on a
thresholding that maximizes the local entropy of the classes in a partitioning result. In the
case of slightly textured images, these criteria of within-class uniformity prove to be effective
and very simple to use. However, the presence of textures in an image often generates
improper results due to the over-influence of small regions.
Complementary to the within-class uniformity, Levine and Nazif [15] defined a disparity
measure between two classes to evaluate the dissimilarity of different classes in a partitioning
result. The formula of total between-class disparity is defined as follows:
99
(ICoIT'17)
10thof April, 2017.
LEV 2 I R

NC
w
k 1 Ck
NC
j 1/ C j W Ck
Ck \C j gl Ck gl C j / ( gl Ck gl C j
p
(3)
k 1wRk
NC
where wck is a weight associated to Ck that can be dependent of its area, gl Ck is the average
of the gray level of Ck and pCk \C j is the length of the boundary of the class Ck common to
the perimeter of the class C j . This type of criterion has the advantage of penalizing over-
segmentation.
Zeboudj [19] proposed a measure based on the combined principles of maximum
between-class (external) disparity and minimal within (interior) class disparity measured at
the pixels neighborhood.
gl x gl z
Let c x, z be the disparity between two pixels x and z x and z Ci , and
L 1
L be the maximum gray level.
The interior disparity CI Ci of the class Ci is defined as follows:
max c x, z , z V , z C
1
CI Ci s i (4)
NCi xCi
where NCi is the number of pixels in class Ci and Vs is the neighborhood of the pixel x.
The external disparity CE Ci of the class Ci is defined as follows:
max c x, z , z V , z C
1
CE Ci s i (5)
pi xCi
where pi is the length of the boundary of class Ci .

Lastly, the disparity of the class Ci is defined by the measurement D(Ci ) [0, 1]
expressed as follows:
CI Ci
1 if 0 CI Ci CE Ci
CE Ci

D(Ci ) CE Ci if CI Ci 0 (6)

0 otherwise
Zeboudjs criterion is defined by:

1 NC
ZEB I R NCi xD Ci
N i 1
(7)
where N is the number of pixels in the image.
100
(ICoIT'17)
10thof April, 2017.
This criterion has the disadvantage of not correctly taking into account strongly textured
regions.
Another criterion that is based on the combination of the within-class and between-class
disparities is the Davies-Bouldin index [20]. It estimates the within-class disparity based on
the distance from the points in a class to its centroid and the between-class disparity based
on the distance between centroids. It is defined as:

NCi NC j
1 NC
1
NCi d
k 1
Fk , g Ci NC d Fk , g C j
1
j
k 1

DB I R max

(8)
NC i 1 j , j i d g Ci , g C j

where Fk is vector of Nf features representing the pixel xk.

Another criterion of this type is the Silhouette index [21]. This index is a normalized
summation-type index. The within-class is measured based on the distance between all the
points in the same cluster and the separation is based on the nearest neighbor distance.

Let d1 x j be the average dissimilarity of x j with all other pixels of its class Ci d1 x j
indicates how well x j is assigned to its class (the smaller the value, the better the
assignment).

Let d2 x j be the lowest average dissimilarity of x j to any other class
Cl with l 1,2,..., K; l i .
The class with the lowest average dissimilarity is said to be the "neighboring cluster" of
x j because it is the next best-fit class for it; and then the size of the silhouette Sil ( x j ) is
defined as:
d2 x j d1 x j
Sil ( x j )
max d2 x j , d1 x j
(9)
Basing on the definition of Sil ( x j ) , the silhouette of the class Ci is defined as:
Sil ( x )
1
Sil (Ci ) j (10)
NCi x j Ci
Finally the global silhouette for a partition is defined as:

1 NC
Sil I R sil (Ci )
NC i 1
(11)
This criterion is very efficient but its time complexity makes it inapplicable to large
datasets.
101
(ICoIT'17)
10thof April, 2017.
The Dunns index (Du) [22] is another unsupervised criterion that measures the
compactness of a class and the separateness between classes as follows:
Du I R
min
i 1:NC min d g(C ), g(C )
j 1:NC ,i i
ij i j
(12)
max dii ( g (Ci ), g (Ci )
i 1:NC

where dij g (Ci ), g (C j ) is the distance between the center of classes Ci and C j , which is
defined here as the minimum distance between the objects of different classes (see Equation
(13)). dii ( g (Ci ), g (Ci )) is the maximum distance between two objects in the same class (see
Equation (14)).
dij g (Ci ), g (C j ) min d ( x, y) (13)

xCi , yC j
dii g (Ci ), g (Ci ) max d ( x, y) (14)

x , yCi
This evaluation criterion has two disadvantages: firstly, it is very time consuming and
secondly it is highly affected by the presence of noise in the dataset.
In [23], [24] Rosenberger and Chehdi presented a criterion that enables estimating the
within-class homogeneity and the between-class disparity considering the types of regions
(textured or non-textured)1 in the partitioning result. This criterion quantifies the quality of a
partitioning result as follows:
1 D IR D IR
ROS I R (15)
2
The global within-class disparity D I R quantifies the homogeneity of each class
obtained in the partitioning result IR of image I. On the other hand, the global between-class
disparity D I R quantifies how well the classes obtained are separated from each other.
The global within-class disparity D I R reflects the statistical stability of each class. It is
calculated from the within-class disparity D Ci of the different classes in a partitioned
image:
1 NC NCi
D IR D Ci
NC i 1 N
(16)
The weight of the within-class disparity of a class Ci in the global within-class disparity
is proportional to the number of pixels for this class. The same principle is used to calculate
the between-class disparity D I R of the partitioned image I R that measures the disparity
of each class with the other classes:
1 For further reading about region type in an image please refer to [25].
102
(ICoIT'17)
10thof April, 2017.
1 NC NCi
D IR D Ci
NC i 1 N
(17)
This criterion is calculated using the between-class disparity D Ci and the with-in class
disparity D Ci of each class Ci . The calculation of these two criteria is detailed in the
following:
Within-class disparity criterion
This criterion evaluates the homogeneity of a class, i.e. the variation of the statistics in the
interior of this class. In the calculation of the within-class disparity, the nature of the regions
(i.e. textured and non-textured) is taken into account.
In the non-textured case, this criterion for class Ci is defined as:
2
1 1
D Ci gl x
2
g x (18)
NCi2 xCi
l
NCi xCi
This criterion is sufficient to characterize the within-class disparity of a non-textured
region. However, in the textured case, each class is characterized by a set of texture feature
vectors. The dispersion of this set of vectors allows calculating the within-class disparity in
the textured case.
Between-class disparity criterion
The evaluation process of between-class disparity of a class is similar to the with-in class
disparity, but instead of estimating the homogeneity of a class, it is disparity with the other
classes is calculated. The between-class disparity is also calculated according to the nature
of the regions as follows:
Between classes of the same region type:
o The disparity between two classes belonging to uniform regions D Ci , C j is
defined as:
gl Ci gl C j
D Ci , C j (19)
NG
where NG is the number of the gray levels in the image
o The disparity between two classes belonging to textured regions D Ci , C j is
defined as:
D Ci , C j

d g ci , g c j (20)
g ci g c j
where d(.,.) is the Euclidean distance, g(Ci) is the centroid of class Ci , and . denotes the
Euclidean norm.
Between classes of different region types: the disparity between classes of different
region types is set as the maximum value, i.e. 1.
Error! Reference source not found. gives a summery of the previously cited methods in
his section.
103
(ICoIT'17)
10thof April, 2017.
Table 1: Summary of main internal (unsupervised) evaluation criteria
Evaluation Criteria Remarks

Sum of squared errors Measures within-class disparity.
Levine and Nazif (LEV1) [15] Measures within-class disparity.
Levine and Nazif (LEV2) [15] Measures within class and between class disparities.
Zeboudj index [19] Measures within class and between class disparities.
Measures within class and between classes, time
Davies-Bouldin index [20]
consuming.
Measures with-in class and between classes, very time
Silhouette index [21]
consuming.
Measures with-in class and between class disparities, not
Dunn index [22]
effective in case of noisy images.
Measures with-in class and between class disparities, takes
Rosenberger and Chehdi [23], [24]
into account the region type (textured and non textured)
The evaluation criteria described in the previous section are tested on an image database
containing 100 images. These images are clustered by the well-known k-
means, and fuzzy C-means [26] algorithms. As these algorithms are unstable
algorithm, the algorithms are run 100 times on the same image and then the
result that has the best objective function value of that algorithm is selected. In
this database the images are composed of five regions (extracted from the
Brodatz album [27]). The number of classes k is set to 5, as the images contain
5 regions for both algorithms, and the fuzzification factor m is set to 4 to get
more stable results [28].
Figure 2 shows two examples of one synthetic image of this database. The
unsupervised evaluations are compared to Overall Correct Classification Rate
(OCCR), which compares the clustering results with the ground truth (GT) of the
image that is calculated as below:
No.of pixelsclassifiedcorrectly
OCCR = (21)
TotalNo.of Pixels
According to the results, the Rosenberger and Chehdi index is the best index that in the
88% of the cases the index scores high values and high OCCR. The Silhouette index is the
second best index with a score of 79%. Davies-Bouldin and Dunn gave 74% and 72%
respectively. The remaining Zeboudj and the 2 indices of Levine and Nazif gave scores below
45%.
104
(ICoIT'17)
10thof April, 2017.
(a) (b)
Figure 2: Examples of images used for performance comparison:

(a) Original image 1 and (b) Original image 2.
5. CONCLUSION
The evaluation of clustering results is inevitable to assess the quality of the results
obtained.
In this paper, various unsupervised evaluation criteria used to assess the quality of a
clustering result is presented. These criteria are also called internal criteria because they do
not use any external information in the evaluation process. Each of these criteria are effective
in specific cases, for example in image datasets, some are effective the case of non-textured
or slightly textured images, while others give effective results in the case of textured images.
None of the evaluation methods can prove satisfactory in all the cases. Therefore, to correctly
evaluate the algorithms and their results; more than one evaluation technique should be used
and their results should be combined.
6. ACKNOWLEDGEMENT
Many thanks goes to the team of TSI2M in the engineering school of ENSSAT in names
(Prof. Dr. Kacem Chedi and Asst. Dr. Claude Cariou), Rennes 1 University, France, as this
paper is a part of my work in corporation with them.
7. REFERENCES
[1] R. Xu and I. Wunsch, D., Survey of clustering algorithms, IEEE Trans. Neural Netw.,
vol. 16, no. 3, pp. 645678, 2005.
[2] C. C. Aggarwal and C. K. Reddy, Data Clustering: Algorithms and Applications. CRC
Press, 2013.
[3] A. K. Jain, M. N. Murty, and P. J. Flynn, Data Clustering: A Review, ACM Comput
Surv, vol. 31, no. 3, pp. 264323, Sep. 1999.
[4] A. Biswas and B. Biswas, Defining quality metrics for graph clustering evaluation,
Expert Syst. Appl., vol. 71, pp. 117, Apr. 2017.
[5] F. Zaidi, D. Archambault, and G. Melanon, Evaluating the Quality of Clustering
Algorithms Using Cluster Path Lengths, in Advances in Data Mining. Applications and
Theoretical Aspects, 2010, pp. 4256.
105
(ICoIT'17)
10thof April, 2017.
[6] M. Halkidi, Y. Batistakis, and M. Vazirgiannis, Cluster Validity Methods: Part I,

SIGMOD Rec, vol. 31, no. 2, pp. 4045, Jun. 2002.
[7] M. Halkidi, Y. Batistakis, and M. Vazirgiannis, Clustering Validity Checking Methods:
Part II, SIGMOD Rec, vol. 31, no. 3, pp. 1927, Sep. 2002.
[8] W. M. Rand, Objective Criteria for the Evaluation of Clustering Methods, J. Am. Stat.
Assoc., vol. 66, no. 336, pp. 846850, 2012.
[9] M. K. Pakhira, S. Bandyopadhyay, and U. Maulik, Validity index for crisp and fuzzy
clusters, Pattern Recognit., vol. 37, no. 3, pp. 487501, Mar. 2004.
[10] O. Arbelaitz, I. Gurrutxaga, J. Muguerza, J. M. Prez, and I. Perona, An extensive
comparative study of cluster validity indices, Pattern Recognit., vol. 46, no. 1, pp. 243
256, Jan. 2013.
[11] K. Chehdi, A. Taher, and C. Cariou, Stable and unsupervised fuzzy C-means method
and its validation in the context of multicomponent images, J. Electron. Imaging, vol.
24, no. 6, p. 061117, Dec. 2015.
[12] A. K. Jain and R. C. Dubes, Algorithms for Clustering Data. Upper Saddle River, NJ,
USA: Prentice-Hall, Inc., 1988.
[13] R.M. Haralick and L. G. Shaprio, Image Segmentation Techniques, presented at the
Computer Vision Graphics and Image Processing, Arlington, 1985, vol. 29, pp. 100
132.
[14] J. S. Weszka and A. Rosenfeld, Threshold Evaluation Techniques, IEEE Trans. Syst.
Man Cybern., vol. 8, no. 8, pp. 622629, Aug. 1978.
[15] M. D. Levine and A. M. Nazif, Dynamic Measurement of Computer Generated Image
Segmentations, IEEE Trans. Pattern Anal. Mach. Intell., vol. PAMI-7, no. 2, pp. 155
164, Mar. 1985.
[16] B. S. Mehmet Sezgin, Survey over image thresholding techniques and quantitative
performance evaluation, J. Electron. Imaging, vol. 13, pp. 146168, 2004.
[17] W. G. Cochran, Some Methods for Strengthening the Common X 2 Tests, Biometrics,
vol. 10, no. 4, p. 417, Dec. 1954.
[18] N. R. Pal and S. K. Pal, Entropic thresholding, Signal Process., vol. 16, no. 2, pp. 97
108, Feb. 1989.
[19] R. Zboudj, Filtrage, seuillage automatique, contraste et contours: du pr-traitement
lanalyse dimage. Saint-Etienne, 1988.
[20] D. L. Davies and D. W. Bouldin, A Cluster Separation Measure, IEEE Trans. Pattern
Anal. Mach. Intell., vol. PAMI-1, no. 2, pp. 224227, Apr. 1979.
[21] P. J. Rousseeuw, Silhouettes: A graphical aid to the interpretation and validation of
cluster analysis, J. Comput. Appl. Math., vol. 20, pp. 5365, Nov. 1987.
[22] J. C. Dunn, Well-Separated Clusters and Optimal Fuzzy Partitions, J. Cybern., vol.
4, no. 1, pp. 95104, Jan. 1974.
[23] C. Rosenberger, Mise en oeuvre dun systeme adaptatif de segmentation dimages,
Phd Thesis, University of Rennes 1, 1999.
[24] C. Rosenberger, K. Chehdi, and C. Kermad, Adaptive segmentation system, presented
at the 5th International Conference on Signal Processing, WCCC-ICSP, 2000, vol. 2,
pp. 918921 vol.2.
[25] A. Taher, K. Chehdi, and C. Cariou, Hyperspectral image segmentation using a
cooperative nonparametric approach, presented at the Image and signal processing for
remote sensing XIX, Dresden-Germany, 2013, vol. 8892, p. 88920J88920J8.
106
(ICoIT'17)
10thof April, 2017.
[26] V. K. Dehariya, S. K. Shrivastava, and R. C. Jain, Clustering of Image Data Set Using
K-Means and Fuzzy K-Means Algorithms, presented at the International Conference
on Computational Intelligence and Communication Networks (CICN), 2010, pp. 386
391.
[27] P. Brodatz, Textures: A Photographic Album for Artists and Designers. Dover
Publications, Incorporated, 1999.
[28] A. Taher, Approche cooprative et non supervise de partitionnement dimages
hyperspectrales pour laide la dcision. Rennes 1, 2014.
107
(ICoIT'17)
10thof April, 2017.
Adaptive Data Collection protocol for Extending Lifetime of Periodic

Sensor Networks
Ali K. M. Al-Qurabat
Department of Software, College of Information Technology, University of Babylon - Iraq
alik.m.alqurabat@uobabylon.edu.iq
Ali Kadhum Idrees
Department of Computer Science, College of Science for Women, University of Babylon - Iraq
ali.idrees@uobabylon.edu.iq

In Periodic Sensor Network (PSN), the periodic energy
Article History: efficient collection of a large amount of sensed data by the
Received: 20 March 2017 sensor nodes is considered as one of the main challenges in
this type of network. An adaptive sampling approach to
periodic data collection is needed for energy optimization
Published: 10 April 2017 due to the limited lifetime nature of the sensor batteries. In
this paper, we propose a protocol, called an Adaptive Data
Keywords: Periodic Collection protocol (ADaC), which collects periodically
Sensor Networks, Data sensor readings and prolong the lifetime of a Periodic
Collection, Adaptive Sensor Network (PSN). The lifetime of ADaC protocol is
Sampling Rate, Network divided into cycles. Each cycle is composed of two stages.
Lifetime. First, data collection. Second, sampling rate adaptation
based on the similarity between periods of one cycle using
Euclidean distance measure to adapt its rate of sampling
according to the dynamic modification of the monitored
environment. ADaC allows each sensor to adapt its
sampling rate in accordance with the monitored
environment conditions. We conduct extensive simulation
experiments on real sensor data by applying OMNeT++
network simulator to explain the effectiveness of the ADaC
protocol in comparison with a method without using
adaptive sampling.
1. INTRODUCTION
S
ensor nodes are spatially deployed in the region of interest in order to monitor
the physical or environmental phenomena like temperature, humidity, light,
pollution, pressure and sound. They collect the sensed data from the monitored
environment, manipulate the data locally, and transmit them to the sink for
further analysis. These sensor nodes work in a collaborative manner and constitute a Wireless
Sensor Network (WSN) [1, 2, 3, 18, 19]. In sensor node, the radio unit represents the
principal source of energy consumption. Therefore, it is important to remove redundant
108
(ICoIT'17)
10thof April, 2017.
sensed data before reporting them to the sink to save the energy and improve the lifetime of
sensor node [4].
Data collection approaches determine the way of sensors work in data collection and
sending to the base station. There are two models for data collection in WSNs: time driven
and event-driven [5, 20]. This work considers time-driven data collection which is named
Periodic Sensor Networks (PSNs). In PSN, every sensor node transmits the sensed data of
the monitored area to the sink periodically. Several PSNs applications use the periodic way
to monitor certain conditions regularly such as pressure, humidity, temperature, etc. Two
main challenges in PSN. First, PSN has to provide adequate lifetime in order to satisfy
applications needs. Second, data management is more difficult due to the huge amount of
collected data by this network. In PSN, the change in the monitored environment can slow
down or speed up. The energy consumption can be decreased when the sensor node modifies
its sampling rate based on the dynamic modification of the monitored phenomena. Therefore,
to prolong the network lifetime, adaptive sampling for periodic data collection is required
for energy optimization and data reduction [6].
This paper introduces the following contributions.

i) A protocol named ADaC is devised to collect the sensor data in an adaptive
way. The principal idea of ADaC protocol is to utilize the similarity of
collected data and adapts its sampling rate accordingly. ADaC works into
cycles. Two stages in each cycle: data collection and sampling rate adaptation
using Euclidean distance measure. The sensor node provides a new sampling
rate after each cycle based on the similarity between the periods of one cycle.
ii) A new adaptive sampling rate algorithm based on the Euclidean similarity is
suggested. In each cycle, the speed of readings capturing inside the sensor
node depends mainly on the previously calculated sampling rate adaptively.
iii) The simulation results are accomplished by OMNeT++ network simulator to
illustrate the effectiveness of the ADaC protocol. The ADaC protocol has
been compared to the results of the method without using adaptive sampling.
The rest of this paper is organized as follows: - Next section exhibits literature review.
Section 3 explains the description of ADaC protocol. Protocol evaluation is shown in
Section 4. Finally, we present conclusion and future works in Section 5.
This section reviews some related literature concerning the adaptive data collection in
WSNs. Adaptive collection approaches are considered as a good candidate to save energy
and extend the network lifetime of PSNs. The major objective of an adaptive collection
technique is to make the sensor node be able to change its sampling rate dynamically in
accordance with the monitored environment conditions. This can reduce the repetitive
gathered data, consume less energy, and decrease the processing load at the base station [6].
Some researchers used prediction as a way to adjust the sampling rate of sensor nodes [7,
8]. An energy saving information gathering scheme is proposed by Liu et al. [7] to predict
the sampling rate inside sensor using ARIMA model. In [8], the authors presented an
algorithm for adaptive sampling using Box-Jenkins approach to estimate the future sensor
readings, depending on the existing readings.
The work proposed in [9] consider adaptive sampling schemes based spatial correlation
among the physical sensed data. In [9], the sampling rate is adapted by the base station.
109
(ICoIT'17)
10thof April, 2017.
Initially, the base station activates a set of sensors to get the sensed data of monitored
environment. The correlation percentage is computed for the received sensed data to increase
or decrease the activated sensors. Chatterja and Havinga [10] present a sampling algorithm
based temporal correlation among sensed data. In this algorithm, the sampling rate is
modified depending on the stability of the monitored environment. The sampling rate
increases when the environment conditions are unstable, otherwise the rate decreases.
Masoum et al. [11] introduce an energy-saving mechanism for data collection. Their scheme
exploits spatio-temporal correlation among sensors and their sensed data to determine the
candidate sensors which are responsible for sampling and transmission. The selected sensors
are adaptively changed.
In recent years, Laiymani and Makhoul [12] proposed a scheme for adaptive sampling
using ANOVA model and Fisher test in PSNs. This algorithm works at the sensor node to
adapt its sampling rate. The authors in [13] proposed method to remove the repetition of
collected data in PSN called Prefix Frequency Filtering (PFF).
This paper suggests an Adaptive DAta Collection (ADaC) protocol for PSNs. The major
goal of ADaC is to remove redundant sensor readings, save energy, and improve the network
lifetime. ADaC performs two main phases. First, data collection according to adaptive
sampling rate. Second, ADaC allows to each sensor node to adapt its sampling rate for each
cycle (cycle = 2 periods) based on the Eculidean similarity. ADaC is simulated on the
OMNeT++ network simulator using real data of sensor nodes. The comparison results show
that our protocol can provide a better performance and prolong the network lifetime.
3. DESCRIPTIO OF THE ADaC PROTOCOL

ADaC protocol is given in more details in this section. The main objective of this
protocol is to enable each sensor to modify its sampling rate adaptively in accordance with
the dynamic changing of the monitored environment. Consequently, this reduces the
amount of redundant gathered data and minimizes energy consumption (extend the PSN
lifetime) whereas the quality of collected data is maintained sufficiently to allow significant
analysis. Figure 1 illustrates the flowchart of the proposed ADaC protocol. Table 1 explains
some parameters used in this paper.
110
(ICoIT'17)
10thof April, 2017.
Start
Initialize the parameters

SMPR ; MINSMP
Is No Sensor is excluded from

Residual Energy > 0? joining the current cycle
Yes
Set Cycle to j Periods
End
Set Periods Counter to 1
Sample Capturing
= Collect measures at SMPR
Update
Residual
Energy
Send Data to Sink (SPc )

= + 1
Is Yes Find Similarity between Periods

P > j? PSim = DISTSIM (S1 , S )
No
R PSim = 100 PSim
Update Sampling Rate Yes Is

SMPR (MinSMP /100) < MinSMP ?
No
Update Sampling Rate
SMPR (R PSim /100)
Figure 1. Flowchart of proposed ADaC protocol.
Table 1: Some parameters used in this paper

Parameter Value
Sampling Rate=
S Temperature readings series = {1 , 2 , . , 1 , }
Application criticality (minimum amount of sampling rate)
Similarity percentage
Reverse of similarity percentage
3.1. Data collection

Each sensor node senses the data reading periodically. These sensed time-arranged
data readings set forms time series. Therefore, ADaC protocol treats the sensor
readings as a time series. In our work, we named this time series as a temperature
readings series.
ADaC protocol is periodic and works into cycles. The cycle includes two periods
(j=2). The period is partitioned into time slots. Therefore, in this stage, the sensor
111
(ICoIT'17)
10thof April, 2017.
node n catches one temperature reading Si each time slot. At the end of each period,
the temperature readings series of sensor n is formed such that S =
{s1 , s2 , . , s1 , s }, where is the total number of temperature readings captured
during the period. The sensor node collects the temperature readings at sampling rate
(SMPR ) speed. The SMPR is initiated to temperature readings per period. The
redundant temperature readings captured by the sensor node increase in two states:
short time slot and slowly variation of a monitored area of interest.
3.2. Adaptive sampling rate

In this stage, ADaC protocol modifies its sampling rate based on the percentage of
similarity between temperature readings of different periods in the cycle. The main
purpose of this stage is to calculate the similarity among periods after each finished
cycle to acclimate the rate of sampling according to the new similarity rate.
Therefore, it uses the Euclidean distance measure to find the amount of similarity
between periods of each cycle.
Euclidean distance represents one of the most famous distance measures. ADaC
protocol uses similar function to identify the Similarity between two temperature
readings series Q and C of the same length using the following formula [14]

DISTSIM (, ) = =1(q i c )2 (1)
After that, in order to measure the similarity percentage (PSim ), it is defined as

follow:
PSim = DISTSIM (Q, C) 100 (2)
Algorithm 1 gives the similarity percentage (PSim ) calculation between two

temperature readings series Q and C.
Algorithm 1. Similarity Algorithm

Input: Q, C (two temperature readings series with dimension).
Output: PSim .
Process:
: 0
: 1
: Sum Sum + ( )2
:
: (Sum = 0)
: DISTSIM 0
:
: DISTSIM Sum
:
: PSim DISTSIM 100
: PSim
3.3. Verification the similarity of periods

In ADaC protocol, every node able to adapt its rate of sampling according to the
amount of similarity among temperature readings series collected during different
112
(ICoIT'17)
10thof April, 2017.
periods. The aim of computing the similarity between the temperature readings series
every cycle is to adapt the rate of sampling based on the new calculated similarity.
Therefore, the Euclidean distance is employed to discover the similarity percentage,
PSim among several periods per cycle. On one hand, if PSim is high, it means the
monitored condition is changed at a slow speed. Therefore, the sensor node will
decrease its rate of sampling to the minimum value to prevent collecting redundant
readings. On the other hand, if PSim is low, the sensor node will collect temperature
readings at approximately maximum sampling rate so as to prevent losing significant
readings. Therefore, to adapt the rate of sampling of sensor node in accordance with
the computed similarity among periods, the reverse of similarity percentage for
Euclidean coefficient (RPSim) is computed as follows
R PSim = 100 PSim (3)
Consequently, the process of adapting the sampling rate in the sensor node depends
on the RPSim, thus the application criticality will be taken into consideration in this
process. In ADaC protocol, the criticality of application is expressed as a minimum
amount of sampling rate in a period for a sensor node, MINSMP. MINSMP takes values
in the range 0 to 100 which represent the criticality level either low or high
respectively. The sensor node adapts the new sampling rate to the MINSMP (not to the
RPSim) when the recently calculated sampling rate is less than MINSMP. Depending on
the requirements of the application and before the deployment, all the sensor nodes
initialize their MINSMP. It is also possible to change MINSMP dynamically during the
lifetime of the network for the whole sensors or for just a given subgroup of sensors
if there are some types of management and control schemes are available. Algorithm
2 illustrates an adaptive sampling rate approach.
Algorithm 2. Adaptive Sampling Rate Algorithm

Input: J (one cycle = j periods), , MINSMP .
Output: SMPR (new sampling rate).
Process:
: SMPR // (initialize sampling rate to readings per period)
: ne > 0
: 1
: (S ) SMPR
: Send_TSink (S )
:
:
: PSim Similarity(S1 , Sj )
: R PSim 100 PSim
: (R PSim < MINSMP )
MINSMP
: SMPR ( )
100
12:
RP
13: SMPR ( Sim )
100
113
(ICoIT'17)
10thof April, 2017.
14:
15:
16:
17: SMPR
4. Protocol evaluation
The evaluation of ADaC protocol is given in more details in this section. The main
objective of this protocol is to enable each sensor to modify its sampling rate adaptively in
accordance with the dynamic changing of the monitored environment. Consequently,
ADaC protocol will remove redundant sensor readings, save energy, and improve the
network lifetime.
4.1. Simulation framework

To study and evaluate ADaC protocol, extensive simulations are performed with discrete
event simulator OMNet++ [15]. ADaC protocol is distributed at each sensor node and it is
based on the dataset of Intel Berkeley Research Lab [16]. PSN in this Lab includes 54
Mica2Dot sensors localized at the lab. The base station is located at the center of the lab. It
receives sensed readings from each sensor node by a single hop.
The sensed data of the weather (such as temperature, humidity, and light) are
periodically collected by these sensors once each 31 seconds. In our simulation, the sensor
nodes use a log file contains about 2.3 million readings collected previously by Mica2Dot
sensor nodes in the Lab. This article uses only one measure of sensor node measurements:
temperature2. There are 7 sensor nodes do not used in our simulation because its data may
be missed or truncated. Therefore, the temperature readings of 47 sensor nodes are selected
and stored. The results are the average of 47 sensor nodes. Table 2 gives the selected
parameters settings.
Table 2: Simulation parameters for PSN initialization

Parameter Value
PSN size 47 nodes
20, 50, 100 and 200 readings
MINSMP 20, 40 and 60
50 nJ/bit
100 pJ/bit/2
In the experimental simulations, (the total number of temperature readings captured

during the period) is set to 20, 50, 100 and 200 readings per period because almost the
papers (e.g. [6], [13] and [17]) use these values and we need to compare with these papers
in the next step of our work. In addition, some performance metrics are applied to assess
the effectiveness of the ADaC protocol such as sampling rate adaptation, number of
collected and sent temperature readings by a sensor node, and energy consumption.
ADaC protocol uses the same energy consumption model discussed in [17]. Energy
consumed by the sensor node depends only on the periodically collected and sent
temperature readings to the base station. The cost of transmission is calculated for a m
bits message and for a distance d as follow
(, ) = + 2 (4)
2 The others are done by the same manner.
114
(ICoIT'17)
10thof April, 2017.
The radio expends energy as described in [17], where is the energy consumed for
radio electronics and it is equal to 50 nJ/bit, is the energy consumed by the amplifier
and it is equal to 100 pJ/bit/2 .
The energy consumption required for capturing mbits by the sensor node is calculated
as follow
(, ) = (, ) 7 (5)
These experiment simulations consider the length of data reading m equal to 64 (we
choose this length because almost the papers (e.g. [6], [13] and [17]) use this length). In the
case of transmission, 64 bits are added to m bits message which corresponds to the
frequency of data reading m. Consequently, Energy consumption is defined as the total
energy dissipated at each sensor node during the collection and transmission of data
readings and formulated as follow
= (, ) + (, ) (6)
4.2. Sampling rate adaptation

Figure 2 shows the adaptation of sampling rate and for two sizes of temperature readings
(50 and 100 respectively). The results illustrate the ability of sensor device to modify its
rate of sampling dynamically depending on the application criticality level. The risk level
MINSMP can be determined according to the type and requirement of application used to
monitor the disaster.
(a) (b)
Figure 2. Sampling rate adaptation (A) = 50 AND (B) = 100.
In this experiment, MINSMP uses two values: 20 for low risk level disaster and 60 for
high risk level disaster. As shown in Figure 2, the adaptation of sampling rate is dynamic
and after each cycle based on the application criticality level (i.e., MINSMP = 20 or 60). The
results in Figure 2 (a) and (b) validate the good performance of our protocol.
4.3. Percentage of sent readings

When collecting the data readings at each period, ADaC protocol at the sensor node able
to decrease the number of sent readings to the base station by using adaptive sampling
method. Figure 3 demonstrates the percentage of sent readings by a sensor node to the base
115
(ICoIT'17)
10thof April, 2017.
station at the end of simulation for ADaC protocol compared with the results of the method
without using adaptive sampling technique.
The results illustrate that ADaC protocol at each sensor node decreases from 26% to
69% of the Percentage of sent readings to the base station as comparing to 99% without
using adaptive sampling. Therefore, ADaC protocol removes the redundant collected
readings successfully and the percentage of sent readings to the base station is reduced. We
can also see that the volume of sent readings from the sensor node to the base station
decreases when increases or MINSMP decreases. This is due to the percentage of sent
readings rely on the number of collected readings and the risk level of application.
120 ADaC(S min=20)

ADaC(S min=40)
110 ADaC(S min=60)
WithoutAdaptiveSa mpling
Percentage of Sent Readings
100
90
80
70
60
50
40
30
20
10
0
20 50 100 200
Number of Readings/ Period
Figure 3. Percentage of sent readings.
4.4. Energy consumption

Figure 4 shows the energy consuming by ADaC protocol at the sensor node compared
with the results of the method without using adaptive sampling technique.
0.6 ADaC(Smin=20)
ADaC(Smin=40)
0.55 ADaC(Smin=60)
WithoutAdaptiveSampling
0.5
0.45
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
20 50 100 200
Number of Readings/ Period
Figure 4. Energy consumption at each sensor node.
116
(ICoIT'17)
10thof April, 2017.
As shown in Figure 4, ADaC protocol saves energy because it reduces both collected
and sent readings at the sensor node. The consumed energy of a sensor node using ADaC
protocol is minimized from 26% to 68% as compared with the results of the method without
using adaptive sampling technique. It can be observed that ADaC protocol is effective in
terms of reducing energy consumption for the applications with high and low risk level,
where more energy is saved when MINSMP decreases.
5. Conclusion and future works

This paper presents a protocol, called Adaptive Data Collection protocol (ADaC), which
collects periodically sensor readings and improves the PSN lifetime. ADaC protocol works
into cycles and consists of two phases. First, collecting the data readings. Second, sampling
resolution to adapt the rate of sampling at the sensor node in accordance with the dynamic
changing of observed environment. ADaC protocol considers the risk level of an application
by fixing the minimum sampling rate that permits to sensor node to collect readings at a
minimum rate while maintaining a good quality of the collected readings. To assess the
effectiveness of ADaC protocol, we compared it with the results of the method without using
adaptive sampling technique using several performance metrics like a percentage of sent
readings and energy consumption. Simulation results show the efficiency of ADaC protocol
to conserve the energy at the sensor nodes thus prolong the PSN lifetime.
In future, we plan to improve our work to consider the sensing overlap among sensor
nodes at the aggregator level to optimize both the aggregated readings and lifetime while
maintaining a good accuracy. In addition, we plan to compare our proposed method with an
existing method in the literature [20].
6. References
[1] I.F. Akyildiz, and M.C. Vuran, Wireless sensor networks, Vol. 4. John Wiley & Sons,
2010.
[2] M. Abdelaal, O. Theel, C. Kuka, P. Zhang, Y. Gao, V. Bashlovkina, D. Nicklas, and M.
Frnzle, Improving energy efficiency in QoS-constrained wireless sensor
networks, International Journal of Distributed Sensor Networks, 2016.
[3] A.K. Idrees, K. Deschinkel, M. Salomon, and R. Couturier, Distributed lifetime
coverage optimization protocol in wireless sensor networks, The Journal of
Supercomputing, vol. 71, no.12, pp.4578-4593, 2015.
[4] X. Tang and J. Xu, Adaptive data collection strategies for lifetime constrained wireless
sensor networks, IEEE Transactions on Parallel and Distributed Systems, vol. 19, no. 6,
pp. 721734, 2008.
[5] G. Campobello, A. Segreto, and S. Serrano, Data gathering techniques for wireless
sensor networks: A comparison, International Journal of Distributed Sensor Networks,
vol. 2016, 2016.
[6] A. Makhoul, H. Harb, and D. Laiymani, Residual energy-based adaptive data collection
approach for periodic sensor networks, Ad Hoc Networks, vol. 35, pp. 149160, 2015.
[7] C. Liu, K. Wu, and M. Tsao, Energy efficient information collection with the arima
model in wireless sensor networks, in Global Telecommunications Conference, 2005.
GLOBECOM'05. IEEE, vol. 5, pp. 5pp., 2005.
117
(ICoIT'17)
10thof April, 2017.
[8] Y. W. Law, S. Chatterjea, J. Jin, T. Hanselmann, and M. Palaniswami, Energy-efficient

data acquisition by adaptive sampling for wireless sensor networks, in Proceedings of
the 2009 International Conference on Wireless Communications and Mobile Computing:
Connecting the World Wirelessly. ACM, pp. 11461151, 2009.
[9] R. Willett, A. Martin, and R. Nowak, Backcasting: adaptive sampling for sensor
networks, in Proceedings of the 3rd international symposium on Information processing
in sensor networks. ACM, pp. 124133, 2004.
[10] S. Chatterjea and P. Havinga, An adaptive and autonomous sensor sampling frequency
control scheme for energy-efficient data acquisition in wireless sensor networks, in
International Conference on Distributed Computing in Sensor Systems. Springer, pp. 60
78, 2008.
[11] A. Masoum, N. Meratnia, and P. J. Havinga, An energy-efficient adaptive sampling
scheme for wireless sensor networks, in Intelligent Sensors, Sensor Networks and
Information Processing, 2013 IEEE Eighth International Conference on. IEEE, pp. 231
236, 2013.
[12] D. Laiymani and A. Makhoul, Adaptive data collection approach for periodic sensor
networks, in 9th International Wireless Communications and Mobile Computing
Conference (IWCMC). IEEE, pp. 14481453, 2013.
[13] J. M. Bahi, A. Makhoul, and M. Medlej, A two tiers data aggregation scheme for
periodic sensor networks. Adhoc & Sensor Wireless Networks, vol. 21, no. 1, 2014.
[14] J. Lin, E. Keogh, S. Lonardi, and B. Chiu, A symbolic representation of time series,
with implications for streaming algorithms, in Proceedings of the 8th ACM SIGMOD
workshop on Research issues in data mining and knowledge discovery. ACM, pp. 211,
2003.
[15] A. Vargas, Omnet++ discrete event simulator, [Online] 2003, https://omnetpp.org/
(Accessed: 10 January 2017).
[16] S. Madden, Intel berkeley research lab, [Online] 2004, http://db.csail.mit.edu/labdata/
labdata.html. (Accessed: 25 August 2016).
[17] H. Harb, A. Makhoul, A. Jaber, R. Tawil, and O. Bazzi, Adaptive data collection
approach based on sets similarity function for saving energy in periodic sensor networks,
International Journal of Information Technology and Management, vol. 15, no. 4, pp.
346-363, 2016.
[18] Idrees AK, Deschinkel K, Salomon M, Couturier R. Coverage and lifetime optimization
in heterogeneous energy wireless sensor networks. ICN 2014. 2014 Feb 23:60.
[19] Idrees AK, Deschinkel K, Salomon M, Couturier R., Perimeter-based coverage
optimization to improve lifetime in wireless sensor networks, Engineering Optimization,
vol. 48, no. 11, pp. 1951-72, 2016.
[20] Harb, H., Makhoul, A., Jaber, A., Tawil, R. and Bazzi, O., Adaptive data collection
approach based on sets similarity function for saving energy in periodic sensor networks,
International Journal of Information Technology and Management, vol. 15, no. 4, pp.346-
363, 2016.
118
(ICoIT'17)
10thof April, 2017.
Alarm system using Microcontroller , GPS and GSM

devices
Nadia Mahmood Hussein
Department of Computer Science,, College of Science, Almustansiriyah University -Baghdad - Iraq
nadia.salam97@yahoo.com
Saad Najim AlSaad
Department of Computer Science, College of Science, Almustansiriyah University -Baghdad -Iraq
alsaad6060@ gmail.com
ABSTRACT
ARTICLE INFO
During last ten years, it is noticed an increasing number of
Article History: accidents causing losing of large number of victims This
Received: 17 March 2017 paper related to an explosion alarm system to inform the
emergency center to make fast response for saving and
rescuing life of people. The whole system is controlled by
Published: 10 April 2017 arduino uno microcontroller unit which senesces the
Keywords: SMS, environment and gives reaction based on what it sensed. Also
Arduino, GSM, GPS, GPS (global Position System) is used that gives the latitude
Sound sensor, Alarm and longitude of the event and sends this information as a
short message to GSM (Global System for Mobile
Communication) mobile using SMS (Short Message Service)
. In other side, database of location (latitude, longitude) of
each help center place is saved to be used for determining the
nearest center to the accident through shortest path formula.
Finally the path between emergency center and accident
location is tracked through Google map application.
1. INTRODUCTION
W
ith the advancement of human civilization, its safety has a prime concern.
Alarm system is one of the most important means of assistance. It alarms the
concern persons that an undesired accident is happened. Automatic alarm
systems are utilized for decreasing the likelihood of danger occurring. Automatic alarms
system operate by using an intelligent devices to detect the accident quickly and to get a
shortest time between the moment of the problem occurs and the moment of help responding
in order to reduce the substantial losses in a short period of time. The alarm system will give
a way to intervene quickly and effectively control on the problem and thus reducing the
amount of the losses.
Researches had implemented alarm systems, for example alarm system for earthquake [1] ,
unwanted events in home [2] and fire detection [3].
This paper focuses on design and implementation of alarm system. The paper is structured
as follows: section 2 represents the proposed system as block diagram and flowchart.
119
(ICoIT'17)
10thof April, 2017.
2. PROPOSED SYSTEM
Figure (1) illustrates the block diagram of the system. It consists of two parts: The
transmitting part and emergency monitoring part. The mechanism work of the system is
started at the transmitting part. The sensor vibrates to indicate an explosion is occurred. GPS
retrieves the location information of the event from satellites in the form of latitude and
longitude to the microcontroller. GSM sends the location and other information in the form
of SMS from microcontroller to the emergency monitoring unit. SMS is a GSM mobile
technology that can perform remote communication wherever they are. Through this facility
messages can be sent quickly, accurately and at a low cost. All the modules of transmitted
unit are controlled by arduino open source microcontroller.
Figure 1: The block diagram of the system.
The emergency monitoring unit received the alarm message by any android phone. This
developing system is based on android phone to get much more mobility and user friendly
.Android application give a many service to reduce the complexity in receiving and sending
gathered information (alarm message). The gathered information is sent and stored on web
server (cloud) utilities from android application. Cloud computing is an Internet-based
computing, Wi-Fi Internet services used to get the information from android mobile to cloud
and from cloud to PC. SMS gateway enables the application to send/receive SMS messages
to mobile devices with the computer. Gateway is most fast and reliable way for SMS sending
and receiving. Now all the transmitted information arrives in computer to be subjected to
analysis and then a formula is run to find the nearest helping center to the explosion place.
120
(ICoIT'17)
10thof April, 2017.
This is performed by a database of all the helping centers3 in Province by stored the location
of each helping center in form of (latitude, longitude) and the phone number for each one.
Also all the alarm messages is stored in database for future reviewing. The transmitted
information is send as SMS to the nearest helping center. Finally the path is tracked
graphically using Google map application; figure (2) illustrates the shortest path on Google
Map.
Figure 2: The shortest path on Google Map
3. FLOWCHART OF SYSTEM
Figure (3) shows the flow chart of the system. It gives the basic idea about how the system
works. First of the initialization of the system is carried out in which it checks if the system
in normal state (no accident) and it is properly or not. If the system is not in normal state then
an explosion detect. Then it tracks position of event through GPS and send SMS through
GSM module to the emergency center. In the emergency center checks the shortest path
between explosion position and helping center position.
3
The helping centers is any places that can present help
Such as the police center, hospital, etc.
121
(ICoIT'17)
10thof April, 2017.
Figure 3: The flowchart of a system
4. HARDWARE REQUIREMENT
The proposed system contains the following hardware components that participate in
performing the task of alarm system.
4.1 ARDUINO
Arduino is an open source physical computing platform based on small microcontroller
board with simple input/output board. Arduino is composed of two major parts: the arduino
board, which is the piece of hardware and the arduino IDE, which is the piece of software. It
is multiplatform environment that may run on Windows, Macintosh, and Linux. Arduino is
a flexible technique; it can communicate with the computer with an easy way also it is a
powerful way to build better prototypes [4].The proposed system used arduino uno.
4.2 GLOBAL POSITION SYSTM (GPS)

GPS is formally known as the NAVSTAR (Navigation Satellite Timing and Ranging).
GPS includes 28 active satellites that are uniformly settled on six different circular orbits.
GPS is a powerful technique that gives the current date, time, latitude, longitude, altitude,
122
(ICoIT'17)
10thof April, 2017.
speed, and travel direction [5] . Although receiver can calculate the position, the accuracy of
positioning is dependent on the number of satellites in view, signal qualities, weather
conditions, etc. It can be used in a many applications including navigation, tracking systems,
mapping and robotics. All these information are provided by GPS in any time and in any
weather condition [6]. The proposed system is used GPS modem to determine the (latitude,
longitude) of the vehicle.
4.3 GLOBAL SYSTEM FOR MOBILE COMMUNICATION (GSM)

GSM is an open, cellular network used for transmitting mobile voice and data services. It
is a wireless modem that works with a GSM wireless network [7] . GSM modem is a
specialized type of modem that accepts SIM card and operates over a subscription to a mobile
operator just like a mobile phone [8] . The proposed system is used arduino GSM shield for
sending the gathered information as an alarm message to android mobile.
4.4 SOUND SENSOR

The sound sensor module provides an easy way to detect sound intensity. In this system,
it is used to detect the sound of explosion. This module can be used for switch, security, and
monitoring applications [9].
5. SOFTWARE REQUIREMENTS
5.1 Arduino software
In the transmitted part the system used (Arduino c language) to perform the connection of
the board with PC using serial port such as COM1 or COM2. The ATmega328 on the arduino
uno comes with a boot loader that allows you to upload new code to it without the use of an
external hardware programmer. It communicates using the STK500 protocol [10]
5.2 (c# language)

In the central part the c# programing language is used to design the interface of the system,
also SQL data base is used to store the information of the message content, real position of
emergency center in the form of (latitude, longitude) and its phone number.
6. RESULT
A figure (4) illustrates the detailed message and its content, the first word (alarm) is a
keyword to differ between alarm message and other messages, the second word is (danger)
means there is an explosion occur , the (3322.872) is latitude , and (4422.759) is longitude
of the event . The proposed system deals with the critical situation, so we should be take into
consideration the time span between the moment of detection and the moment of intervene.
The maximum time need to transmit the message from GSM to android mobile is 11.5 second
and the minimum time is 6 second, the maximum time need to transmit the message from
android mobile to PC is 11 second and the minimum time is 5.5 second.
123
(ICoIT'17)
10thof April, 2017.
Figure 4: The alarm message contains in

android mobile
7. CONCLUSION
Our prime objective in this system is to reduce the numerous losses in human life beside
of the economic loss , this system can detect the problem that occur by using sensors that
integrated with a microcontroller , GSM and GPS modules . SMS is used to alert user via
mobile phone when possible problem occurs. The message sent to center emergency that
decided the nearest helping center to the vehicle, also it view the shortest path on the Google
map application.
8. RECOMMENDATIONS
The system can be further developed with added hardware like camera in transmitted
part to describe the situation online and send it as a video to emergency monitoring part.
9. REFERENCES
[1] Hima MS, Josphin Subha, Kalyani Devi G, Karthika Sankaran, Naveen S, "Earthquake
Indicator Using Arduino," International Journal of Research in Advent Technology, Vol.4,
No., vol. 4, no. 8, 2016.
124
(ICoIT'17)
10thof April, 2017.
[2] A. Gupta, "Intelligent Home security using GSM communication module," Vols. 13 No. 1
Jan. 2015, pp. 239-242, no. 1, pp. 239-242, 2015.
[3] Omar Asif , Md. Belayat Hossain , Mamun Hasan , Mir Toufikur Rahman, Muhammad E. H.
Chowdhury, "Fire-Detectors Review and Design of an Automated, Quick Responsive Fire-
Alarm System Based on SMS," Int. J. Communications, Network and System Sciences, 2014,
7, 386-395, vol. 7, pp. 386-395, 2014,.
[4] M. Banzi, Getting started with Arduino, 2d ed., Make:Books, 2011.
[5] AHMET IRKIN and SERKAN KARAKIS, Relative distance measurement using GPS and
internal vehicle sensors, CHALMERS UNIVERSITY OF TECHNOLOGY, 2011.
[6] Basavaraj Chougula, Archana Naik , Monika Monu, Priya Patil and, "Smart girls security
system," International Journal of Application or Innovation in Engineering and Management
, vol. 3, pp. 1-4, 2014.
[7] P. Chandra, Bulletproof wireless security GSM, UMTS, 802.11 and ad hoc security, 1d ed.,
Elsevier,, 2005.
[8] Ramchandra K. Gurav and Rohit Jagtap, "Wireless digital notice board using GSM
technology," " International Research Journal of Engineering and Technology (IRJET) , vol.
2 , pp. 1-3, 2015..
[9] Jiankai.li, Sound sensor user manual, seeed Grow the di_erence, 2015.
[10] Santoso Budijono ,Je_ri Andrianto and Muhammad Axis Novradin Noor, "Design and
implementation of modular home security system with short messaging system," 2014.
125
(ICoIT'17)
10thof April, 2017.
An Analytical Study for Some Drawbacks and Weakness

Points of the AES Cipher (Rijndael Algorithm)
Omar A. Dawood
Computer Science, College of Computer, University of Anbar - Iraq
The_lionofclub@yahoo.com
Othman I. Hammadi
Computer Science, College of Computer, University of Anbar - Iraq
othmanibraheem@yahoo.com.

The present paper includes a research study about the weak
Article History: points and the major weakness aspects of the Rijndael cipher
Received: 19 March 2017 from the point of view design. Although several published
papers for most researchers around the world included either
the developing models or improving techniques by depending
Published: 10 April 2017 on Rijndael cipher. In this study, opposite matter about some
scientific criticism for certain essential points in the AES
Keywords: Rijndael construction will be discussed. When the AES was selected
cipher, MARS, Serpent, 16 years ago, the digital technologies were quite different
Twofish, RC6, DES, AES, from now and the magnitude of the challenges was less, so
Block Cipher, Symmetric with the recent advanced technology and the emergence of
Cipher. new applications like Big datas applications in addition to
the applications have run with 64-bit and a lot of other
applications, it has become a necessity for designing a new
contemporary algorithm for the current demands. Especially
young Rijndael that has faded and its sun had set as it has been
believed by many researchers. Since the experts and designers
of information security in previous time determined its retired
date for ten years. In this study, a list of drawbacks and
vulnerabilities for the Rijndael internal structure in addition
to new recommendations for the future work will be
diagnosed. No one denies that the selection of Rijndael was a
good choice for civil applications on software and hardware
implementations and on many of various platforms, but the
excessive speed for the IT progress leads to take in to account
recalculation of the security level for the current and
perspective future requirements.
126
(ICoIT'17)
10thof April, 2017.
1. INTRODUCTION
I
n 1997, the National Institute of Standard and Technology (NIST) unit of the US
commerce department announced an open competition, and invited the cryptography
and the data security specialists from around the world to participate in the conference
to select a replacement for the old standard cipher algorithm (DES) and Triple-DES, because,
during the last years DES has become obsolete for its too short key and block sizes,
notwithstanding the current advances in computing technology [1]. The new encryption
standard should support block lengths of 128-bits and key lengths of 128-bit, 192-bit and
256-bits. This demand determined by the NIST. In this conference applied twenty-one
candidates algorithms, the specification of these algorithms varies widely in structure and
form and their coverage of implementation some of these candidate using formal
mathematical style and other depending on combinations text diagrams and pseudo code.
The first round of conferences held in California in 1998 and fifteen algorithms have been
submitted. The initial evolution criteria determined by three elementary categories security,
cost and algorithm and implementation characteristics. The large majority of the candidates
algorithms satisfy the criteria determined by the NIST [2]. In March 1999 the second AES
conference was held in Rome, Italy submitted the final evolution criteria that include general
security, ease of implementation on both software and hardware, implementation attacks and
flexibility in encryption and decryption, ciphering key and other factors. Later fifteen
algorithms were reduced to five algorithms [3].
2. THE SELECTION OF DES REPLACEMENT
There were several algorithms that were not accepted because either they did not satisfy the
security requirements presented by NIST or because the weakness such as LOKI97, DAL,
HPC and another some candidates algorithms had not been eliminated because of the security
problems but for other reasons such as cost, slow and bad performance for example Cast-
256, Crypton, DFC, E2 and SAFER+. Only five algorithms were accepted in this conference
that were considered as the optimal algorithms and have a good specification with margin of
security these algorithms include (MARS, RC6, Rijndael, Serpent, and Towfish) [4]. The
origin of each candidate algorithm with submitters and countries can be shown in the
following Table 1.
TABLE 1: The First of the AES Candidates Algorithms
Countries Submitters Algorithms
Canada Entrust Technologies, Inc. CAST-256
South Korea Future Systems, Inc. CRYPTON
Canada, Richard Outer bridge, Lars DEAL

Norway Knudsen
127
(ICoIT'17)
10thof April, 2017.
France CNRS - Centre National pour la DFC

Recherche Scientifique Ecole
Normale Suprieure
Japan NTT Nippon Telegraph and E2

Telephone Corporation
Costa Rica Tec-Apro Internacional S.A. FROG
U.S.A Richard Schroeppel HPC
Australia Lawrie Brown, Josef Pieprzyk, LOKI97

Jennifer Seberry
Germany Deutsche Telekom AG MAGENTA
U.S.A. IBM MARS
U.S.A. RSA Laboratories RC6
Belgium Joan Daemen, Vincent Rijmen Rijndael
U.S.A. Cylink Corporation SAFER+
U.K., Israel, Ross Anderson, Eli Biham, Lars Serpent

Norway Knudsen
U.S.A. Bruce Schneier, John Kelsey, Doug Twofish

Whiting, David Wagner, Chris
Hall, Niels Ferguson
The selection process was in accordance with a hundred points that were distributed
according to seven criteria and these criteria gave the algorithms variations in access to capture
points in these seven criteria [5]. As it shown in Table 2.
TABLE 2: The Main Criteria for the Selection of AES Cipher
The main Criteria determined by NIST Distribution Points
Algorithm Design & Presentation (10 points)
Security (30 points)
Ease of implementation (10 points)
Usage Flexibility (10 points)
Performance/ Computational Efficiency (10 points)
Performance/Adaptability on Smart cards (10 points)
Demonstrated/Expected strength against Cryptanalysis (10 points)
Future Resilience (10 points)
128
(ICoIT'17)
10thof April, 2017.
3. SPEED OVERVIEW ON THE FIVE FINALISTS CANDIDATES

There is no doubt about the AES candidate algorithms have been designed by the worlds
best academic cryptographers. The majority of these candidates are highly regarded for their
security and there seems little doubt that when the finalists are selected. In the second and
the first round of conferences all of them were be very high quality and performance. But the
last five AES finalists (MARS, RC6, Rijndael, Serpent, and Twoftsh) were believed to be
secure, and none was clearly superior in all aspects. So the choice of Rijndael was based on
its balance of flexibility, ease of implementation and speed in both hardware and software
[6]. The stats for the number of votes obtained by each algorithm during the voting process
in the conference. Rijndael obtained on 86 votes, Serpent 59 votes, Towfish 31 votes, RC6
23 votes and MARS 13 votes. In October 2000 NIST announced that the Rijndael algorithm
selected as a winner because it took most of the votes. All the five finalist algorithms were
not built from scratch since all of these algorithms were based on the predecessors
algorithms. Originally these finalists block ciphers similar in their structure and internal work
to their predecessors. Each of these ciphers inherited properties from them and enriched with
new ideas to be more resistance structure for facing all the known attacks. Each of these
competitions ciphers deduced good ideas and strong points from their predecessors and
addressed the weakness points. Like MARS, Twofish and RC6 that are dependent on DES
cipher [7]
3.1. ADVANCE ENCRYPTION STANDARD (MARS CIPHER)
One of the five finalist block cipher chosen as AES, IBM s' algorithm uses innovative
techniques including multiplication data-dependent rotation. One of the components of the
cipher is 9*32-bits of S-box. MARS is a 128 bit-block cipher and a variable key size ranging
from 128-bit to 444-bit. As most symmetric-key block ciphers MARS is based on a Feistel
network structure and a round function. This function provides the basis for the encryption
mechanism and combines several linear and non-linear operations [8]. MARS operates on
32-bits microprocessor and run slowly. MARS designed to be more secure than tripleDES
and to be used in computers today but it has been observed that it is more complicated overall
the AES candidates. The main strength of MARS is its robustness. This was the main design
goal but MARS contains more fail stop mechanisms than any of the other finalists. Due to
the heterogeneous structure and the large variety of strong operations in MARS. Because
the fail-stop mechanisms in MARS makes its hardware implementation more involved than
the other finalists [9].
3.2. ADVANCE ENCRYPTION STANDARD (SERPENT CIPHER)
Serpent cipher Created by R.Anderson and E.Biham, two of the leading experts in
cryptanalysis, which can be viewed as an indication to its strength. It is one of the AES
candidates and it has been implemented on 8-bit, 16-bit, and 32-bit platform like other AES
submissions. Serpent is a (SPN) that uses 32-rounds. The algorithm has two different modes
of implementation: standard and bitslice mode. The standard mode operates on individual
bits or groups of four bits, while the bitslice mode improves software efficiency by operating
on entire 32-bit words [10]. Serpent has a block size of 128-bits and support a key size 128-
bits, 192-bits, and 256-bits. The main operations on a block are implemented by four word
(32-bits). The S-box of Serpent appears to have a high security margin. Serpent uses the set
129
(ICoIT'17)
10thof April, 2017.
of different S-boxes with four bits inputs and four bits outputs. In each round there is used
one of S-box working in parallel 32 times. After the first eight rounds, all S-boxes are used
and this unit of eight rounds is repeated four times. The main selling point of Serpent is its
very conservative number of rounds. Serpent does not have fail-stop mechanisms as in
MARS [11].
3.3. ADVANCE ENCRYPTION STANDARD (TWOFISH CIPHER)
Symmetric key block cipher proposed by Counterpane. The Twofish block cipher
employs a 16-round feistel network with additional whitening of the input and output. The
128-bit plaintext block is split into four 32-bit words. In the input whitening step each 32-bit
word is XORed with a different 32-bit input whitening key. Towfish work with 128-block
size and accept a variable length key up to 256-bit sufficient on a various platforms, with
objective function work on 32-bits word with four key dependent 8*8 S-box followed by
4*4 maximum distance separable matrix over GF(28) a pseudo hadamard transform and key
addition [12]. Twofish cipher was designed for flexibility, and indeed it offers a wide variety
of implementation tradeoffs. It is also a very fast cipher. However, the same design for
flexibility also resulted in a cipher which is very hard to analyze. The designers used many
tricks to obtain flexibility which security implications are not clear. Towfish has the
efficiency and speed on both software and hardware implementable on a wide variety of
platform and applications and suitable for stream cipher, hash function, and MAC [13].
3.4. ADVANCE ENCRYPTION STANDARD (RC6 CIPHER)
RC6 is very simple algorithm and the easiest of the AES candidates to implement
developed by Ron Rivest in collaboration with associates from RSA laboratories. It is based
on carefully crafted ciphers such as RC2, RC4, and RC5 exactly it is strengthened version of
RC5 which was proposed in 1995. The RC6 cipher works with 20-rounds feistel structure
with objective function work on 32-bit modular multiplication, addition, and XOR with
addition key [14]. The main advantage of RC6 is its simplicity and speed which may serve
as an indication of the suitability of the current design as well. The main argument against
RC6 is single point of failure for the design method. There are also lingering concerns
regarding the number of rounds used in RC6. The main difference between the RC6 and its
predecessor RC5 is the RC6 algorithm depends mainly on the usage of four working registers
each with size of 32-bits, this means it handles 128-bits input/output blocks in contrast to the
RC5 cipher which works in two registers of 64-bits [15].
3.5. ADVANCE ENCRYPTION STANDARD (RIJNDAEL CIPHER)
The Rijndael cipher is the winner algorithm in the NIST competition designed by Joan
Daemen and Vincent Rijmen from Belgium and recently selected as the official (AES). It is
well suited for work across all platforms and involve clean mathematical structure and seem
as a unique among the candidates ciphers. It is an iterated block cipher that has a very simple
structure and easily implemented in both hardware and software. Rijndael operates with
block size 128-bits and with variable key length of 128-bits, 192-bits and 256-bits. It is a fast
and flexible cipher. Rijndael is somewhat similar to SQUARE cipher and the lessons from
SQUARE are incorporated in its design [16].
130
(ICoIT'17)
10thof April, 2017.
4. ANALYSIS AND JUSTIFICATIONS FOR THE DRAWBACKS OF THE

RIJNDAEL CIPHER
After deep study and an intensive analysis of the internal structure and the algebraic
foundation of the AES cipher, there is evidence that the AES algorithm has many suspicious
aspects and it has suffered from several vulnerabilities from the design term that brings the
attention and can be listed as follows:
1- Most of the developing methods of the previous studies for most researchers and
specialists deal with the AES improvement and exactly focus on increasing the
number of rounds or increasing the block size in order to increase the security level
and this clue does not consider the best solution for the development experiments.
2- Most of the AES algorithms (MARS, Serpent, RC6, Twofish, and Rijndael) designed
to work with platforms of 32-bits and could not work with 64-bit and this is
considered a negative factor in some operating system of 64-bit from the designing
perspective as it was mentioned in the personal blog of Bruce Schneier.
3- The three keys of changeable length (128-bit, 192-bit, and 256-bit) of the AES cipher
have not real length except the first one with a length of (128-bit) and the two others
are unreal as it claimed in the most references. Since the XORing operation between
the key and state matrix is not fit to the same size or in another word it only increased
the probability of the search space for the key generation to the rest rounds.
4- The R-Con table with fixed values calculated according to specific formula can be
considered as an awkward point in the designing map for the key F-function
especially for the key generation that should depend on a complicated one-way
technique or intractable methods.
5- The decryption process with AES structure is slower than the encryption process
especially in the embedded devices and this feature refers to the unbalancing structure
in encryption and decryption process for some devices.
6- The encryption process for thousands of bits will give an obvious difference time
between the encryption and decryption process due to the accumulator for repetition
of thousand rounds, so the difference will be a clearly evident.
7- The AES cipher work with 128-bit of plaintext or 16-byte and all internal operations
rely on the byte-oriented implementation that means the maximum value can be
implemented in Hexa-decimal is (FF).
FF (byte) * FF (byte) mod M= (FF) Byte
So the multiplying byte by byte also will give byte as a final result.
8- The selection only one algorithm (Rijndael) and discarding the other four best
algorithms (MARS, RC6, Twofish, and Serpent) according to the essay that was
mentioned in NIST report (do not put all eggs in the one basket) is an unsatisfactory
reason. Since there is a possibility to combine all the candidates algorithms in one
package of software and give more options and flexibility for selecting an appropriate
algorithm according to the application nature.
9- Several modern attacks have proven their effectiveness from a theoretical perspective
like the algebraic attacks, implementation attack and side channel attacks.
10- The AES ciphers life time is dedicated for ten years to face the advance of
technology progress at that time. So many researchers think that finding an alternative
131
(ICoIT'17)
10thof April, 2017.
model becomes a vital issue according to the requirements of the last decade. So the
current state may need more options from all aspects
11- AES cipher with secure of 128-bit may not be appropriate for the big data applications
and others modern big applications like secure cloud storage. Therefore; these
applications with huge data may need a larger algorithm with a larger order of
mathematical and structure foundations with tradeoff speed.
12- The Rijndael cipher got the higher voting in the NIST competition but this does not
means it is the best from all factors since some algorithms exceed the Rijndael from
certain aspects but the accumulated scores were from the Rijndael share.
13- The X-time function designed to work with (02) recursive multiplications clue for a
number of times according to the coefficient values instead of the tradition
mathematical multiplication operations that have taken modular on the irreducible
polynomial equation of degree eight.
14- The last round without mixcolumn has no effective role in the security factor as it
was mentioned by the authors of the Rijndael. So it can be added or discarded.
15- The fixed S-Box with two tables of 256-values in hexa-notation may form a good
objective for the cryptanalysts attacks, since each value has a mapping to the
corresponding value with inverse picture.
5. RECOMMENDATIONS
After this contrastive study, there are several important recommendations for
future work that can be summarized here:
1. Design a dynamic algorithm with the changeable stage for all linear and non-
linear layers.
2. Develop a new model with a higher mathematical order of finite field or
Galois Field (GF) and with higher irreducible polynomial to increase the
security level and the complexity.
3. Improve the AES cipher to work with the system of 64-bits and eliminate any
numerical constants that can be exploited by the attackers.
4. Made the improved algorithm more balanced in encryption and decryption
process with embedded devices and restricted hardware devices.
5. Develop new techniques for the key generation that generated key with real
length specifically for the long keys that comprises 192-bit, 256-bit and upper
than that increase the guessing probabilities effectively.
6. CONCLUSIONS
The current research presented a simple literature review and an analytical study for
the AES selection and the main standard criteria for the design principles. The basic
target of this study is to highlight weaknesses and vulnerabilities in addition to explain
the gaps of the design elements that possible to be exploited in the AES structure. This
paper discussed the Rijndael features on the one hand constructions fails stop and what
are the best alternatives by given a set of fundamental factors diagnostic for the negative
aspects from the point of view author and the scientific researchers articles around the
world. The introduced research also includes some future recommendations for the
132
(ICoIT'17)
10thof April, 2017.
designers and academic specialists in addition to the essence solutions for developing a
modern algorithm.
7. REFERENCES
[1] A. M. Sagheer, S. S. Al-Rawi, and O. A. Dawood, Proposing of Developed Advanced
in Encryption Standard AES, IEEE Computer Society DOI 10.1109/DESE, Page No.
197, 2011, The Fourth International Conference in Developments in E System
Engineering DESE, Dubai, 2011.
[2] J. Daemen and V. Rijmen, "The design of Rijndael: AES the advanced encryption
standard", Springer-Verlag, 2002.
[3] Dr B. Gladman, "Implementation Experience with AES Candidate Algorithms", Second
AES Conference, 28th February 1999.
[4] B. Gladman and Worcester, "The Need for Multiple AES Winners", United Kingdom,
7th April 1999.
[5] J. Dray, Report on the NIST Java AES Candidate Algorithm
Analysis,http://csrc.nist.gov/encryption/aes/round1/r1-java.pdf, November 8, 1999.
[6] O. A. Dawood, A. S. Rahma and A. J. Abdul Hossen, The Euphrates Cipher, IJCSI
International Journal of Computer Science Issues, Volume 12, Issue 2, March 2015, ISSN
(Print): 1694-0814 | ISSN (Online): 1694-0784.
[7] Federal Information Processing Standards Publication 197 "Announcing the
ADVANCED ENCRYPTION STANDARD (AES)", November 26, 2001.
[8] IBM MARS Team, "MARS and the AES Selection Criteria", May 15, 2000.
[9] "A Request for Comments on Candidate Algorithms for the Advanced Encryption
Standard (AES)", Federal Register, Volume 63, Number 177, September 14, 1998.
[10] T. Pornin, "Automatic software optimization of block ciphers using bitslicing
techniques", Paris, France, 1999.
[11] E. Biham, RJ. Anderson, and LR Knudsen, \Serpent: A New Block Cipher Proposal",
in Fast Software Encryption | FSE 98, Springer LNCS, vol 1372, pp 222.
[12] B. Schneier and his colleagues, "The Twofish Encryption Algorithm: A 128-Bit Block
Cipher", John Wiley & Sons, 1999. www.counterpane.com/twofish-paper.html.
[13] A. J. Elbirt, W. Yip, B. Chetwynd, and C. Paar, An FPGA-Based Performance
Evaluation of the AES Block Cipher Candidate Algorithm Finalists, IEEE
Transactions On Very Large Scale Integration (VLSI) Systems, VOL. 9, NO. 4,
AUGUST 2001
[14] R. L. Rivest1, M.J.B. Robshaw and Yiqun Lisa Yin, "The Security of the RC6TM Block
Cipher", RSA Laboratories Version 1.0, August 20, 1998.
[15] M. Y. Rhee, "Internet Security Cryptographic Principles, Algorithms and Protocols",
John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, England, 2003.
[16] O. A. Dawood, A. S. Rahma and A. J. Abdul Hossen, The New Block Cipher Design
(Tigris Cipher), I.J.Computer Network and Information Security (IJCNIS).
133
(ICoIT'17)
10thof April, 2017.
An Overview of Software Defined Environment

Ali Dheyaa Jalil
Networks Engineering Department, College of Information Engineering, Al-Nahrain University - Iraq
ali.dheyaa@coie-nahrain.edu.iq
Dr. Abdulkareem A. Kadhim
Computer Engineering Department, College of Engineering, Al-Nahrain University -Iraq
abdulkareem.a@coie.nahrainuniv.edu.iq

Datacenters are a complex environment consisting of network
Article History: equipment, server hardware and storage systems. The rapid
Received: 17 March 2017 evolution in Information Technology applications and the
widely used services and datacenters provided by different
vendors make them a big challenge. Traditional datacenters
Published: 10 April 2017 are managed by administrators manually. The growth of
Keywords: Networks, datacenters thus reveals problems in managing and
Software Defined controlling. A new technology called Software Defined
Environment, Software Environment (SDE) is introduced to solve such problems.
Defined DataCenters, SDE solution separates the controlling functions of the
Software Defined datacenters from its physical components and performs such
Network, OpenFlow operations in a centralized controller. The present work
protocol, Software introduces an overview of SDE and its features to solve
Defined Storage, common problems in datacenters such as the controlling
Software Defined problems and the management problems in adding an extra
Compute.
physical hardware. Suitable scenarios are suggested with the
assistance of Software Defined Network (SDN) and its useful
features. Possible tools and resources to implement SDE are
also presented.
1. INTRODUCTION
F inding a standard way for controlling and managing the infrastructure of datacenters
is an outstanding demand. Thus the separation between the control plane and data
plane is proposed. Software Defined Environment (SDE) is a new approach where
the software based management and configuration of all resources, based on the
workload, can be achieved. Software Defined Network (SDN) is considered as the main
component of SDE. The latter is a term used by IBM for its "software-defined everything"
vision [1]. A part of SDN, the umbrella of SDE technology covers SD Compute (SDC), SD
DataCenters (SDDC) and SD Storage (SDS). In SDE environment, intelligent software is
responsible of managing and controlling the network, storage and/or datacenter infrastructure
instead of their hardware components [1]. L. MacVittie discussed the idea of Software
Defined and compares it with Object Oriented idea where the implementation of the data
134
(ICoIT'17)
10thof April, 2017.
layer is abstracted from the interface the control layer [2]. So it is easy to make
implementation changes with modifications in the interface.
The idea of software defining the datacenter components is discussed by several works.
V. Trhnen has designed his own SDDC by utilizing multi-vendor hardware [3]. His
proposed datacenter is based on VMware as the virtualization layer, HP server hardware as
the computing component, NetApp storage, and HP and Cisco networking equipment. All
the components provide an Application Programming Interface (API) that is used to control
the datacenter [3]. VMware in 2015 introduced an architecture using their software
components to create an SDDC as follows [4]:
- VMware vSphere: as a Server virtualization platform (instead of Software defined
compute).
- VSAN and vSphere: VSAN provides distributed storage capabilities and vSphere is
responsible of the Policy-based storage capabilities.
- VMware NSX: NSX provides a platform for Network virtualization (instead of
SDN(.
Until the present time there is no emulator designed to emulate SDE, SDS or SDDC.
Mininet is one of the widely used tools for emulating SDN [5]. Mininet is an emulation
tool which can build a virtual SDN network consists of hosts, switches, controllers, and links.
Mininet switches support OpenFlow protocol for more flexible customized routing and SDN
applications.
Darabseh et. al. customized Mininet to emulate several Software Defined experimental
frameworks [6-9]. They customized Mininets elements by adding storage parameters,
tables, and functions to the hosts, switches, and controllers, respectively to build SDS_host,
SDS_switches and SDS_controller. They proposed an SDsecurity framework in the same
way. They also integrated their SDS and SDSecurity frameworks into a software defined
secure storage. SDDC is built by integrating the SDN (using Mininet), SDS, SDSecurity
frameworks and with another block; the Software Defined compute (SDC) that is considered
as part of the virtualization and the computing resources.
The remaining parts of the present paper are; part-2 covers the concept of traditional
network architecture, while part-3 deals with SDE based network with its useful components
(SDC, SDS, and SDN). Part-4 gives an overview of the available tools for SDE. Finally,
the main concluding remarks are given in part-5.
2. TRADITIONAL NETWORK ARCHITECTURE

In this section, the architecture of traditional networks and its main differences
compared to SDN are introduced. There are some difficulties to manage the traditional
networks. they consist of different types of equipment such as routers, switches and
middleboxes (e.g. intrusion detection systems, network address translators (NAT), load
balancers and firewalls). Routers and switches run a control software that is very complex
and usually locked and proprietary. Usually, such software implemented network protocols
which subject tests for standardization and interoperability for years. Administrators can
configure each device with its configuration interfaces. These interfaces are different from
one vendor to another and also among the same vendor products. This leads to increase the
complexity, slow innovation and inflate both the capital cost and the cost running the network
[10].
135
(ICoIT'17)
10thof April, 2017.
Datacenter is a facility used to house servers and computer hosts which are connected
using dedicated switches and links. Current Datacenters may contain huge number of hosts
with bandwidth and delay requirements [11]. The topology of the network has an important
effect on the agility and reconfigurability of the datacenter infrastructure to respond to
changing application demands and service requirements. Basic tree topology consists of
either two or three levels of switches or routers, with the host as leaves. In 3-tier topology;
at the root there is the core tier, at the middle there is the aggregation tier, and at the bottom
there is the edge tier where the switches are connected to the servers. In a 2-tier topology,
the aggregation tier does not exist [12]. A 3-tier tree topology is shown in Figure-1.
FIGURE 1. 3-Tier datacenter tree topology
A more complex type of tree topology called Fat Tree topology is used by different works
[12, 13]. In Fat Tree a large number of routers and switches are used where the traffic is
forwarded based on the limited view of each network device. This networking model is
unable to meet the proper flexibility required with the current massive data amounts. The
coupling between the software and the hardware of the routers and the switches leads to an
expensive scaling and complexity in maintenance as well [14].
3. SDE BASED NETWORK

In this section the Software Defined Environment (SDE) and its main components are
presented. SDE is a general term which covers multiple different products, technologies and
solutions. Defining what the needs are will determine the selection of correct components.
These can be put in three main areas [1]:
3.1 CLOUD WORKLOAD ORCHESTRATION AND MANAGEMENT

It is the foundation of SDE environment. IBM SDE with OpenStack provides a platform
for cloud computing and enables configuration of clouds, application workloads to each
deployment topology along with the other SD technologies [1].
3.2 SOFTWARE DEFINED COMPUTE
136
(ICoIT'17)
10thof April, 2017.
SDC term is used to describe the functions of server virtualization technology. These
serve as resources to virtual machines guests [1]. Hypervisors can be used as an SDC in the
SDE. Examples of these hypervisors are KVM for Linux Operating Systems [15], Hyper-V
for Microsoft [16], and ESXI vSphere from VMware [17].
3.3 SOFTWARE DEFINED STORAGE

SDS is new storage architecture for wide variety of data storage requirements based on a
set of loosely coupled software and hardware components. IBM admitted that the SDS
architectural concepts are new and evolving technological approaches. So it is not currently
fully developed into mature, straightforward implementations [18]. The main idea of SDS
is that all storage functions are embedded in software so there is no need for sophisticated,
dedicated storage equipment. Thus, SDS will be more flexible than disk arrays or the
traditional hard disks [1]. The main characteristic of the SDS reference architecture is the
separation of the storage functions into two main layers:
- SDS control layer: This layer is responsible for managing the storage resources. It is
based on a software that provides flexible, scalable, optimized, and rapid provisioning
storage infrastructure capacity.
- SDS data layer: This layer is responsible for processing the data. The data layer
represents the interface of the data storing hardware infrastructure. It provides range
of data access possibilities and spans the current access methods such as block I/O
(e.g. iSCSI) [18].
There is no specific emulator to emulate and test SDS. An experimental framework
emulator for SDS systems was presented but with the use of SDN Mininet emulator [6]. By
following such approach the Mininet SDN elements are customized to make their framework
cover the following components:
- Host: where some extra SDS parameters are added to the hosts, such as no. of
directories, no. of files inside each one, and the size of each file.
- Switch: where a table is created to store the updating hosts information about their
status while keeping such information available to other hosts.
- Controller: where POX controller is used with some functions being added like "Get
Number of Stored Files, isFull, used space, addFile and available space" [6].
3.4 SOFTWARE DEFINED NETWORK

SDN is a new network architecture where all the networking functions and tasks are
performed by software in a controller (the control plane) separated from the network
equipment (the data plane) like switches and routers. The limitations of the current network
architectures are; scalability problem, different vendors interfacing problem, and
configuration management problem. The idea of programmable network is older than SDN.
Active Network has common visions with SDN, it showed a programmable networks, but
with an emphasis on programmable data planes. Active network programming interface
exposed the resources (e.g., processing, storage, and packet queues) on each node of the
network. The problem with Active Network is that it has a lack in the clear use case and the
incremental deployment path [10].
The interaction between the control and the data planes is organized by a protocol called
the OpenFlow protocol, which is an open source protocol by the Open Networking
Foundation (ONF) [19]. Each OpenFlow network consists of one or more Open flow
switches; each switch with one or more flow table. Each port and flow table in the switch is
137
(ICoIT'17)
10thof April, 2017.
associated with many counters that gather various statistics describing the events that the
switch must handle. The flow tables and their entries are all created by the controller.
SDN Controllers are considered as the main part of SDN. It acts as strategic point of
controlling, and managing the flow control to the routers or switches. The controller
communicates via southbound API with the forwarding plane, the infrastructure layer.
through the OpenFlow protocol. The controller communicates via northbound API with
application layer to deal with various network applications [20]. Figure-2 shows the SDN
architecture.
FIGURE 2. SDN Architecture
There are several emulators that can emulate the SDN networks, such as Mininet [5], ns-
3 and Estinet [21]. Mininet is widely used by SDN researchers because it is simple, open
source and written in Python programming language. It is based on Ubuntu Linux
distribution. Mininet consists of three main components: the host, which sends and receives
the packets, the switch, which stores the rules for forwarding the packets to their destinations,
and the controller, which is responsible for the operations of controlling and managing the
network [5]. Mininet also gives the ability to connect the SDN network to a remote
controller. There are several controllers that can be connected to Mininet, such as NOX,
POX, Beacon and Floodlight. NOX is the first SDN controller developed at Nicira networks
[22]. The main difference between NOX and POX is that POX is based on Python
programming language, while NOX is based on C++ programming language [23]. Beacon
and Floodlight are written in Java programming language [24,25]. The Mininet emulator has
a problem of its inability to handle large-scale networks. MaxiNet is created as an extended
version of Mininet for large scale networks. Its main idea is to build a distributed emulation
environment [26].
138
(ICoIT'17)
10thof April, 2017.
4. COMMON TOOLS FOR SDE

The available tools for SDE are introduced in the following;
4.1 SDE MANAGEMENT TOOLS

The OpenStack project started in 2011 as the main component in cloud management [27].
It is developed by the large number of companies and developers all around the world. The
OpenStack project used Apache 2.0 license and runs on variety of Linux operating systems.
OpenStack is software that controls the resources of datacenter such as the compute, storage,
and networking resources. These resources are managed through the OpenStack API.
OpenStack can work with well-known enterprise and open source technologies that make
OpenStack as a good management tool for heterogeneous infrastructure [27]. It is worth to
be noted here that an excellent programming proficiency is required when operating
OpenStack API.
.
IBM suggested their tools to manage clouds and deploy orchestrated services for the
clients cloud environment. One of these tools is IBM cloud manager, which is based on
OpenStack, easy to use and to deploy cloud management software, with additional IBM
enhancements and enterprise class support [1]. It also supports many different hypervisors,
storage, and network solutions.
4.2 SDE COMPUTE TOOLS
Hypervisors can be used as an SDC in the SDE. The main advantage of SDC tools is that
multiple servers can run on one physical server, which translates into more efficient use of
resources. The main disadvantage here is that if there is a problem in this physical server,
all the virtual servers will be affected. Examples of the hypervisors are:
- IBM PowerVM: This is a virtualization technology available on IBM POWER
servers and can provide virtualized or physical server resources to the virtual
machines. PowerVM enables dynamic, non-disruptive, manual or automatic
resource relocation, such as dynamic CPU or RAM reassignment between different
logical partitions [1].
- Hyper-V: This is a virtualization technology introduced by Microsoft [16]. It runs
on x86 servers and supports Linux and Windows guest virtual machines. It is also
supported by IBM Cloud Manager as one of the underlying cloud technologies.
- ESXI : This is a virtualization technology provided by VMware [17]. It runs on x86
servers and supports Linux and Windows guest virtual machines. It is also supported
by IBM Cloud Manager as one of the underlying cloud technologies. The main
drawback when using such tool is the difficulty in accessing it by the administrator
directly from the working server. vSphere client software need to be installed on
another host and used to access the ESXI.
4.3 SDE STORAGE TOOLS

In SDE, SDS tools run on the physical storage and provide the advantage of multiple
virtual storage pools to the hosts. Different storage products and solutions can be used as
in the following:
139
(ICoIT'17)
10thof April, 2017.
- IBM SAN Volume Controller (SVC) is a well-known and proven, robust storage
solution. SVC can virtualize huge number of external disk arrays to provide common
storage pool to hosts [1].
- IBM Storwize V7000 is a disk storage system which provides block storage to hosts.
Storwize and SVC run the same code with the same features as SVC. It differs from
SVC that it actually handles up to several hundreds of real physical disks, which can
be gathered in RAID arrays and then provide common storage pool for physical or
virtual servers [18].
- As mentioned, VSAN and vSphere can be used together as SDS tools, where VSAN
can provide distributed storage capabilities and vSphere is responsible for the policy-
based storage capabilities [4].
All the above tools are based on the virtualization idea. One of the main disadvantages
of such approach is that a more complex troubleshooting effort is required.
4.4 SDE NETWORK TOOLS

The IBM Software Defined Network for SDE has a set of capabilities. It provides
integrated management of OpenFlow networks and Distributed Overlay of Virtual Ethernet
(DOVE) networks. IBM Software Defined Network for Virtual Environments (SDN-VE) is
software for networking implementation, based on the OpenDaylight open source platform.
SDN VE creates virtual networks that are isolated from physical networks in the same way
similar to how a virtual machine is separated from a physical host. SDN VE integrates with
OpenStack Havana via the neutron network component of OpenStack [1].
Cisco Application Centric Infrastructure (ACI) is an SDN architecture proposed by Cisco
[28]. It is a policy-based automation solution that supports a greater scalability through a
distributed system, and a greater network visibility. These benefits are achieved through the
integration of physical and virtual environments under one policy model for networking,
servers, and storage. Cisco ACI is implemented only with Cisco Application Policy
Infrastructure Controller (APIC) and Cisco Nexus 9000 series switches.
Most of the SDE tools, for instant the storage and server system tools, are vendor-based.
The source code of these tools is usually unrevealed, so there is a difficulty to improve the
system that uses such tools. The improvement is required to support the legacy compute and
storage systems in datacenters. For network tools, there are many open source controllers.
The users can develop their own application to manage and improve their networks. As an
example for such applications is the user security policy.
5. CONCLUSION
Virtualization and Software Defined technologies can be considered as the basic
elements of the next generation datacenters where all the components are either virtualized
or Software Defined. In this paper, the idea of Software Defined Environment (SDE) is
covered. SDE technology can be divided into three main technologies, SDN, SDC and SDS
with an orchestration or management tool. Software Defined Networking (SDN) is the most
well-known technology where the controlling functions are separated from switches and
routers. There are different research topics linked to SDN such as traffic engineering and
security. The idea of Software Defined Storage (SDS) is close to SDN idea where all storage
functions are dealt with in software. Software Defined Compute (SDC) can be considered
140
(ICoIT'17)
10thof April, 2017.
as the server virtualization which is implemented in a physical server to create multiple

virtual servers. Hypervisors can be used to give the functionality of SDC.
It seems that there is no general emulation tool that covers all SDE functions and
components. The advantages of SDE implementation in datacenters can be seen clearly in
the management and scalability of such datacenters. In SDE, there is no need for specific-
purpose devices. Thus SDE seems to be a very promising technology for implementation
when there are multi-vendor devices and requirements for scalability. With SDE, great
reductions in physical space and power requirements of datacenters will be achieved as a
result of software defined and virtualization processes.
REFERENCES
[1] D. Quintero, W. Genovese, K. Kim, M. Li, F. Martins, A. Nainwal, D. Smolej, M.
Tabinowski, A. Tiwary, IBM Software Defined Environment, IBM Red-Book, August
2015.
[2] L. MacVittie, "Stop conflating software-defined with software deployed", Devcentral
2013.
https://devcentral.f5.com/articles/stop-conflating-software-defined-with-software-
deployed (Accessed on 14 March 2017).
[3] V. Trhnen, "Designing a Software-Defined Datacenter", M.S. thesis, Tampere
University of Technology, Tampere, Finland, December 2013.
[4] VMware Team, "Software-Defined Data Center", Technical White Paper, VMware Inc,
2015. (www.vmware.com)
[5] Mininet, Mininet Official Site, [online] 2017, http://mininet.org (Accessed on 14
March 2017).
[6] A. Darabseh, M. Al-Ayyoub, Y. Jararweh, E. Benkhelifa, M. Vouk, A. Rindos,
"SDStorage: A software defined storage experimental framework", in IEEE
International Conference on Cloud Engineering (IC2E 2015), 2015.
"SDSecurity: A software defined security experimental framework", in Third
Workshop on Cloud Computing Systems, Networks, and Applications (CCSNA-2015),
2015.
[8] A. Darabseh, M. Al-Ayyoub, Y. Jararweh, E. Benkhelifa, M. Vouk , A. Rindos, " A
novel framework for software defined based secure storage systems", Simulation
Modelling Practice and Theory 2016.
"SDDC: A software defined datacenter experimental framework", in Third
International Conference on Future Internet of Things and Cloud (FiCloud-2015),
Rome, 2015.
[10] N. Feamster, J. Rexford, and E. Zegura, "The Road to SDN: an intellectual history of
programmable networks", ACM SIGCOMM Computer Communication Review, New
York, April 2014.
[11] S. Seo, "Dynamic Traffic Engineering for Improving Energy Efficiency and Network
Utilization of Data Center Networks", Ph.D. dissertation, Pohang University, Korea,
December 2013.
[12] M. Al-Fares, A. Loukissas, and A. Vahdat, "A scalable, commodity datacenter network
architecture", in ACM SIGCOMM2008 Conference on Data Communication, 2008.
141
(ICoIT'17)
10thof April, 2017.
[13] R. Niranjan Mysore, A. Pamboris, N. Farrington, N. Huang, P. Miri, S. Radhakrishnan,

V. Subramanya, and A. Vahdat, "Portland: a scalable fault-tolerant layer 2 data center
network fabric" ACM SIGCOMM Computer Communication Review, vol. 39, no. 4,
2009.
[14] S.A.Chasib ,Dr. A. A. Khadim. "Software Defined Network for Data Center Using
OpenFlow Protocol ",International Journal of Scientific Engineering and Technology
vol.5 ,no.12, December 2016.
[15] Kernel Virtual Machine, KVM, [online] 2008, https://www.linux- kvm.org, (Accessed
on 14 March 2017).
[16] Microsoft Hyper-V, Microsoft Official Site, 2017.
https://www.microsoft.com/en-us/cloud-platform/server-virtualization (Accessed on
14 March 2017).
[17] vSphere ESXI, VMware Official Site, 2017.
http://www.vmware.com/ products/esxi-and-esx.html (Accessed on 14 March 2017).
[18] L. Coyne, M. Defiebre, P. Gilmer, S. Gopalakrishna, M. Lindstrm, B. Mann, G.Perilli,
M. Portak, IBM Software-Defined Storage Guide, IBM RedBook, January 2016.
[19] N. McKeown, T. Anderson, H. Balakrishnan, G. Parulkar, L. Peterson, J. Rexford, S.
Shenker, and J. Turner,"OpenFlow: enabling innovation in campus networks" ACM
SIGCOMM Computer Communication Review, vol. 38, no. 2, 2008.
[20] What are SDN Controllers (or SDN Controllers Platforms)?, SDx Central, 2015.
https://www.sdxcentral.com/sdn/sdn-controllers/ (Accessed on 14 March 2017).
[21] S. Wang, C. Chou, C. Yang, "EstiNet OpenFlow Network Simulator and Emulator",
IEEE Communications Magazine, vol.51, no.9, pp.110-117, 2013.
[22] N. Gude, T. Koponen, J. Pettit, B. Pfaff, M. Casado, N. Mxkeown, "NOX: Towards an
Operating System for Networks" ACM SIGCOMM Computer Communication Review,
Vol. 38, No. 3, 2008.
[23] M. McCauley, A. Al-Shabibi, "POX Wiki".
https://openflow.stanford.edu/display/ONL/POX+Wiki (Accessed on 14 March 2017).
[24] D. Erickson, "What is Beacon?", [online] 2013, https://openflow.stanford.edu
/display/Beacon/Home. (Accessed on 14 March 2017).
[25] Floodlight, Project Floodlight official Website, 2017.
http://www.projectfloodlight.org/floodlight/ (Accessed on 14 March 2017).
[26] P. Wette, M. Draxler, A. Schwabe, F. Wallaschek, M. Zahraee, and H. Karl, "MaxiNet:
Distributed emulation of software-defined networks", in 2014 IFIP Networking
Conference, Norway, June 2014.
[27] OpenStack, OpenStack Official Website, [online] 2017, https://www.openstack.org/
(Accessed on 14 March 2017).
[28] Cisco Application Centric Infrastructure, Cisco Official Website.
http://www.cisco.com/c/en/us/solutions/data-center-virtualization/application-centric-
infrastructure/index.html (Accessed on 17 March 2017).
142
(ICoIT'17)
10thof April, 2017.
Building New Tool for Software Project Planning Based

on Meta-modeling Concepts and CMMI Meta-Standards
Husam Bofaoor
Software Engineering, computer Engineering , Damascus University- Syria
husambofaoor@hotmail.com

CMMI (Capability Maturity Model Integration) for
Article History: development is a quality system targeting to improve the
Received: 7 March 2017 software development processes. It allows to measure
organization maturity level and to improve its capacity in
Accepted: 1 April 2017 software development.
Published: 10 April 2017 MDE (Model Driven Engineering) is an approach that aims
to automate the use of models in order to build more complex
systems
Keywords: CMMI The software industry became a vital economic sector that
(Capability Maturity employs a wide range of people. A quality assurance system
Model Integration), becomes a real need to ensure the quality of software and to
MDE (Model Driven increase abilities of software companies to meet the
Engineering), GMF requirements of customers.
(Graphical Modeling The paper aims to use a MDE approach to develop tools
Framework), WBS supporting any software company to adjust their software
(Work Breakdown development process and then to meet the CMMI goals.
Structure), PP (Project this paper suggests a new structure of software development
Planning). process in several models, and shows how to specify a meta-
model for each step of this system.
One of the advantages of the approach is very resilient and
can support the software process development for several
software companies and fit rapidly the changes in their
software development process.
1. INTRODUCTION
S
oftware companies depend more and more on enhancing and improving their work
procedures and processes to achieve their quality objectives.[1]
CMMI offers to software companies the ability to improve their software quality
system by exercising specific practices related to specific and generic goals to reach certain
required maturity level in software development process [2][4].
CMMI is a quality standard whose target is to improve the software development process. It
allows to measure the organization maturity level and to improve its capacity of software
development. This meta-standard directs software companies to establish theirs processes in
order to develop high quality software products. CMMI defines how to enhance and evaluate
143
(ICoIT'17)
10thof April, 2017.
processes among a set of activities called Process Areas which fully support software
development operations (like Project Planning and Management, etc.) [3][5].
CMMI meta-model allows software companies to measure their maturity level and then to
determine the needed set of best practices to improve their capacity of software development,
and to enhance quality of their processes and products [6].
So, CMMI meta-model offers:

Defined structure of practices to follow in order to reach a selected maturity level.
Documented procedures and processes to enhance selected maturity level.
Set of control and evaluation processes to validate improvements.
It is important to note that CMMI identifies a set of practices for a software project
development and gives procedure examples to follow [7], but it doesn't guarantee the
complete success of software projects.
We should also remember that models are abstractions of systems. High levels of abstraction
allow easier understanding and handling of systems. Moreover, new approaches like the
MDE [12] aim to automate the use of models. MDE encourages the identification and the
separation between the different system functionality. MDE structures the application
development in several models and model transformations. The MDE framework specify a
meta-model for each system aspect. So, it can design separately the system different
functionalities through the definition of system models [12].
After that, the MDE framework builds the whole system through the model transformations
[12]. These transformations permit the integration of system different functionalities. Briefly,
MDE allows us to specify a methodology for defining problems and how to go towards
solutions [13].
This paper provides a model-driven framework of software development process with

respect to CMMI guidelines [8]. This framework is supported by modeling tools which are
automatically produced themselves by meta-models. Each meta-model defines the concepts
related to each CMMI guideline. This paper discusses how to meta-model some of CMMI
guidelines. Then, it will show how to organize the produced framework tools.
Section 2 of this paper presents briefly CMMI maturity levels of and their related Process
Areas. Section 3 extracts requirements of CMMI tool for software project planning and
management. Section 4 describes Meta-Model that it used to build CMMI tool. Section 5
presents implementation strategy of suggested tool. Finally, in section 6 displays case study
to use the proposed model.
2. CMMI maturity levels (MLs):

CMMI for development has 22 process areas distributed into 4 categories [9]: Process
Management, Project Management, Engineering and Support as it is described in the
table of figure 1.
144
(ICoIT'17)
10thof April, 2017.
M Process Project Engineering Support

L Management Management
2 -Project Planning (PP) -Requirement -Configuration

Management (REQM) Management
-Project Monitoring and
Control(PMC) (CM)
-Supplier Agreement -Process & Product

Management (Sam) Quality Assurance
(PPQA)
-Measurement and
Analysis (MA)
3 -Organizational Process -Integrated Project -Requirements -Decision Analysis

Focus (OPF) Management (IPM) Development (RD) and Resolution
(DAR)
-Organizational Process -Risk Management - Technical Solution (TS)
Definition (OPD) (RSKM)
-Product Integration (PI)
-Organizational Training
(OT) -Validation (VAL)
-Verification (VER)
4 -Organizational Process -Quantitative Project

Performance (OPP) Management (QPM)
5 -Organizational -Causal Analysis and

Innovation and Resolution (CAR)
Deployment (OID)
Figure 1: CMMI for Development process Areas.

The CMMI maturity level 2 has 7 process areas. One of them is project planning process
area.
As we know, project management phase is one of the important stages in the development
of software products. Most of the software companies suffer from weaknesses in managing
their software development projects to deliver products of high quality. The majority of these
weaknesses are the result of bugs in work processes implementation. So, CMMI can give
us a new thinking manner to establish work processes to manage software projects in order
to get rid of problems and obstacles that could occur during software development process.
2.1 Scope of work

CMMI contains five levels, each level contains many of process areas.
Every process area is supported by a set of generic and specific goals, each goal has a set of
practices.
In level 2 of CMMI cover all management process in any software company, this level
contains seven process areas.
145
(ICoIT'17)
10thof April, 2017.
Figure 2 shows an outline of the process areas related to project management and their
relationships
Fig 2 Process Areas related to project Planning

2.2 Weaknesses in software project management
The process of project management may suffer from several problems, such as:
1- Delays in product delivery.
2- Misunderstanding of customer requirements.
3- Non-accurate estimation of project cost and time.
4- Lack of teamwork commitment according to software engineering principles.
To overcome these obstacles, the user should introduce improvements in project
management work processes and develop new work platform for project planning according
to CMMI recommendations.
146
(ICoIT'17)
10thof April, 2017.
3. Requirements of CMMI tool:
The process of manufacturing of a software tool for project planning and management
must be based on several important requirements [11]. These requirements are:
3.1 Work Breakdown Structure (WBS)
WBS helps in determining the list of actions that describe all project activities. These
activities are not the same and vary from one project development stage to another. So, these
activities must be determined carefully for each stage, to provide project manager and
company administration the ability to fully control their projects. Although it is not easy to
determine these activities at the beginning of the project, this task becomes more straight
forward at advanced stages of project development process.
3.2 Project Schedule
Project Schedule is a timeline of project execution. It depends directly on the list of

activities specified previously during WBS phase.
In addition to project activities, the project schedule has many other factors that play an
essential role in determining project schedule, like:
- Start and end dates of each activity.
- which resources are assigned to perform it.
- availability dates of assigned resources.
- dependencies between activities.
Any change at the level of project activities and any change at the level of time estimation
for project activities, will necessarily affect project schedule.
3.3 Estimates
Estimation process plays an essential role in project planning. This process aims to
estimate cost and effort that needed to complete each activity of the project.
Estimations are determined by using many methodologies such as: Planning Poker, Function
points, use-case points [3].
Estimation should be done at the level of:
1- Project activities.
2- Management, training and follow-up tasks.
3- Change requests.
For each one of these estimation operations should fallow specific steps to provide good
results. Also the estimation takes consideration that any change in WBS project activities
will affect their estimation which will in its turn affect project schedule.
3.4 Tracking and monitoring project activities
147
(ICoIT'17)
10thof April, 2017.
This activity is used to monitor and trace all changes and projects activities, such as:
detections of non-compliances, change requests, bugs, defects This activity aim to ensure
the good study of all discovered issues related to the project and to perform the appropriate
amendments as required.
4. Using Meta-Modeling Concepts to Build CMMI Tool:
The Paper target to define an approach to build development processes conform to CMMI
recommendations and dedicated for several companies. And to built process to be easily
modified according to the changes of company work and/or to be easily adapted to other
companies. So, the CMMI tool creates a meta-model for each adopted CMMI process area,
then produces modeling tools starting of the meta-models for all the desired phases. The
models defined using some produced modeling tools are inputs to other produced tools. The
CMMI recommendations are reflected into the built process by using meta-modeling
techniques. This approach is flexible because the user can easily change the process tools by
introducing the changes in the meta-models that drive the development process, Then re-
produces the modeling tools.
Now, the paper will define the conceptual diagrams of project planning procedures and
practices as recommended by CMMI. Then, it will apply meta-modeling techniques to build
process tools as it is shown in the following sections.
4.1 Conceptual Diagrams compatible with Project Planning Process Area

Figure 3 shows the conceptual diagram related to CMMI project planning process area.
This diagram contains the following classes:
Estimation
Project phases
Project schedule
Procedures
CMMI Pas
PA Goals
Goal Practices
To build a project schedule, the tool needs to define its tasks. These tasks are extracted as
follows: the tool starts with the class of project phases which are the definition of key stages
of the project. Each project phase is associated with class Procedures that represents a
range of activities (Activities) to be executed sequentially and interconnected one to many.
Procedures in their turn are defined by work stages containing a number of actions.
148
(ICoIT'17)
10thof April, 2017.
Activities depend on defined goals and practices in the CMMI Platform.
Estimation Project Schedule
CMMI PAs Project Phases
PA Goals Procedures
Goal Practices Project Activities
Figure 3: Conceptual diagram of CMMI project planning PA.
Figure 4 presents the conceptual diagram to generate tracking reports. This report should
contain:
Basic data date of the report and project stage.
Automatically collected data about: risk, general problems, inconsistencies,
deficiencies, change requests, defects.
Risk Bugs NCs
Commitme
Tracking Report Meetin
nt
g
Generic Defects Change request

Issues
Figure 4: Conceptual diagram to generate tracking reports.
4.2 Specific goals of project planning
- SG 1: Establish Estimates
Estimates of project planning parameters are established and maintained [1].
149
(ICoIT'17)
10thof April, 2017.
Classes that apply the specific

Specific practice Name
practice in CMMI tool
Establish a top-level work breakdown
structure (WBS) to estimate the scope of Project Activities
the project.
Establish and maintain estimates of work -Estimation
product and task attributes. -ProceduresProject Activity
Link(1 n)
Define project lifecycle phases on which -procedures
to scope the planning effort. -Project Phases
-Project Schedule
-ProceduresProject Activity
Link(1 n)
-Project Phases Procedures
Link(1 n)
-Project Schedule Project
Phases
Link(1 n)
Estimate the projects effort and cost for
work products and tasks based on Estimation
estimation rationale.
- SG 2 Develop a Project Plan:
A project plan is established and maintained as the basis for managing the project [1].
Establish and maintain the projects -Project Schedule
budget and schedule. -Estimation
Identify and analyze project risks. -Risk
Plan for the management of project data. -CMMI PAs Project Activity
Link(1 n)
-PA Goals Project Activity
Link(1 n)
-Goal Practices Project Activity
Link(1 n)
Plan for resources to perform the project. Schedule
Plan for knowledge and skills needed to - Project Schedule
perform the project. - Procedures
Plan the involvement of identified - Project Schedule
stakeholders. - Project Activity
Establish and maintain the overall project - Project Schedule
plan. - Tracking Report
- SG 3: Obtain Commitment to the Plan

Commitments to the project plan are established and maintained [1].
150
(ICoIT'17)
10thof April, 2017.

Review all plans that affect the project to -Project Schedule
understand project commitments. -Estimation
Adjust the project plan to reconcile -Generic Issues

available and estimated resources. - Meeting
Obtain commitment from relevant -Commitment
stakeholders responsible for
Performing and supporting plan
execution.
5. Implementation of CMMI tool

Eclipse Graphical Modeling Framework (GMF) use to produce CMMI tools. GMF
produces graphical modeling tools starting from EMF (Eclipse Modeling Framework)
meta-models [13]. So, to express business requirements as EMF meta-models. Then, the
paper is explaining sub relationships and attributes between classes that it needed to build
tools.
- In the project planning defined class activity, this class relation with class
estimation by one to one, in class estimation defined three attributes, these attributes
are time, cost and effort.
- Also the project planning defined class Project schedule, this class contains
attributes, these attributes are: task id, task name, start task date, end task date and
human resources. This class relation one to many with class Project phases, the
project phases contains attributes, these attributes are phase number and phase
name.
151
(ICoIT'17)
10thof April, 2017.
- The project planning defined class Procedure, this class contains two attributes,
these attributes are procedure number and Procedure name.
The class Procedure relations one to many with class Project Activities, also
relations with class Project phases by many to one.
5.1 How transform diagrams to generate code
First this paper used hierarchical diagram presented in figure 5 to handle any CMMI
process area.
152
(ICoIT'17)
10thof April, 2017.
Application
PP
SG1 SG3
SG2
Figure 5: Hierarchical diagram to handle a project planning PA.
it shows in the following sections how it did to integrate GMF with CMMI meta-model
previously presented.
so, it will be able to build special model according to CMMI meta- model taking into
account all constraints in this meta- model.
The following diagram shows how the user can drag and drop any specific goal that he needs.
Figure 6: interface for implementation of PP_ Process Area.

By double click on SGs the user will have PP_SG1 execution interface. So, the user can
drag and drop any class related to any attribute and any relation he needs to take into account
for meta model constraints, as the display in the following figure.
153
(ICoIT'17)
10thof April, 2017.
Figure 7: interface for implementation of PP_ SG1.

Finally, after building special model the XML file is generated from meta model code
(GMF) tool, the XML file contains all attributes and information that the user needs them to
use the new application that is compatible with CMMI standard. The following figure
display sample on XML file and displays the information of SG_1:
Figure 8: Sample of XML file that is exported from CMMI Tool.
154
(ICoIT'17)
10thof April, 2017.
6. Case study
The paper displays how apply the ideas on software companies, MIS (Management
Information System) is software company in Damascus Syria, this company applied
CMMI tool in project planning process area.
MIS sets the attributes that are Compatible with the MIS requirement, so the tool directly
builds detailed class diagram, this class diagram is converted to XML File by meta-model
code, after that import the xml file to eclipse java tool (for example) to parse the content of
xml file, and the user can use the application easily.
Now, the detailed class diagram for MIS company in project planning Specific Goal_1 like
the figure 9:
FIGURE 9: DETAILED CLASS DIAGRAM FOR MIS COMPANY OF PP_ SG1
The detailed class diagram is changed automatically when the requirement is changed for
another software company, thats mean the software companies just sets the attributes that
are Compatible with the its requirement and all the next steps are built automatically by
CMMI tool.
7. CONCLUSION
CMMI is concept of how make software industry and MDE is an approach that aims
to automate use models in order to build more complex systems,
This paper displays how the software company builds its projects compatible with
CMMI Standard, and displays Flexibility when the software company changed its
requirement, the CMMI tool automatically builds new detailed class diagram after that
export the final XML file that is mean the software companies just define the
requirement and another steps are generated automatically by CMMI tool.
Also the CMMI tool can easily add new phase or modify any phase concepts used in
defining the software product model.
155
(ICoIT'17)
10thof April, 2017.
The paper explained approach on the Project planning process area in the CMMI
maturity level 2, the idea can be applied for all process area in five level in CMMI
Standard.
8. REFERENCES
[1] Mary Beth Chrissis And Mike Konrad And Sandy Shrum, CMMI For Development,
Version 1.2. CMMI Second Edition Guidelines for Process Integration and Product
Improvement.
[2] Ahern, Dennis M.; Clouse, Aaron; & Turner, Richard.
CMMI, Distilled: A Practical Introduction to Integrated Process, Improvement, Second
Edition. Boston: Addison-Wesley,2003.
[3] Ahern, Dennis M.; Armstrong, Jim; Clouse, Aaron;Ferguson, Jack R.; Hayes, Will; &
Nidiffer, Kenneth E. CMMI,SCAMPI Distilled:
Appraisals for Process Improvement.Boston: Addison-Wesley, 2005.
[4] Chrissis, Mary Beth; Konrad, Mike; & Shrum, Sandy. CMMI:Guidelines for Process
Integration and Product,Improvement. Boston: Addison-Wesley, 2003.
[5] Crosby, Philip B. Quality Is Free The Art of Making .Quality,Certain. New York:
McGraw-Hill, 1979.
[6] Curtis, Bill; Hefley, William E.; & Miller, Sally A. The People Capability Maturity
Model Guidelines for Improving the Workforce. Boston: Addison-Wesley, 2002.
[7] Deming, W. Edwards. Out of the Crisis. Cambridge, MA:MIT Center for Advanced
Engineering, 1986.
[8] Dymond, Kenneth M. A Guide to the CMMI: Interpreting the Capability Maturity Model
Integration. Annapolis, MD:Process Transition International Inc., 2004.
[9] Dymond, Kenneth M. A Guide to the CMMI: Interpreting the Capability Maturity Model
Integration. Annapolis, MD:Process Transition International Inc., 2004.
[10] CMMI for Development, Third Edition, Addison-Wesley Professional, by Mary Beth
Chrissis, Mike Konrad, Sandy Shrum, 2011.
[11] SEI 2006b CMMI Product Development Team. SCAMPI v1.2, Standerd CMMI
Appraisal Method for Process Improvement, version 1.2: Method Definition Document.
[12] Model Driven Architecture Applying MDA to Enterprise Computing, Wiley Publishing,
Inc, David S. Frankel 2003
[13] The Eclipse Graphical Editor Framework (GEF), Pearson Education, by Dan Rubel,
Jaime Wren, Eric Clayberg, 2012.
156
(ICoIT'17)
10thof April, 2017.
Computer Aided Diagnosis System for Breast Cancer using

ID3 and SVM Based on Slantlet Transform
Mohammed Salih Mahdi
BIT, Business information College, University of Information Technology and Communications- Iraq
Ms.mahdi99@gmail.com

Lately, the woman's chest cancer is the 2nd reason of deaths
Article History: in females. Mammogram images are medical images, which
. .
Received: 20 March 2017 can be read by physicians to detect breast carcinomas. In this
paper, proposed Computer Aided Diagnosis system that can
assist the doctors in hospitals to improve the diagnosis of the
Published: 10 April 2017 disease to detect cancer cells. Enhance the undesirable effects
the of mammogram images by using slantlet transformer , set
. .
Keywords: Breast cancer, of different stages and classifies as normal, abnormal

CAD, Slantlet, ID3, SVM according to ID3 and SVM. For the same testing set, the
practical outcomes displays SVM classifier with an
accuracy of 95% and ID3 classifier with an accuracy of
92% based on MIAS database.
INTRODUCTION
T
he Medical im ages helpful in the education area for learners by clarifying these
-
im ages will assist them in their projects [1]. The explanation of medical im ages
- -
acts the most significant and exciting part of technology. For many experts,
aiding physicians is becoming of high and advantage value by evolving a computer
assisted diagnosis for carcinoma [2]. Mammogram im ages are top-resolution x-ray
. . -
imaging of breast. This includes ray transmission during tissue and the dropping of
.
anatomic structures on a lm display or im age sensor. A linked with the x-ray imaging
-
dropping is a decrease in anatomic data into a 2D im age. Studies have display that the
-
death rate could decrease by 34% if all females age fifty and older have regular
mammogram [3]. In exercise of mammography imaging, there exist essentially two kind
. .
of normal tissue differentiable in the im ages normal and abnormal. A sample

-
mammogram displaying the breast anatomy is shown in Figure.1 [4].

. . .
157
(ICoIT'17)
10thof April, 2017.
Figure 1 :Mammographic breast Anatomy[4]

.
Some proposals have been drawn from the scientific literature: Sara D. And Abbasi,
M. [2], breast cancer is scourge diseases amongst females, needing CAD system in
.
diagnosing tumors of br east cancer and helping medical technicians using classifier
.
based on SVM. Subbiah B. [5], reduce the number of false positives and false
negatives. by implemented a CAD that can help the radiologists. Leena J. Baskaran S.
and Govardhan A. [6] building a CAD for classification in digital mammograms using . .
contourlet transform and SVM.
THEORY
The transformation is a procedure that converts an object from scope to another scope so
as to have some paramount implied data, that used for CAD system [7] The aim of
transformation is to replace the data from time-space scope to time-frequency scope, that
produce leading compression outcomes [8]. The information on spatial scope is generally
less fixed than frequency scope. Thus, presented lazy and complex in spite of providing
higher accuracy [9].
Slantlet transform (ST) is founded based on an improved version of the usual discrete
wavelet transform filter bank where the support of the discrete-time basis functions is
reduced. However, Slantlet transform (ST) is according to an amended version of the
discrete wavelet transform (DWT) where reducing the backing of DWT-time basis
functions. For variant scales, ST designing variant filters unlike iterated filter bank
technique for the DWT. For each level, It utilizes distinct filters while DWT is
performed in the form of an iterative filter bank. [10]. Some feature of ST are [11]:
ST is a multi-resolution and better time localization.
158
(ICoIT'17)
10thof April, 2017.
ST better compression as compared to DCT

Filter bank of ST is orthogonal.
Filter bank of ST-Level , using the down-pointing arrows ( ) as shown in Figure (2)
appear down-sampling by 4 and transform matrix same size of im age. -
Figure 2: Analysis ST-Level2 filterbank and transform matrix.
The in equation (1,2,3,4), filters coefficients of ST level 2 filter bank can be gained
as [12]:
H2= [0.27 , 0.395, 0.52, 0.645, 0.23, 0.1, -0.01, -0.148] ------------------------- (1)
F2= [-0.083, -0.12, -0.79, -0.197, 0.753, 0.344, 0.064, -0.474] ----------------- (2)
G1= [-0.512 , 0.828, -0.12, -0.195] ------------------------------------------------ --- (3)
G1r= [-0.195, - 0.12, 0.828, 0.512] ----------------------------------------------------- (4)
159
(ICoIT'17)
10thof April, 2017.
The decision tree is a common approach for pattern classification and consider as
a class-labelled. Decision tree same to tree structure, so very inner node indicates a test
on an attribute, each branch acts an result of test, and any leaf node catch a class label [13]
.
. decision tree algorithm by Quinlan Ross presented in the year of 1986. ID3 utilizes
information gain measurement for choosing the attribute of splitting. It only accepts
categorical attributes in constructing a model of tree. 3 widely known diversity functions are
depicted below. Let there be a data-set S (training data) of C outputs. Let P (I) represent the
proportion of S that belongs to a class I where I is different from one to C for the classification
problem with C class [14]:.
Simple Diversity index = p (I)-------------------------------------------------------(5)
The goodness of a split measures of entropy, which measures the amount of

information on a feature.
Entropy(S) = (() log 2 ()) (6)

=1
Gain(S,A) defined as
||
Gain(S, A)= Entropy(S) (( || ) ()) (7)
Support Vector Machines : the modern technique in machine learning is SVM in recent
years, that is common in many pattern classification issus including texture classification
and pattern recognition . According to different kernels SVM maximizing marginal
range between classes with decision outlines drawn. Binary SVM is intended to job with
only 2 class by determining hyper-plane to split the 2 class. The patterns which nearest
to the margin which was chosed to define hyper-plane which known as support vectors.
[15].
METHODOLOGY
The phase structure of the Computer Aided Diagnosis system for breast cancer (CAD-BC)
.
made up of 2 phase: training and testing. Every one of the steps has predetermined functions.
Every function is depicted in detail in the upcoming sub-sections which can be shown in
following Figure 3. Mammography im ages Analysis Society (MIAS) implemented in this
-
research was taken from , which is a UK research group organization related to the
breast cancer. 250 im ages selected from the data base, 175 im ages used for training and 75
. - -
160
(ICoIT'17)
10thof April, 2017.
im ages used for testing. Enhance the undesirable effects the of mammogram im ages ,
- . . -
improving the fitness of the mammogram im age and produce the feature extraction step
. . -
simple and extra credible by using slantlet transform with clearness , rid from noise and
smooth by blocking detailed information. Region of interest(ROI) is the most important
step in CAD-BC. breast object is segmented from the background using means
.
thresholding segmentation. It is easy to be performed. The conception is to make a

threshold which represented by the mean of all pixels of mammogram im ages .such that. . -
each pixel in mammogram im ages is more than density to threshold remained otherwise
. . -
removed. extracted features are clue of ROI outcomes . For any mam mogram im ages, Gray . . -
Level Co-occurrence Matrix (GLCM) is build with distance=1 and aveage direction of
(0,45,90,135) degree. attributes are extracted based on GLCM are Contrast, Dissimilarity,
Entropy, Homogeneity and Standard deviation. These attributes are the input of the classifier
model . The classifier model built by using ID3 and SVM. The classifier model to test a new
woman's chest im age which is not exist in training set.
-
A- B-
Figure 3: The structure of CAD-BC: A- training of CAD-BC. B- testing of CAD-BC.
RESULTS
The performance of the CAD-BC is estimated by using confusion matrix, running
time and classification accuracy which has been obtained from the testing part by using
slantlet transforms can be shown in a table (1) , table (2) and table (3). Also, im ages quality -
after slantlet transforms measured by PSNR can be shown in a table (4). Table (5). Shows
the Comparison of CAD-BC with related work system and finally illustrated the simple
im ages of stages of mam mogram processing in figure 4.
- . .
161
(ICoIT'17)
10thof April, 2017.
Table 1: Confusion matrix using ID3.

Predicate Class
Actual Class Normal Abnormal
Normal 43 2
Abnormal 4 26
Table 2: Confusion matrix using SVM.

Predicate Class
Actual Class Normal Abnormal
Normal 42 3
Abnormal 1 29
Table 3: Comparison between ID3 and SVM

Class Type ID3 SVM
Running time Accuracy Running time Accuracy
Average 90 second 92% 116 second 95%
Table 4: Performance measures of the Slantlet transforms.

Class Type Slantlet transforms
PSNR
Normal 31.052
Abnormal 30.061
The implementation of CAD-BC compared with other related system with same data
set (MIAS database) as following table.
TABLE 5: Comparison with related work system
Item Accuracy
Sara D. and Abbasi 87%
Subbiah B. 91
Leena J. Baskaran S. and 93%
Govardhan A
Mohammed salih (CAD-BC) 92 % ID3
Mohammed salih (CAD-BC) 96% SVM
162
(ICoIT'17)
10thof April, 2017.
Figure 4: Stages of mam mogram processing

. .
CONCLUSION and RECOMMENDATIONS
True indicator of breast cancer diseases are mammogram im ages. Automated

. . . -
segmentation using means threshold, which extracted the object (breast), For the same .
testing set, the practical outcomes displays SVM classifier with accuracy of 95% and
ID3 classifier with accuracy of 92% based on MIAS database. it is strict to assume that
every individual attribute is superior to others based on outcomes . The five attributes used
in this paper which can be extracted from GLCM matrix, and which contain
information about im age features.
-
Enhancing the woman's chest cancer by using different transforms like framelet.
Appling a various classification techniques to build the classifier like naive , ANN.
REFERENCES
[1] Smitha P., Shaji L. and Mini M., "A Review of Medical Image Classification
Techniques", International Conference on VLSI, Communication and
Instrumentation (ICVCI), Karunagapally, 2011.
[2] Sara D. and Abbasi M., " Br east Cancer Diagnosis System Based on Contourlet
.
Analysis and Support Vector Machine", Department of Computer, Science and Research
Branch, Islamic Azad University, Khuzestan, Iran, 2011.
[3] Yajie S., "Normal Mammogram Analysis", PhD Thesis, Department of Computer
Science, University of Purdue, 2004.
[4] Monika S., "Computer Aided Diagnosis in Digital Mammography: Classification
of Mass and Normal Tissue", MSc Thesis, Department of Computer Science and
Engineering, University of South Florida, 2003.
[5] Subbiah B.,"Contourlet Based Texture Analysis and Classification Of Mammogram
Images", International Journal of Engineering Science & Technology, Vol. 5 Issue 6, p1228,
2013.
[6] Leena J., Baskaran S. and Govardhan A.," A Robust Approach to Classify Micro
calcification In Digital Mammograms Using Contourlet Transform and Support Vector
Machine", JCA (ISSN: 0974-1925), Volume VI, Issue 1,2013.
163
(ICoIT'17)
10thof April, 2017.
[7] Matheel E, "Color Image Denoising Using Discrete Multiwavelet Transform" , PhD
Thesis, Department of Computer Science and Information Systems of the University of
technology, 2000.
[8] Panrong X., "Image Compression by Wavelet Transform ", the faculty of the Department
of Computer and Information Sciences, East Tennessee State University, Masters
Abstracts International, Volume: 40-01, page 0197, 2001
[9] Jing Y., Yong H. and Phooi Y. ,"Recent Trends in Texture Classification: A Review",
Symposium on Progress in Information & Communication Technology, pp. 63-68, 2009.
[10] Ivan W. Selesnick, "The Slantlet Transform", IEEE Transactions on Signal
Processing, Vol. 47, No. 5, May, 1999..
[11] Iman M. Gafer and Afrah L. Mohammed, "Slantlet Transform based Video
Denoising", Baghdad Science Journal, Vol. 8, No.2, 2011.
[12] N. B Patil, V. M. Viswanatha and S. Pande, "Slant Transformation as a Tool for Pre-
Processing in Image Processing", International Journal of Scientific & Engineering
Research, Vol. 2, No. 4, April, 2011
[13] Matthew N. and Sajjan G.," Comparative Analysis of Serial Decision Tree
Classification Algorithms ", IJCSS, 2013.
[14] Lavanya V. and Rani U.,"Performance Evaluation of Decision Tree Classifiers
on Medical Datasets", IJCA, 2011.
[15] C. Chen, C. Chen and C. Chen, A Comparison of Texture Features Based on SVM
and SOM, ICPR, vol. 2, pp. 630-633, 2006.
164
(ICoIT'17)
10thof April, 2017.
Design a Complete Website for the Sustainable

Development Goals (SDG) in Iraq
Amer S. Elameer
Methaq Khamis
Amjed Abbas
The Informatics Institute for Postgraduate Studies (IIPS-ICCI)

17 goals and 169 targets have been announced by the United
Article History: Nations at the summit held in September 2015 to achieve
Received: 20 March 2017 comprehensive development in all social, economic and
Published: 10 April 2017 environmental aspects to improve life. However, the achievement of
these goals requires a change in mentality and culture of design
Keywords: : MDGs development strategies where information plays a key role in its
(Minimum Development success if it has been delivered to communities effectively. Websites
Goals) Indicators, SDGs have active role in providing information for different people free of
(Sustainable charge. The research aims to improve the current developed
Development Goals), condition of Iraq through publishing the cultural awareness of
Web development, Web sustainable development and its goals among people via the new
Design. technology tools, as well as helping professionals in the field of
sustainable development in identifying problems and finding proper
solutions. The problem of research is educating the community to
actively participate in achieving the goals of sustainable
development, especially after the significant decline experienced by
Iraq in achieving the MDG after 2014. The first website is designed
for the purpose of SDGs. This website is designed to publish the
SDGs and their targets as well as the statistics that relate to them.
Different programming languages have been used in designing this
website such as PHP, HTML, CSS, JavaScript, Jquery, AJAX and
MySQL.
1. INTRODUCTION
B rought about scientific developments, technique, revolution era of information and

communication, a lot of achievements that has gotten the attention of observers,
At the same time neglected humanitarian survival chances, newest industrial
progress lot of problems, Concerning nature, Increased rates of extinction of living
organisms, as ozone layer depletion And the contamination of water and air.[1]
As a result of the great success of the Millennium Development Goals in 2002, for the
advancement by the world and the protection of human rights in the present and the future,
without affecting to the environment and the depletion of its resources, Goals came in (2030)
to reinforce those successes and complements it. [2].
165
(ICoIT'17)
10thof April, 2017.
Iraq as other countries works to satisfy targets of these goals through educating the
community to actively participate in achieving the goals of sustainable development,
especially after the significant decline experienced by Iraq in achieving the MDG after 2014
as a result of carrying the burden of the government to achieve those goals without the
participation of community. [3]
Website is one of the most common tools to disseminate information, thus this website is
designed to publish the SDGs and their targets as well as the statistics that relate to them.
Different programming languages have been used in designing this website such as PHP,
HTML, CSS, JavaScript, JQuery, AJAX and MySQL.
2. MILLENNIUM DEVELOPMENT GOALS INDICATORS IN IRAQ

In year 2000, the UN approved eight Millennium development Goals with the view of
addressing the most problems that are begging to be addressed of the time, below are the
most important issues within the goals achieved by The Government of Iraq towards the
Millennium Development Goals: [4]
First goal: The eradication of poverty and hunger:
Iraq was able to achieve a reduction in the poverty rate from 23% to 19%, although it was
aspired to reach 15% by the end of 2015due to the economic prosperity during 2014.but the
wave of displacement due to "ISIS crimes", and the drop in oil prices which have negatively
affected the economic situation rising poverty rate to reach 22.5%.
The second and third goal: To ensure primary education for children (male and female):
The elementary school net enrolment rate has come to 94.7% in 2015, up from 87.3% in
2000
The Fourth, fifth and sixth goals: crises and health status:
The number of deaths of children under five has pointed to 17.4% in 2014
Seventh goal: ensuring the availability of environmental sustainability:
In 2014, 89.3% of the Iraqi population is utilizing an improved drinking water source.
Eighth goal: Creating a global partnership for development:
Internet using has increased from 5.9 % of the total population in 2014 to 15.9 % in 2015.
[5]
166
(ICoIT'17)
10thof April, 2017.
3. WEB DEVELOPMENT
"Web development" usually refers to the main non design aspects of building websites:
writing mark-up and coding. In small organizations may only require a single developer,
while in large organizations and businesses, web development teams consist of hundreds of
people (web developers). There are three kinds of web developer specialization; Front End
Developer, Back End Developer, and Full Stack Developer. [6]
3.1. Front End Development:
It's also known as client side development which is referring to all operations that are
performed in client side. Front end developers are responsible for a websites user facing
code and the architecture of its user activities. The main languages used at this side are
HTML, CSS and JavaScript numerals. [7]
3.2.Back End Development:
It's also called server- side which refers to all operations that done at server side. The back
end of a website consists of a server, an application, and a database which responsible of
make front end possible. The languages used at that side are PHP, ASP.net and MySQL. [8]
4. DESIGN AND IMPLEMENTATION

This website has been designed to help citizens to know the concept of sustainable
development and Specialists in sustainable development to share information about SDG
with others.
The languages that have been used in designing this website are PHP, HTML, CSS,
JavaScript, Jquery, AJAX and MySQL. To design any simple static website there is a need
to use HTML&CSS while building dynamic website by using JavaScript to add interactive
effects in page, MySQL to store data and PHP to retrieve data from database. In this site
jquery and AJAX are used to get good performance because they use less number of codes
in programming that decrease time of processing as well as updating only a section of a page
not all.
4.1.WEB DESIGN:
It's what representation to a client and within seconds they will decide, rightly or wrongly if
you have what they need. Web designing is subset of web development
4.1.1.DATABASE AND TABLES
This website has five tables in MySQL database to store data as illustrated in fig.1 and fig.2.
167
(ICoIT'17)
10thof April, 2017.
Fig.1. Database tables
Fig.2. Table's structure
4.1.2.WEBSITE STRUCTURE:
The information and data gathering process include collecting all the required
information about the SDGs, writing the required needs and preparing all the necessary
resources to build the website.
The information architecture used to describe the planning of a website's structure. There
are two types of navigation structures wide and deep. In wide navigation, the main pages
are all visible together and for small sites, and deep navigation simplifies the main
navigation and groups related pages into categories, which selected for this website,
because it is the optimal solution for large websites as shown in Fig. 3.
168
(ICoIT'17)
10thof April, 2017.
Fig.3. website structure
4.2. IMPLIMENTATION
The main interface of the website is designed by using HTML & CSS codes. The interface
includes many pages that show information and images about SDGs. All components of this
interface are explained below:
Header: Designed by Photoshop CC and consists of the IIPS (Informatics Institute

for Postgraduate Studies) logo.
Navigation Menu: Consist of links to the pages: Home, SDG goals'(each goal has
separate page), Sustainable Development Concepts, Contact Us and Search engine. As
shown in fig.4 and fig.5.
Fig.4. Navigation Menu
169
(ICoIT'17)
10thof April, 2017.
Fig.5 SDGs Link
Page Content: It differs according to the page information. The most pages consist of
texts and images such as Home, SDG concepts and Read more pages. While other
pages like SDG pages consist of:
- Name of goal and image related to that goal in its page as shown in fig.6.
Fig.6 Page of SDG6
170
(ICoIT'17)
10thof April, 2017.
- Statistical, Targets, Indicators of each goal which is designed as a tab within the page.
It is scripted by JavaScript to exploit small space to write a lot of information.
- Each page of SDGs contains either text or input information of visitors that are stored
in a database. The database contains tables with the data of targets of each goal, user
information (name, email, and workplace), comments and replay to comment as
shown in fig.7.
Fig.7 Page of SDG9
News Slideshow: Its a set of images of news animated on same position each of which
has a brief text linked to its full story page. The slideshow is coded by JavaScript, set the
animation time to 5 seconds and animation type to fade.
Contact Form: User can send a message to the email account of the website admin by
this contact form. It contains 3 input fields (name, email address and message). The form
is coded by PHP used mail function to send message and used validation method in case
the user inserts wrong format of the input field. After clicking on the send button an alert
message will appear showing "message sent successfully" message as shown in Fig.8.
171
(ICoIT'17)
10thof April, 2017.
Fig.8 Contact Form
Footer: It is the last part of the website, consists of

Links to the related websites such as: Iraqi Economists net, Iraqi Central Statistical
Organization UNSECO, UNITED NATIONS, and the copyright.
5. CONCLUSION
To achieve the goals of sustainable development, there must be political will and
participation of the fact of all categories of community
In this project, website for SDGs was created as a first site in Iraq that illustrates the
concept of SDG (targets and indicators) to spread cultural awareness of the importance
of sustainable development and its goals, it also works to help the specialists in the field
of sustainable development and those affected in to identifying problems and finding
solutions through displaying the statistical information data and users comments.
In this site: PHP, Jquery, and Ajax languages are used to get more performance for page
in loading. PHP is free fed, Jquery means less number of JavaScript code. Ajax used
when you have database request, can reload only part of page not all.
References
[1] UNEP, Integrating the three dimensions of sustainable development: UNEP, 2015, How
to achieve a balanced, ambitious and inclusive framework.
[2] D. Sachs and Jeffrey, The Age of Sustainable Development, 1st Ed. USA: Colombia
University press 2015.
172
(ICoIT'17)
10thof April, 2017.
[3] A. Neameh, "Transforming our world: the sustainable development agenda for 2030,"
presented at the Good governance and sustainable development goals, Istanbul, 2016
[4] UN, "The Millennium Development Goals Report," New York2015.
[5] Ministry of planning, COSIT. The Millennium Development Goals in Iraq 2014 and view
on post 2015.Iraq: COSIT, 2016.
[6] Hoar and Ricardo, Fundamentals of web development. USA: person education, 2015.
[7] Murugesan, San, handbook of research on web 2.0, 3.0, and x.0:Technologies, Business,
and Social application, USA: published by O'REILY,1st Ed,2010.
[8] Nixon and Robin, Learning PHP, MySQL JavaScript, CSS &HTML5, 3rd ed. USA:
published by O'REILY, 2014.
173
(ICoIT'17)
10thof April, 2017.
Design and Development of a MOOC

for Academic Institution in Iraq
Inas Ismail Imran
Amer S. Elameer
Rafah Al-Hamdani
The Informatics Institute for Postgraduate Studies

Massive Open Online Courses, shortly MOOCs, are a
Article History: phenomenon in online education nowadays because it
Received: 20 March 2017 supports teaching and learning processes in academic
Published: 10 April 2017 institutions. Therefore In this paper a new design of MOOC
website have been developed for Iraqi academic institution;
Keywords: : MOOC, E- specifically for Informatics Institute of Higher Studies. The
learning, programming new website is designed and implemented as an interactive
languages (PHP7, educational platform using PHP7 scripting language (back
MySQL, JavaScript), end code that control front end logic), MySQL to build the
HTML and CSS database, HTML language used to design the layout of
languages, Tools and website, CSS language is used to add style to the web page
Technique of MOOCs. finally JavaScript programming language is also used to add
dynamic elements and action to the web page. The result of
the project is to establish MOOC website based on the general
curriculum provided by the Iraqi ministry of higher education
to spread usage of E-learning in Iraq and also establish the
first E-Iraqi University.
1. INTRODUCTION
E
-learning is one of self- learning patterns in which the individual learns via
computer, smartphone, tablets, and the Internet [1]. E-learning provides almost no
hindrances for self-learning as long as all the requirements are available, therefore
this method became well-known among those who desire to learn via Educational
technology, hence this method is considered as an important part of society today. One of
the well-known terms of E-learning is MOOC system [2].
MOOC term is an abbreviation of Massive Open Online Courses, which represent a
recognized model of self-learning. It is considered as one of the most developed, modernized
methods and the latest evolution of an open education resources [3]. It is an educational
courses and curricula prepared by accredited Universities and educational institutions [4, 5],
moreover; it is free and open on the Internet. MOOC have interesting features which include
feedback, homework assignments, evaluation, testing, and certification [6].
MOOC has developed a high-quality educational system which is accessible to the
masses over the world, therefore many educational experts and technologists predicted that
it will change the patterns of traditional education methods in the near future [12].
Now, international organizations and UNESCO adopt these types of learning and supports
these trends. Although new, adopting such system proved to have a high potential to direct
174
(ICoIT'17)
10thof April, 2017.
the society towards knowledge, increasing growth of learning in the world, and promoting
the concept of lifelong learning [13].
Recent researches and articles showed that MOOC usage had spread all over the world,
hence MOOC is considered as a huge opportunity for learners to increase their knowledge
and experience in various fields [14].
MOOC encourages the exchange of knowledge, ideas and information among the
learners and teachers therefore it contribute to the enrichment of an open educational
environment [15]. Therefore In this paper a new design of MOOC website have been
developed for Iraqi academic institution; specifically for Informatics Institute of Higher
Studies to spread usage of E-learning in Iraq and also to establish the first E-Iraqi University.
2. DEFINITION OF MOOCS
Different definitions have been offered for the term MOOC by illustrating the four
words in the MOOC abbreviation. The key components of MOOCs are shown in Figure (1)
[16]:
1. Massive: massiveness in MOOCs refers to the number of course participants. Whereas
most of the MOOCs had a hundred participants so that some courses reached over
150,000 registrations. Massive reflects the capacity of the course to expand to huge
numbers of learners. The challenge is to find the correct balance between a large
number of participants, individual needs of learners and content quality [17].
2. Open: Openness consist of four dimensions (4Rs) Reuse, Revise, Remix, and
Redistribute. In the terms of MOOCs, it refers to providing a learning experience to a
large number of participants around the world regardless of their location, age and
level of education, without any requirement meaning no course fees to access high-
quality education, also openness can refer to providing open educational resources
(OER) e.g. course notes, video lectures, PowerPoint presentations, and assessment
[18].
3. Online: In MOOCs, the term online refers to the accessibility of these courses from
every spot of the world via the internet connection to provide synchronous and
asynchronous interaction between the course participants. In some types of MOOCs
example (blended MOOCs), learners can learn in part face-to-face as well as the online
interaction possibilities [19].
4. Courses: The term course in higher education is defined as a unit of teaching. In
MOOCs, it refers to the academic curriculum to be delivered to the learners,
Including OER, learning objectives, networking tools, and assessments [20].
175
(ICoIT'17)
10thof April, 2017.
Figure (1): Abbreviation of MOOC (Massive Open Online Courses).
3.COMPONENTS OF MOOC
MOOCs are the definition of dynamic social learning. It provides for participants the
opportunity to interact, connect and engage with others all over the world. MOOCs are
interactive and use quizzes ,assignments, and tests, moreover, videos blogs, wikis, Facebook
pages, twitter and forums to participate with students. Every MOOC course has a syllabus
that lists the learning goals and defines the scope of the topics accompanying with
discussions, assignments, and quizzes or tests [21]. The syllabus also consist of a weekly
schedule specifically for assignment submissions. The contents of an MOOC course consists
of video lectures, suggested reading materials, discussion forums and also assignments and
quizzes [22].
The video section is the most valuable and unique component of any an online course as it
enables an expert (the instructor) to deliver his /her expertise (built up over years of
researching and teaching a subject) in an extraordinarily efficient way. Students generally
dont need to purchase books for these courses, due to the fact all reading material is either
be provided inside the MOOC content or is linked to open access texts[23].
MOOCs offer interactive user forums that assist build a community for the students and the
instructors. Discussion forums in MOOCs aim to create the interaction between course
participants. Learners are able to share and discuss information and opinions with others
from all over the world .
In a typical MOOC, students view video lectures online and interact with one another using
discussion forums. Some MOOC courses involve students to take an online quiz or test with
multiple choice answers that can be graded automatically.
The assignments are an essential component of MOOCs .Some MOOC courses need students
to finish assignments and upload the answers into the MOOC platform. These assignments
may be evaluated and graded automatically or by instructor [24].
4. MOOC TECHNIQUES
MOOC techniques may differ from one platform to another, also in two different
courses within the same platform. Most techniques of the MOOC are the same as those used
176
(ICoIT'17)
10thof April, 2017.
in traditional classrooms, such as having a start and end time, quizzes, assignments, final
exam and certificate. But, MOOC platform might handle these differently depend on the
tools and technologies available to MOOC. Another technique that might be hard to apply in
classical classrooms and can be used in MOOC, is to create a community among learners by
integrating social media in the communication between participants [25]. In this section, we
discuss some of the main MOOC techniques in three categories. Although each technique is
listed under one category, some techniques might have an overlap with other categories [26].
4.1 Design Techniques
Design techniques are those that used the MOOCs connectivism theory as the main
method. They also draw the basic of MOOC principles and values to be achieved. Here are
some of these techniques :
Personalized curriculum: Course content of MOOC is designed mainly for online use
so that the lecture is divided into smaller parts, each of which represents a topic [27].
Hence, these topics could be taken independently based on the students need,
knowledge and available time. Moreover, each student can assessment their gained
knowledge at the end of each topic, either in a small quiz or in questions. Due to this
way, the student could focus on the actual topic rather potentially being lost in the
middle of the class while the instructor moves on to other topics [28].
Active retrieval practice: Incorporating the recovery of knowledge into MOOCs
improves the learning process. As research suggested, retrieval of knowledge is not
simply restating the facts, the act of reconstructing knowledge itself enhances
learning. Retrieval practice is not restricted to online learning. However, in an online
setting, all students are required to go through the same process [29]. Take for
example if the instructor asks a question during a lecture, one student might answer
the question while possibly half of the students were not aware that there was a
question posed in the first place. In MOOC, every student is forced to answer any
question posted during the video and engage with the material [30].
One-on-one tutoring: Peer tutoring has been proven to give the best learning result.
Using this technology, MOOC can simulate, to an extent, a tutor who is
customized and personalized for each student. Furthermore, if the course content
itself is satisfactory for some students, they could act as tutors for other students [31].
Global community: Having a massive number of students all over the world would
make a community around the MOOC course, so the students have the opportunity
to connect with each other by posting a question and providing an answer in an
impressively short response time [32].
4.2 Development Techniques
These are techniques that implement the design and ensure that the tool used achieves
the design goal. Example, the usage of customizable tools targets the personalized curriculum
technique in design. Moreover, inserting short quizzes within the course material targets
active retrieval practice. Below illustrate some additional development techniques [33]:
177
(ICoIT'17)
10thof April, 2017.
Automated grading: The automatic grading is interactive, hence, it provides instant

feedback to students and provides them the opportunity to take the quiz again. The
technology has been improving to allow automatic grading in a lot of fields like math,
science, financial models and in programming [34].
Peer grading: On the subject of disciplines that require critical thinking, automatic
grading does not work. Coursera has provided peer grading to solve this issue. Note
that peer grading does not only solve this problem whereas a massive number of
assignments in such disciplines as sociology and business. Importantly, peer grading
improves the learning process whereby students get the opportunity to correspond
with the knowledge and can explore the solutions from the different point of view
[35].
Data collection and analysis: The amount of data that be collected and the analysis
applied could arise the learning process to a different dimension. This can be done by
understanding the students needs and providing more interactive and personalized
responses to students, which ultimately improves the educational process [36].
4.3 Deployment Techniques
Deployment techniques are the delivery methods for MOOC platform or MOOC
courses .
Open-source platform: Systems like Moodle, Open MOOC7 and edX are open-
source- based platforms that provide MOOC courses. These systems offer the option
and flexibility of either being an independent MOOC platform setup or just hosting
a MOOC course in an existing platform. Some of the programing languages used in
these systems include PHP, Python, Ruby on Rails, and JavaScript [37].
Proprietary system: Blackboards Course Site, Desire2Learns Open Course, and
In structure's Canvas are examples of these systems. Some provide limited courses,
hosting, and functionalities free of charge, but in order to get the full package,
organizations are required to purchase the license [38].
Software as a Service (SaaS): "mooc.org" is a newly-built platform by Google and
edX to allow creating and hosting MOOC courses. At first, Google started its own
Course Builder platform which provided users with the capability to create MOOC
courses and deploy them on the Google app engine [39].
5. ARCHITECTURE/ TECHNICAL METHODS OF MOOCS
The main interface of the platform is designed by using the following techniques and as
illustrated in Figure (2):
1. HTML5 & CSS3: HTML language is used over all to design the page layout, while
the style of the web page is created using CSS language.
2. JavaScript: the animations and validation task is created by JavaScript language.
3. PHP: the back end codes that control the front end logic is applied by PHP language.
4. MYSQL: is used as a database for the created project and manipulation by the
MYSQL language.
178
(ICoIT'17)
10thof April, 2017.
5. Apache Server: Platform will be operated, using Apache server.

6. Windows 10: is the used system for the Platform.
Apache Window
Server s 10 HTML
5&
CSS3
MySQL
JavaScri PHP
pt
Figure (2): MOOCs Technical Platform
6. MOOC TOOLS/TECHNOLOGIES
There is many numbers of tools that could be used in MOOC, a lot of which we might
not be able to mention in this paper, however, we will point out the most widely used tools
in MOOC and highlight their importance to the MOOC platform. On the other hand, most of
these tools can fall into one or more of these categories [40]:
6.1 Content Management Tools
An example of these tools: video, audio, presentation and text .These are the tools that
deal with the main content of the course and they are the primary tools of MOOCs as they
are the course objects or the delivery tool of the objects that contain the initial knowledge
shared. The designer of the MOOC platform usually integrates more than one of these tools
to improve interaction. The content is uploaded or posted by the course instructor using one
of these tools, either via video, audio, presentation, or text document. In MOOCs, the
contents are rather short in time or small in size in order to enhance the students
understanding and gain of knowledge. Moreover, the contents usually include an assessment
tool to motivate the students understanding by applying a retrieval practice of the main idea
into the lecture. Althoughwe see the combination of the tools, content management, and
assessment tools, the content does not include all the knowledge that the students require.
Based on the original MOOC, or what is now called cMOOC, knowledge is created
and generated during the course as participants contribute to the course material through their
discussion about the topic. Although there is an area for discussion, contribution and
knowledge sharing in MOOC as we will see in the discussion tools is not linked to the
179
(ICoIT'17)
10thof April, 2017.
contents like the assessment tools, which makes discussions sometimes seem to be irrelevant
information rather than concrete knowledge [41].
6.2 Social Media Tools
An example of these tools: Facebook, Twitter, Google+, Linked in and My Space .
These are the tools that create the learning community and encourage collaboration among
participants to support the educational process in MOOC. Moreover, they open the learning
materials to the outside world and attract interest from others, as users begin to share their
MOOC activities and experiences within their own social networks [42].
6.3 Discussion and Collaborative Tools
An example of these tools: Discussion forums, Wikis, Blogs, Micro- blogs, and
Virtual study groups/meeting .
Discussion forums are the major and most common tool of communication used in any LMS.
Especially in traditional LMSs - it is the only tool that allows collaboration among
participants. For that reason, it has become a standard feature of LMSs. This feature as well
carried its way to MOOC platforms. The major purpose of discussion forums is to create a
learning community by encouraging students to share their knowledge and contribute with
their own discussion, either in a form of a question or comments similar to discussions in a
classroom. However, the tool was not intended at first to be used for educational purpose.
Students were not enticed to participate, as neither the system nor the designers promoted
such features [43].
The LMS can also include other discussion tools such as real-time discussions, Wikis, Blogs,
and Micro-blogs, all of which can facilitate engagement among participants and encourage
collaboration in a way that suits each individual. The new MOOC platforms do use some of
these tools to create the learning community, however, most of the time they are used in
isolation of each other and of the main course content [44].
6.4 File Management Tools
An example of these tools: Google Drive, Microsoft Office, and Drop Box .Shared
and collaborative files are the tools mentioned to in this category. Shared files simplify
collaboration between participants. Some of these tools permit single access to files, while
others which are better for collaboration permit multiple synchronous access. For example,
users can access the same file and work on the same document and are capable of seeing
their changes directly. The best example of these tools is Google Drive, which provides the
basic and most widely used collaboration files: documents, presentations, spreadsheets, and
drawings. It has the option to link other existing file types from a preset list. Due to the fact
that these tools are commonly used for collaborative deliverables, MOOC does not need to
use most of their functionalities. However, MOOC focuses on collaborative and shared
knowledge, which make Wikis and Blogs the perfect tools for that purpose [45].
6.5 Notification Tools
An example of these tools: Emails, RSS feed, and Posts. Notification tools are very
important in the MOOC platform as well as in any LMS. They could be used for group or
broadcast communication. On the other hand, not everyone has the desire or the time to
always check emails. Mainly with the many numbers of learners in the system, quantity of
180
(ICoIT'17)
10thof April, 2017.
emails could be massive. Hence, we need a smart tool that allows customization and
personalization of notifications. Having such a tool encourages collaboration and fast
response among participants by enabling notification with the most suitable tool that fits the
learners objectives. In addition, these tools can be included with the students daily used
accounts so it does not become an additional burden. One way to integrate personalized
notification is to partner with social media sites and enable one login feature. Using one login
via the different social media sites or email accounts, the MOOC course is connected to that
site and can drive notifications to the learners via the most convenient method to the learner
[46].
6.6 Assessment Tools
These are the tools that help instructors to create course assessments to record
students grades and progress. Most assessments need to be the automated-grading type. But
when it comes to open-ended questions, the assessment should use a different scheme or
style (i.e. peer assessment) [47].
6.7 Analytical Tools

These tools are used primarily in the background by administrators to offer important
statistics about courses, usage, and users activity. These statistics are used by designers and
decision-makers to improve the platform. For example, Coursera studies these statistics
carefully and makes modifications consequently to improve the user experience [48].
7. SITE DIAGRAM TOOL
Site diagram describes the whole structure of the site and how pages are connected with
each other.
The structure of the website page must be chosen before considering the visuals. The
architecture information is used to describe the planning of a website structure .There are
two types of navigation structures that will eventually translated to the navigation menus;
wide and deep.
Wide navigation system is used for small web sites; the main pages are all visible together.
Deep navigation, coordinate the pages into categories.
The deep navigation system is selected for the MOOC portal website, because it is the
optimal solution for a large websites, drop-down menus and secondary navigation menu, as
shown below in Figure (3).Tool that used to draw this diagram is lucidchart which
considered better solution for visual communication. Online flowcharts, diagrams, and ER
models [49].
181
(ICoIT'17)
10thof April, 2017.
Figure (3): Shows MOOC Site Diagram
8. WIREFRAME DIAGRAMS
A wireframe diagram shows the structure of web page using only outlines for each
widget and content type, as shown in Figure (4).The purpose of wireframe diagram is to
give basic idea about home page interface without any graphic design. Best toot that used
to draw wireframe diagram is Pencil [50].
Figure (4): Shows the MOOC Wireframe Diagram
182
(ICoIT'17)
10thof April, 2017.
9. CONCLUSION
In this paper a new developed design of MOOC system is offered. The newly developed
design is based on all recent techniques and tools that are needed to run online courses
building them from the ground up and right to provide students with certificates of
completion by using the most efficient languages that are known in nowadays; the PHP7
scripting language ,MySQL language, HTML language, CSS language and finally JavaScript
programming language. The new platform is constructed with advanced features to make
online education more effective than the traditional class room methods. It offers a safe and
an easy way for learners and instructors to exchange ideas, share content or information,
connect and collaborate-anytime, anywhere and other. The new web design is thought to
fulfill the most efficient features including instructor and student dashboards, online quiz and
certificate, progress tracking and support for multimedia classes, in addition to a mobile
friendly themes that are thought to save time and money. The new web page is designed and
implemented to fit the general curriculum provided by the Iraqi ministry of higher education
and is thought to spread usage of E-learning in Iraq and also establish the first E-Iraqi
University.
10. REFERENCES
1. Mosakhani, M., & Jamporazmey, M. (2010). Introduce critical success factors (CSFs) of
elearning for evaluating e-learning implementation success. Educational and Information
Technology (ICEIT), IEEE
2. Agboola, A.K. (2006). Assessing the awareness and perceptions of academic staff in using
e-learning tools for instructional delivery in a post-secondary institution: A case study. The
Public Sector Innovation Journal, 11(3), 1-6.
3. Ajadi, T.O., Salawu, I.O., & Adeoye, F.A. (2008). E-learning and distance education in
Nigeria. Turkish Online Journal of Educational Technology, 7(4), 1-7.
4. D Koller, A Ng, C Do, and Z Chen, "Retention and intention in Massive Open Online
Courses," EDUCAUSE Review, vol. 48, no. 3, 2013.
5. Anderson, D. Huttenlocher, J. Kleinberg, and J. Leskovec. (2015) Engaging with Massive
Online Courses, in WWW. ACM, 2014, pp. 687698. [5] C. G. Brinton and M. Chiang,
MOOC Performance Prediction via Clickstream Data and Social Learning Networks, in
INFOCOM. IEEE.
6. Caswell, T. et al. (2008). Open educational resources: enabling universal education.
International Review of Research in Open & Distance Learning, 9 .)1(
7. B. Fogg. .(2009) A behavior model for persuasive design. In Proceedings of the 4th
international Conference on Persuasive Technology, page 40. ACM.
8. Gilbert, S. Morton, and J. Rowley, "e-Learning: The student experience," British Journal of
Educational Technology, vol. 38, no. 4, pp. 560-573.
9. C. G. Brinton, M. Chiang, S. Jain, H. Lam, Z. Liu, and F. M. F. Wong. (2014) Learning
About Social Learning in MOOCs: From Statistical Analysis to Generative Model,
Transactions on Learning Technologies, vol. 7, pp. 346359.
183
(ICoIT'17)
10thof April, 2017.
10. McCallum, C. M., Thomas, S. and Libarkin, J., The AlphaMOOC: Building a Massive Open
Online Course one graduate student at a time, 2013, Issue 33, ISSN: 1887-1542.
11. Creelman, A., Ehlers, U., & Ossinnilsson, E. (2014). Perspectives on MOOC quality An
account of the EFQUEL MOOC Quality Project. INNOQUAL - International Journal for
Innovation and Quality in Learning , 2 (3), 78-87.
12. Creed-Dikeogu, G., & Clark, C. (2013). Are you MOOC-ing yet? A review for academic
libraries. Kansas Library Association College and University Libraries Section Proceedings,
3 (1), 9-13.
13. D. Russell, et al.(2013), "Will Massive Online Open Courses change the education?," in CHI
'13 Extended Abstracts on Human Factors in Computing Systems, Paris,France, , pp. 2395-
2398 .
14. J. Basque and B. Pudelko. (2010), Exploring the potential of blended learning to promote
retention and achievement in higher education professional study programs, 9th
International Conference on Information Technology based Higher Education and Training
ITHET, Cappadocia, pp. 383-390.
15. R. Gluga, T. Lever, and J. Kay.(2013), Foundations for Modelling University Curricula in
Terms of Multiple Learning Goal Set, IEEE Trans. Learning Technologies, vol. 6, no. 1,
pp. 2537.
16. Cooper, S. & Sahami, M.(2013 ), Reflections on Stanfords MOOCs: New possibilities in
online education create new challenges. Communications of the ACM56(2), 28-30, doi:
10.1145/2408776.2408787.
17. deWaard, I., Abajian, S., Gallagher, M. S., Hogue, R., Keskin, N., Koutropoulos, A., &
Rodriguez, O. C. (2011). Using mLearning and MOOCs to understand chaos, emergence,
and complexity in education. The International Review of Research in Open and Distance
Learning, 12(7), 94-115.
18. Gamage, D., Perera, I., & Fernando, S. (2014). Effectiveness of eLearning through MOOC:
lessons learnt in selecting a MOOC. The 22nd International Conference on Computers in
Education (ICCE 2014). Nara.Japan.
19. K. Veeramachaneni, F. Dernoncourt, C. Taylor, Z. Pardos, and U.-M. OReilly, Moocdb:
Developing data standards for mooc data science, in AIED 2013 Workshops Proceedings
Volume. Citeseer, 2013, p. 17
20. Martin, F. G. (2012). Will massive open online courses change how we teach?
Communications of the ACM, 55(8), 26-28. doi: 10.1145/2240236.2240246
21. Fini, A. (2009). The technological dimension of a massive open online course: The case of
the CCK08 course tools. The International Review Of Research In Open And Distance
Learning, 10 .)8(
22. J. Mackness and R. Williams.(2010), "The ideals and reality of participating in MOOC," in
Networked Learing Conference, London, pp. 266-275.
184
(ICoIT'17)
10thof April, 2017.
23. Kizilcec, R. F., Piech, C., & Schneider, E. (2013). Deconstructing disengagement: Analyzing
learner subpopulations in Massive Open Online Courses. Third International Conference on
Learning Analytics and Knowledge. Leuven, Belgium
24. Hew, K. F., & Cheung, W. S. (2014). Students and instructors use of massive open online
courses (MOOCs): Motivations and challenges. Educational Research Review, 12, 45-58.
doi: 10.1016/j.edurev.2014.05.001
25. Chew, Renee S.Y. (2011). Perceptions of online learning in an Australian university:
Malaysian students perspectives. Doctoral dissertation. Queensland University of
Technology .
26. I. Titov and R. McDonald.(2008), Modeling online reviews with multi-grain topic models.
In Proceedings of the 17th international conference on World Wide Web, pages 111120.
ACM,
27. Russell, D.M., Klemmer, S., Fox, A., Latulipe, C., Duneier, M., and Losh, E.(2013), Will
Massive Online Open Courses (MOOCs) Change Education? , in: CHI '13 Extended
Abstracts on Human Factors in Computing Systems. Paris, France: ACM, pp. 2395-2398.
28. Tomkin, J.H., Charlevoix, D.(2014): Do Professors Matter?: Using an A/B Test to Evaluate
the Impact of Instructor Involvement on MOOC Student Outcomes. In: Proceedings of the
First ACM Conference on Learning at Scale Conference. pp. 7178. L@S14, ACM .
29. Colvin, K. F., Champaign, J., Zhou, A., Fredericks, C., & Pritchard, D. E. (2014). Learning
in an Introductory Physics MOOC: All Cohorts Learn Equally, Including an On-Campus
Class. The International Review of Research in Open and Distributed Learning ( IRRODL)
, 15 (4).
30. Kop, R. (2011), The challenges to Connectivist learning on open online networks: Learning
experience during a massive open online course. International Review of Research in Open
and Distance learning, 12(3), 19-37.
31. M Bali.(2014), "MOOC Pedagogy: Gleaning Good Practice from Existing MOOCs," Journal
of Online Teaching and Learning (JOLT), vol. 10, no. 1, pp. 44-57.
32. K. Masters .(2011), "A brief guide to understanding MOOCs," The Internet Journal of
Medical Education, pp. 2-4,.
33. Manir, K. A. (2011). Implication of ICTs in libraries of higher education institutes: A
panacea catapulting library development in Africa. DESIDOC Journal of Library &
Information Technology, 31(1), 65-71.
34. T. R. Liyanagunawardena, A. A. Adams, and S. A. Williams, "MOOCs: A systematic study
of the published literature 2008-2012," The International Review of Research in Open and
Distance Learning, vol. 14, no. 3, pp. 202-227, 2013 .
35. Pant E., Comas-Quinn A. (2013), The Challenge of Open Education, Journal of e-Learning
and Knowledge Society, v.9, n.1, 11-22
36. Milligan, C., Littlejohn, A., and Margaryan, A.(2013), Patterns of engagement in
connectivist MOOCs, Journal of Online Learning and Teaching ,9(2), pp 149-159.
185
(ICoIT'17)
10thof April, 2017.
37. R. F. Kizilcec, C. Piech, and E. Schneider, Deconstructing disengagement: analyzing learner

subpopulations in massive open online courses, In Third Conference on Learning Analytics
and Knowledge, 2013, pages 170179.
38. Mak, S., Williams, R., and Mackness, J, Blogs and Forums as Communication and Learning
Tools in a MOOC, Networked Learning Conference, 2010, pp. 275-2.
39. S. Jiang, A. Williams, K. Schenke, M. Warschauer, amd D. ODowd, Predicting MOOC
performance with week 1 behavior, in Proc. the 7th International Conference Educational
Data Mining, London, pp. 273-275, July 2014
40. Mackness, J., Mak, S., and Williams, R. The ideals and reality of participating in a MOOC.
In International Conference on Networked Learning 2010. 266275. 23.
41. Skiba, D. J. (2012). Disruption in higher education: Massively open online courses
(MOOCs). Nursing education perspectives , 33 (6), 416-417.
42. Vihavainen, a., Luukkainen, & Kurhila, J. (2012 October). Multifacted support for MOOC
in Programming. SIGITE12 of ACM. 171-176. doi: 978-1-4503-1464-0.
43. Clow, D. MOOCs and the funnel of participation. In LAK 13: 3rd International Conference
on Learning Analytics & Knowledge (2013). 9.
44. Cormier, D., and Siemens, G., The Open Course: Through the Open Door-Open Courses as
Research, Learning, and Engagement, 2010, pp. 30-32.
45. S. Goldrick-Rab. Challenges and opportunities for improving community college student
success. Review of Educational Research, 80(3):437469, 2010.
46. Y. Wang and R. Baker. Mooc learner motivation and course completion rate. MOOC
Research Initiative Conference, 2013.
47. AF Yousef, M, A Chatti, U Schroeder, and M Wosnitza, "What Drives a Successful MOOC?
An Empirical Examination of Criteria to Assure Design Quality of MOOCs," in Advanced
Learning Technologies (ICALT) IEEE14th International Conference, 2014, pp. 44-48.
48. B. Kim and T. Reeves. Reframing research on learning with technology: in search of the
meaning of cognitive tools. Instructional Science, 35(3):207256, 2007.
49. Robbins, J.N., Learning Web Design. Fourth Edition ed. 2012, Inc., 1005 Gravenstein
Highway North, Sebastopol, CA 95472: OReilly Media. 603.
50. Paul Deitel, H.D., Abbey Deitel, Internet and World Wide Web: How to Program. 5th ed.
How To Program Series. Vol. 1. 2012: Prentice Hall. 923.
186
(ICoIT'17)
10thof April, 2017.
Design and Execute a Complete Vehicle System

for the MOHEDSR
Mead Ali Moter
Rafah Al-Hamdani
Israa Shehab
Wasan Ala'a alhamami

Databases touch all aspects of our lives. The database is
Article History: a group of records that is arranged in a way that makes it easy
Received: 20 March 2017 to be managed, accessed and updated. The relational database
Published: 10 April 2017 can be defines as a group of data items arranged as a set of
formally-described tables. Data can be accessed and
Keywords: VDBMS, reassembled in a number of different ways without having to
database management reorganize the database tables. The Ministry of higher
system (DBMS) , education and scientific research currently stores the vehicles
MOHEDSR, Vehicle records manually this makes it hard and difficult to manage
Record, SQL, Oracle and update these records. This project aims to develop a
database management system to store and retrieve the
vehicles records of the ministry of the higher education. The
projects system are developed using Oracle database
management software, the system provides an interface for
the data entry of a new vehicle record. Updating data of a
previously added record is also performed using interfaces.
SQL has been used in developing the system queries to
display reports and tables based on specified constrains.
1- INTRODUCTION
D atabases touch all aspects of our lives. A Database management system

is a computerized record-keeping system. The developed database
management system are designed to manage all the types of vehicles
records that belongs to the MOHEDSRs universities and institutions, and any new
purchased vehicle can also be added. The system is developed in order to reduce the
efforts of the vehicles records management.
A database is a collection of information that is arranged in a manner that makes
it easy to be accessed, managed, and updated. In one way, databases can be classified
according to types of content: bibliographic, full-text, numeric, and images. The aim
of a database system is to obtain a highly organized collection of data in addition to
appropriate tools and applications that aids processing and access to that data. A
database can be simplified as an organized collection of a data which is related in
some form, database are usually stored on disk, and may be accessed by a number of
concurrent users. Generally, databases can be separated into area of applications. For
187
(ICoIT'17)
10thof April, 2017.
example, a database could contain Human Resource (employee and payroll) data; or
it may contain accounting data; another may contain sales data; and so on. Databases
are managed by a database management system (DBMS).
There is confusion between a database system and a database management
system, but the two are different. Whereas a DBMS is a set of software tools that is
used for storing and manipulating data. Database management systems on the other
hand have many different types, such as file store, document store, and relational
database management systems.
Examples of different types of databases include end-user databases, operational
databases, analytical databases, distributed databases, relational databases,
hierarchical databases and database models. Databases can be categorized depending
on the type of their content, application area and technical aspect. For example, a
deductive database is the combination of logic programming and a relational
database, while graph structures are used in a graph database for representing and
storing information. Database types may also include hypertext databases, parallel
databases, mobile databases, cloud databases, active databases, in-memory databases,
temporal databases, spatial databases, real-time databases, probabilistic databases
and embedded databases
2- VEHICLES DATABASE MANAGEMENT SYSTEM
The Vehicles Database Management System (VDBMS) has been developed to
the benefit of the Ministry of Higher Education and Scientific Research
(MOHEDSR). The system is responsible for the management of all the vehicles
owned by the MOHEDSR and that have been delivered to all MOHEDSRs
universities and institutions. Many types of vehicles such as personal use cars, buses
and trucks are all stored and organized by their categories.
The VDBMS has the advantage of being able to manage data effectively and
allowing users to perform multiple tasks easily. The VDBMS uses a single software
application to stores, organizes and manages a large amount of information.
Efficiency of operations is provided in this system as well as the reduction of the
overall costs. The VDBMS is important to the Ministry of higher education and
scientific research because it provides a highly efficient method in handling multiple
types of data. Some of the data that are easily managed with this type of system
include: vehicles records, person in charge, MOHEDSRs institutions, status of the
vehicle and vehicles types. This system was built with the aim to be extremely
versatile.
Without VDBMS, tasks have to be done manually and take a lot of time and
possibly a lot of paper work. Data can be categorized and structured to accommodate
the needs of MOHEDSR. Data is entered into the system and accessed on a routine
basis by specified users. Each user may be given a password to gain access to a certain
part of the system. A number of users can access the system simultaneously in
different ways.
The system database is relational type, this means that multiple tables are used
and relationships between these tables are defined using a schema in addition to data
elements. Data elements and records from each table are merged, depending on the
query to be displayed in specific table or provided as a report. The obtained results
188
(ICoIT'17)
10thof April, 2017.
are displayed in a form that uses the defined data elements whereas only the records
that meet the query criteria are shown.
3- AIM OF THE RESEARCH
This project aims to develop a database to store, organize and manage all the
related to the vehicles owned by the Ministry of higher education and scientific
research. The system is developed to make it easy to retrieve specific information or
build custom reports about any vehicle. The stored data that contain information
about the current status of any vehicle record can be update so the information will
always be up-to-date. Detailed information about every vehicle is stored such as
brand, Model, plate number, type and color. Figure 3.1 shows the categories of
VDBMS interfaces functions.
Figure 3.1: VDBMS Main Functions
3.2 VDBMS DATABASE COMPONENTS

The VDBMS database were build based on the collected requirement, it contains
all the information related to the vehicles. The same information currently being
written in the documents of the vehicles records has been transformed to the created
database tables. The main table of the database is the VEHICLE table which contains
all the related information about the vehicle and code of the other related data.
Vehicle color, status, type and main type are also created in separated tables with a
unique code for each. CODE_RASH table save the data about all the MOHEDSRs
institutions and universities.
VEH_COLOR table contains the color name with a code number for each color,
this code is the primary key of the table and used in the relation with main Vehicle
table as a foreign key. Each vehicle has a current status like new, used, out of service
and broken VEH_STATUS table stores the code of the status (primary key of the
table) and the name of the status, the NO attributes acts as foreign key in the
VEHICLE table. VEH_TYPE table has three attributes, CODE the primary key of
189
(ICoIT'17)
10thof April, 2017.
the table, VEH_TYPE the name of the vehicle type same name registered by the
vehicle manufacture, P_CODE the category code of the vehicle. CODEN_RASH
table has the UCODE attribute as the primary key and other attributes that is used to
store related data to the Institute or university currently using the vehicle such as
division and department names. Figure 3.2 shows the Entity diagram of the created
tables and the attributes of each table.
VEH_STATUS Vehicle CODEN RASH

NO VEH_UNIT UCODE
NAME VEH_STATUS MINST CODE
VEH_M_NO SUB MINST
VEH_DOC_NO DEPART
VEH_TYPE DIVISION
VEH_MODULE OFFICE1
VEH_P_TYPE VEH_TEMPLE_NO OFFICE2
CODE VEH_COLOR MLK CODE
NAME VEH_P_TYPE CODE NAME
PERSON_NO DIVISION1
V_BOOK_NO
V_BOOK_DATE
CITY VEH_COLOR
CODE
VEH_TYPE CODE
V_TRAN_BOOK_N
VEH_COLOR
CODE V_TRAN_BOOK_D
VEH_TYPE ATE
P CODE PERSON_NAME
RANK
Figure 3.2: VDBMS Entity Diagram
The vehicle table contains one primary key which is the CODE attribute and five
foreign keys each one of them is the primary key of the tables mentioned above.
These five foreign keys are VEH_TYPE, VEH_P_TYPE, VEH_COLOR and
VEH_STATUS. Other attributes store the rest of the information related to the
Vehicle. Table 3.1 shows detailed information about the attributes of the VDBMS
Main table (VEHICLE).
190
(ICoIT'17)
10thof April, 2017.
Table 3.1: VDBMS Main Table
4-VDBMS INTERFACES
The system Interfaces were developed to cover all the database management
function that the system perform. Each interface has predefined purpose, these
purposes varies from adding new records, updating records to building reports. This
section will introduce the functionality of each interface.
4.1 Main Interface
Through navigation to all the system interfaces can be done. Figure 4.1 shows
the main interface. This interface has eight command buttons each one of them
represents a link to another interface, and one command button for exiting out of the
system.
191
(ICoIT'17)
10thof April, 2017.
Figure 4.1: VDBMS Main Interface
4.2 Adding New Vehicle Record Interface

Adding new vehicle records could be done using this interface. All the
information and attributes related to the vehicle record are shown as blank text box
do be filled with new vehicle information. Figure 3.4 shows the add new vehicle
interface.
Figure 4.2: VDBMS Add New Vehicle

Interface
192
(ICoIT'17)
10thof April, 2017.
4.3 Update Vehicle Record Interface
This interface was designed based on the SQL command Update. The main
function of this interface is to update a previously added vehicle records same
attributes that has been added in the add new interface can be seen in this interface
in order to enable the user of making the required updates and changes to these
attributes. Figure 3.5 shows the Update interface.
Figure 4.3: VDBMS Update Vehicle

Interface
4.4 Delete Vehicle Record Interface
Vehicle records can be deleted by selecting the delete vehicle record command
button from the main interface. Any vehicle record could be deleted using the delete
interface by the vehicle plate number. The records deleted from this interface will be
permanently removed from the database. Figure 3.6 shows the delete interface.
193
(ICoIT'17)
10thof April, 2017.
Figure 4.4: VDBMS Display Interface
4.5 Records Display Interface

Clicking on the display records command buttons will navigate to another
interface that contains multiple options on how the user would like to view and
retrieve the vehicles records. The display interfaces with the displaying options are
shown in Figure 3.7. Figure 3.8 shows the display vehicle record by vehicle plate
number.
There are nine options of records displaying. All these option could be found at
the display interface. These nine options are:
Display by vehicle plate number.
Display by the name of the employee.
Display by chassis number.
Display by year of manufacturing.
Display by current university/institute owning the vehicle.
Display by color.
Display based on university/institute.
Display by vehicle status.
Display by vehicle category.
Display by vehicle type.
194
(ICoIT'17)
10thof April, 2017.
Figure 4.5: VDBMS Display Interface
4.6 Coding Interfaces

The VDBMS was designed to accept adding new vehicle type, new vehicle color
and adding new institution or university that has not been added previously to the
system.
Adding this type of data could be done through on of the specified interfaces.
Adding new vehicle type could be done by the add new vehicle type Interface
(Figure 3.9) Adding new vehicle color could be done by the add new vehicle color
Interface (Figure 3.10) Adding new institution or university could be done by the
add new institution or university Interface (Figure 3.11).
Figure 4.6: VDBMS Add New Vehicle Type Interface
195
(ICoIT'17)
10thof April, 2017.
Figure 3.10: VDBMS Add New Vehicle Color Interface
5. CONCLUSION
This project introduced a database management system that stores organize and
retrieve records of all the vehicles owned by the Iraqi Ministry of Higher Education
and Scientific Research (MOHEDSR). Currently vehicles records are documented
manually by the employee. This method takes a lot of time when a new vehicle is
purchased out has been transferred between the ministry institutions. In addition
probability of error is high. Updating status or keeping track of vehicle take a lot of
time.
Using the proposed database management system reduces the efforts paid by the
employees in charge of managing these vehicles and saves time while retrieving
reports of the vehicle.in addition a custom built report can be obtained from the
system.
The system provides the following functions:
Storing detailed information about the vehicle. These information
include:
o Vehicle ID (primary key).
o Vehicle plate number.
o Vehicle status (foreign key).
o Document number.
o Vehicle category (foreign key).
o Vehicle year of manufacturing.
o Vehicle type (foreign key).
o Chassis number.
o Vehicle color (foreign key).
o Vehicle owner number.
o Vehicle official letter Number.
o Vehicle official letter date of the.
o City.
196
(ICoIT'17)
10thof April, 2017.
o Number of the Transfer Official letter.

o Date of the Transfer Official letter.
o Vehicle owner name.
o Vehicle owner rank.
Updating the information of the vehicle record attributes.
Building reports of the vehicle records. these report are:
o Report based on university or institute.
o Report by vehicle plate number.
o Report by vehicle type and status.
o Report by vehicle category.
Using Oracle 10.0g database management software and the SQL commands were used to
build the VDBMS. The system interfaces were designed to be easy to learn, use and
administrate and effectively retrieve stored vehicles records
6.REFERENCES
[1] T. Connolly And C. Begg, Database Systems A Practical Approach To Design,
Implementation,And Management, 6th Ed. Pearson, 2015.
[2] M. Mannino, Database design, application development, and administration, 3rd
ed. Boston: McGraw-Hill Irwin, 2007.
[3] A. Kriegel, Discovering SQL A Hands-On Guide for Beginners. Indianapolis,
Indiana: Wiley Publishing, Inc., 2012.
[4] T. Teorey, S. Lightstone and T. Nadeau, Database modeling & design, 4th
ed. Amsterdam: Elsevier, 2006.
[5] K. Kendall and J. Kendall, Systems analysis and design, 8th ed. Upper Saddle
River, N.J.: Prentice Hall, 2010.
[6] S. Bennett, S. Mcrobb And R. Farmer, Object-Oriented Systems Analysis And
Design Using Uml. London: Mcgraw-Hill, 2002.
[7] A. Dennis, B. Wixom And R. Roth, System Analysis And Design, 5th Ed.
Wiely, 2012.
[8] J. Price, Oracle database 11g SQL. New York: Oracle Press/McGraw-Hill,
2008.
[9] C. Zeis, C. Ruel and M. Wessler, Oracle 11g for dummies. Hoboken, N.J.:
Wiley, 2009.
197
(ICoIT'17)
10thof April, 2017.
Design and Implementation of a Rehabilitation Wireless

Networked System
Assist. Prof. Dr. Mahmood Zaki Abdullah
Computer Engineering Department, College of Engineering, Al-Mustansiriyah University, Baghdad
- Iraq.
drmzaali@gmail.com
Assistant Lec. Fadia Noori Hummadi
Biomedical Engineering. Department, Al-Khwarizmy College of Engineering, University of
Baghdad, Baghdad Iraq.
fadia_noori@yahoo.com

Many countries in the world do great efforts to achieve
Article History: modern medical services to a lot of people; they applied
Received: 12 March 2017 modern network systems, electronic applications and
programs to facilitate these issues. This paper will give a
modern approach and algorithmic procedure for a
Published: 10 April 2017 rehabilitation system by using the modern aspects, techniques
Keywords: and protocols of the wireless networks. Firstly, an overview
Rehabilitation, Wireless about previous works will be introduced then a proposal
Network, Servers, modern design for this system will be given, test and compare
Distributed System, LAN, the obtained results with previous ones. A multi stage
Packet Tracer, Zigbee. distributed servers will be used via this new design, this will
assist and facilitate the ability of storage and delivering of the
monitoring medical data, also checking for errors or damaged
frames of data will be occurred at any stage.
1. INTRODUCTION
T
he Rehabilitation system is a modern medical monitoring system that can be
monitored and checked the case of the patients after doing the operation, many
of soft computing companies applied modern techniques and issued the software
that have contributed to assist the doctors to follow their patients remotely and check
their cases.
2. BACKGROUND
The basic architecture of the rehabilitation system can be consisted of medical sensors,
personal servers, clinical database, and the main server of the hospital; the medical sensors
can be connected with the personal servers by using the IEEE 820.15.4/Zigbee standard [1],
while the personal servers can be connected with the main hospital server through the 3G
198
(ICoIT'17)
10thof April, 2017.
networks. When the patients data are arrive to the main hospital server, the data are either
stored in the clinical data base or viewed from the doctors or the clinician via the hospital
LAN, and then the doctors can analyze these data and give the suitable diagnosis advices
accordingly [1]. Figure 1 shows the system architecture of the conventional rehabilitation
system.
FIGURE 1. The architecture of conventional Rehabilitation system [1]
3. RELATED WORKS
This section surveys previous works in structured of Rehabilitation systems as shown
below:
J. Bouchard, B. Prosperi, G. Bavre, M. Daud, and E. Jeandupeux [2], presented a proposed
software that uses a 3D model for representing patient and tool of medicine professionals.
Zhizheng Wu, Feng Li, Wenjun Mi and Jingtao Lei [3], proposed a new approach for Lower
Extremity Rehabilitation Robot (LERR), this approach presented a good design for a robust
controller for Rehabilitation Robot.
Stucki G, Bickenbach J, Melvin J. [4], presented a good discussion for securing access to
functioning data of cases and treatments for Rehabilitation system.
Dirk Nolte, Claudia Tschammler, Markus Henzler, Robert Linsenmann, and Johannes
Angermair [5], proposed a two-phase transplantation concept that represents a new surgical
ways that ultimately arose from the absence of therapeutic alternatives.
Ovidiu Filip, and Tudor Deaconescu [6], presented a good study and explanations about the
Rehabilitation equipment.
The motivation of the proposed work focuses on suggesting a good idea for storing the
database of patients cases and their treatments based on distributed servers concepts, as well
as presents a good engine for auto treatments.
The comparison between the related mentioned works and the proposed algorithm is the
proposed algorithm can give a new ways for managing and administrating the medical issues,
it can present a good way for storing a database with many attributes and entities for patients,
cases, diseases, medical treatments, diagnoses, and information about doctors who work
inside the hospital, as well as gives an auto engine researching algorithm for finding a
suitable doctor for a certain patients case, and monitors the patients remotely.
199
(ICoIT'17)
10thof April, 2017.
THE PROPOSED ARCHITECTURE

The proposed architecture of the Rehabilitation system consists of medical wireless
sensors, as well as the distributed servers that can be used as database storages. The patients
data can be encapsulated according to a certain format including the source and destination
IP addresses for the patients and doctor respectively. The central router delivers the incoming
data to their corresponding destinations. The doctor analyzes these data and gives a suitable
advice for the patient. Figure 2 shows the design of the proposed system.
FIGURE2. Design of the proposed architecture of the Rehabilitation system

To test and check the proposed architecture, a prototype design can be applied. This
prototype design consists of three patients with different medical cases, the data of these
patients are encapsulated to a certain protocol format that contains the source and destination
IP addresses, these packets transmitted wirelessly to a central router locates in the hospitals
position. This router will deliver these packets according to their corresponding destinations.
The doctors will analyze and study the receiving data and give the suitable diagnosis and
advices. The distributer servers can be used for storing the copies of all the transferring data
including their cases and advices. The proposed system suggests an auto treatment engine,
this engine can transmit a suitable advice for an urgent patients case, and can alert for the
dangerous cases. This proposed system can monitor the patients after their medical operation
and leaving the hospital, so it monitors the patients remotely by using the aspects of the
wireless networked systems, like 3G and 4G mobile techniques. The prototype design can
be applied by using the Packet Tracer ver. 3.0 simulator and gave the results as shown in
200
(ICoIT'17)
10thof April, 2017.
Figure 3, through Figure 8. Figure 3 shows that the sensor that connected to patient 1 is
transmitted its data to a smart mobile device of doctor 1.
FIGURE 3. Sensor of patient 1 ready to transmit its packet of data to smart mobile
device of doctor 1
Figure 4 shows the outbound data format of patient 1 that is transferred to doctor 1
FIGURE 4. Outbound data packet format of patient 1
201
(ICoIT'17)
10thof April, 2017.
Figure 5 shows that the data packet of patient 1 are arrived in the central router which is
located in a certain position at the hospital
FIGURE 5. Data packet of patient 1 which arrived at the central router

Figure 6 shows the outbound data format for patient 1 when it is arrived at the central
router in the hospital; this format includes the addresses of the patient 1 and the doctor 1
respectively.
FIGURE 6. Outbound data packet format of patient 1 which arrived at central router
202
(ICoIT'17)
10thof April, 2017.
Figure 7 shows the data packet format of patient 1 which arrived at the smart mobile
device of doctor 1.
FIGURE 7. Data packet of patient 1 which arrived at the smart mobile device of doctor1
Figure 8 shows the outbound data format for patient1 when it is arrived at the smart mobile
device of doctor1.
FIGURE 8. Outbound packet format of patient 1 which arrived at the smart mobile
device of doctor 1
203
(ICoIT'17)
10thof April, 2017.
5. CONCLUSION
The Rehabilitation system can offer many medical services remotely, it is easy for the
doctors to check and monitor cases of their patients remotely. The proposed architecture gave
good results for that, this system can give additional features like storing data for different
patients cases with their treatments; it also offers a good engine for auto medical treatment
of urgent cases if there are no responses from corresponding doctors, as well as give auto
monitoring for patients cases. This proposed system can be applied within a certain zone of
operation, to exceed this zone for a wide range; the other modern techniques of wireless
network system must be adopted like 3G, or 4G networks.
REFERENCES
[1] Rajasekaran.S., Kumaran.P., Premnath.G., and Karthik.M. (2013). "Human Health
Monitoring Using Wireless Sensors Networks (WSN)". International Journal of
Application or Innovation in Engineering & Management, Vol.2, Issue 12, pp. 323-330
[2] J. Bouchard, B. Prosperi, G. Bavre, M. Daud, and E. Jeandupeux (2013). "A Predictive
Rehabilitation Software for Cerebral Palsy Patients". International Journal of Medical,
Health, Biomedical, Bioengineering and Pharmaceutical Engineering, Vol. 7, No. 8, pp.
456-459.
[3] Zhizheng Wu, Feng Li, Wenjun Mi and Jingtao Lei (2014), " Robust Controller Design
for the Cooperative Control of Lower Extremity Rehabilitation Robot". International
Journal of Control and Automation, Vol.7, No.4 , pp. 313-328.
[4] Stucki G, Bickenbach J, Melvin J. (2016), Strengthening Rehabilitation in Health
Systems Worldwide by Integrating Information on Functioning in National Health
Information Systems, US National Library for Medicine, National Institute for Health.
[5] Dirk Nolte, Claudia Tschammler, Markus Henzler, Robert Linsenmann, and Johannes
Angermair (2017), Two-Phase Transplantation (TPTX) Concept: A New Approach for
Instant Rehabilitation of Young Children after Avulsion of Central Incisor, Open
Journal of Stomatology, Scientific Research Publishing, 136-146.
[6] Ovidiu Filip, and Tudor Deaconescu (2017), Finite Element Analysis of Some
Pneumatically Actuated Wrist Rehabilitation Equipment, MATEC Web of Conferences
94, published by EDP Sciences.
204
(ICoIT'17)
10thof April, 2017.
Distance Varieties Wireless Signals Relations Connected

Peer to Peer Computers
Kamaran HamaAliFaraj
Department of Computer, College of Science, Cihan university Sulamani,KRI
Department of Computer, College of Science, Sulamani university Sulamani,KRI
kamaranfaraj@yahoo.com
Banar Fareed Ibrahim
Department of IT, College of Science, Lebanese French University, Erbil, IRAQ-KGI
bnar.fareed@lfu.edu.krd
Azhi A. Salih
Department of Mathematics ,college of Basic Education ,Salahaddin university ,Erbil,IRAQ-KGI
azhi.salih@su.edu.krd

Static website by hypertext markup language (HTML)
Article History: published over a computer networks with the wireless
Received: 14 March 2017 distances variety are connected by peer-to-peer as a
client/server(C/S) based network. The evaluating test started
from distanced of (twenty meters, forty meters, sixty meters
Published: 10 April 2017 and up to hundred meters) between both points with the
communication media guide-wire and unguided-wireless. In
Keywords: No less than phase-1 (guide-wire evaluation) testing for three different
five keywords separated distances between both C/S computers which are (20 meters,
by comma. Times New 40 meters, 60 meters). Each mentioned distances which tested
Roman Font, Point Size and record different result. The phase-2 (unguided-wire
112, Not Justified, In evaluation) wirelessly for three different distances between
Italic. both C/S computers which are (20 meters, 40 meters, 60
meters). The network Quality of Service (QoS) in private
Network (PRN) or peer to peer (P2P) network is in different
distances discovered. The private wired and wireless
networks utilized in our proposed system and the
performance latency of website will be determinate for the
reason of solving traffic/bottleneck problem. The designed
constant size of static website-html and installed on Server-
side, also connected in different length of (cable UTP or
wireless) to several clients in order to test and discover lowest
response time. QoS will be finding out in least response time
in Network different distance wire and wirelessly with a
proper solution in right distance.
205
(ICoIT'17)
10thof April, 2017.
1. INTRODUCTION
T
he consideration WLANS allow users in a local area, such as a university
campus or library, to form a network or gain access to the internet. Also
Performance or Latency by response time is an important factor and hot topic in
the world of computer communication and networking [1]. The ways people and
devices connect to the Internet have changed significantly in recent years. Numerous
highly capable wireless networking technologies have been developed and widely
deployed. Todays commercial wireless communications landscape is operating a
revolution in the way people access and share information [2]. In this paper concentrate
variety distance is up to hundred meters to connect computer1 and computer2 wired and
wirelessly methods. The recent society we live in relies heavily on communication [3]
is a wireless more than before. A computer network consists of a number of computers
joined together for the purpose of communicating and sharing resources [4]. Wireless
networking takes into consideration the range, mobility, and the several types of
hardware components needed to establish a wireless network [5], the range of distance
between two points wirelessly and wire connected is under investigation by performance
(time response). The least response time with longer distance is a target for both methods
of communication. Thus, the result test of distance time-consuming between two
connected computers with least response will be an appropriate solution for our paper.
A temporary network can be formed by a small number of users without the need of an
access point [6], and protocols needed for optimum performance. In the proposed
configuration of wireless networks with peer to peer LANs-private network and with
wired and wireless methods. Proposed wired private and wireless private network as
multipoint is shown in Figure (1).
Figure (1) proposed private-wired and private-wireless Networks as multipoint
Since our main objective is to advise on the LAN requirements of a prototype real-life
business and fine best result in the respect of least response time with longer distance. We
aim at producing an informed report on the LAN needs of any business. In order to do this,
the network designer must understand how computers communicate in two different phases
namely one and two. The two phases were mentioned in the previews section. The figure (2)
Show the phases unguided and guided communication in only two points, and it is suitable
for our evaluation. The paper starts with the background used for this research. Then, a brief
206
(ICoIT'17)
10thof April, 2017.
description of the system performance as well as the trial that took place, subsequently a
discussion of the survey results is present. Finally, paper finished with a summary and
conclusions.
Figure (2) phases unguided and guided proposed communication in only two points
2. BACKGROUND
Network or Networking of computers has an important role in nowadays society; from past
until now there are lots of modifications occurred in Computer networks and create new
computer networks generation. Each generation has improved the computer networks
performance [5]. Thus, the performance and computer generation have a direct relation
between them. Increasing of performance is a hot top in networking science and technology.
each modification in performance network is creating new generation; for example Two
decades ago, few people had access to a network. Now, computer communication has
become an essential part of our infrastructure. Networking computer are everywhere and
used in Government, Education, Commerce, military [6] and etc. The popularity of computer
network is returned to several aspects; time, space, distance, better quality of service (QoS),
Speed, power, real time activity [6] and etc.
The unguided networked houses or offices the make smarter house or smarter office, but its
not grantee for the better security than the guided network or designs for maximizing QoS in
complex networks must end with several important criteria [7]. An operational perspective
is the ability of the network to service an application effectively, without affecting its
performance and functionality [8]. The most important criteria are performance, reliability
and security. The performance criteria are mostly considerate for evaluation the speed of
wired and wireless network in our proposed system, in this paper the performance is a
selected criteria for testing and evaluation for guided and unguided point to point network in
different distance. The distance is variable:
2.1. PERFORMANCE
Investigation for wireless and wire point to point Performance is in many techniques with
transmit time and response time. Transmit time is the amount of time required for a message
(i.e. open index or home page) that travel from one device to another. The response time is
the elapsed time between an inquiry and a response. The response time is the elapsed time
between end of enquiry and the beginning of the response [9] (i.e. the response time in our
proposed system is the inquiry start the ping to server domain name by client until beginning
207
(ICoIT'17)
10thof April, 2017.
of response. its end of ping the domain name and beginning of response.). The figure (3)
Show the performance by response time by elapsed time between end enquiry and the
beginning of the response.
Figure (4) requesting time by elapsed time between end requesting files
Transfer and starting files transfer
The performance of a network depends on a number of factors, including the number of
users, the type of transmission medium [9] either wire or wireless. The capabilities of the
connected hardware and the efficiency of the software [9] are very important factor to
increase the performance. The better hardware capability and software efficiency improve
the response time. But in this paper are only two (connected guided and unguided) stations,
and used two different IP. Thus, the investigation is for the limited performance with variable
distance.
3. MEASUREMENT TECHNIQUES AND TOOLS
The reliability of wireless communication deteriorates due to many factors such as

interference, path loss and blockage. One of the most likely sources of wireless
communication errors is fading [10]. In order to find out the fading of wireless and wire
communication signals with different distance will be an investigation of the paper. For that
reason the performance measurements could be done by a tool is commonly used for example
Ping is very commonly used tool for network. Availability and round-trip delay are measured
by Ping and used for our process system in order to test all performance aspect. Ping is the
name of a software utility originally written by Mike Muuss [11]. In wireless and wire
communication; the time needs it takes a "packet" of data to travel from client computer to a
server on the Internet and return is calculated by Ping measurement. The fading signal is
occurred by delayed responses in Internet applications - this is possible due to a higher than
preferred ping (e.g. packet loss is lesser will be better when it comes to ping). The best to use
is a very good suggestion to use echo request and echo reply messages to check if the remote
host is reachable or not, that is, there is a network connection from the local host to the remote
host. The ping command is a very easy way to confirm that your computer connection is
active and run a ping test to ensure connectivity. In order to receive a four-line return and a
synopsis explaining how much data was transferred from the remote domains and how
quickly. There is some important information about ping. This information is very important
for verify time in millisecond:
The figure (5) shows all steps from start button to DOS prompt.
208
(ICoIT'17)
10thof April, 2017.
1) Select to Start. 2) Click on Run. 3) Type COMMAND and click OK. 4) Once the DOS
prompt comes up type the following:
The Figure (5) shows all steps from start button to DOS prompt.
ping cihan.krd. [enter]
The black screen of MS DOS-prompt appear with all information about the connected client
to server and vice versa.
All Steps Complete. And all information are shown as below about ping in table (1) [10].
In order to find network latency statistics [8], the activity need to be performed on any
connected and worked network for determination measure and evaluate network latency over
time, and during different periods of the day to capture a representative sample of typical
network activity. This will be able to by analysing the return delay from a distant computer
with the ping command. Return delay times, measured in milliseconds, will be summarized
by computing the average latency (mean) and the range (maximum and minimum) of the
delay times.
The ping is a useful for any network activity or finds out the best performance, also can be
used to calculate the round-trip delay to the remote host. Ping software is working as open
source, which has resulted in the software that supports almost every OS [12]. The index or
home page designed as a static and Ping is the measurement of time taken for home or index
page to travel from one point to another over a network.
For example, Speed of information traveling from client computer to a server or appearing
the index page by client computer, thus traveling from any client to a server (down load a
website by any client). A 'ping' estimates a round trip time using interval timing responses
and will generally give a result that's measured in 'milliseconds'.
209
(ICoIT'17)
10thof April, 2017.
Table (1) describes all ping information and applications that used for designing proposed
system[11].
4. TESTS AND EVALUATION

The Measurement techniques and tools is Ping for evaluation of our proposed system. The
measurement of time taken for information to travel from one point to another over a network
is achieved to analyse best result between guided and unguided in the respect of distance.
4.1. TWENTY METERS DISTANCE:
The results of ping test shows in Figure (6a) by command prompt for wired communication
between server and client. Ping between two computers in an (peer to peer wired LAN), one
of them is server with the IP first computer is a server 192.168.1.1, and the other are clients
with IP second computer is a client 192.168.1.2.
However, in Figure (6b) by command prompt for wireless communication between server
and client. Ping between two computers in an (peer to peer wireless LAN), one of them is
server with the IP first computer is a server 192.168.1.1, and the other are clients with IP
second computer is a client 192.168.1.2
Figure (6a) test result of wire peer to peer LAN in ping by prompt command
210
(ICoIT'17)
10thof April, 2017.
Figure (6b) test result of wireless peer to peer LAN in ping by prompt command
There must be different between two results ping test show witch one of them is better
private network (wire or wireless) in the distance of twenty meters. The result of private
network-wired is better than the private network-wireless with constant distance of twenty
meters between client and server. In general, the response time of the private network-wire
is lesser than the private-wireless network (LAN). The private network-wireless (LAN) is
very limited up to sixty five meters, but the private network-wire (LAN) distance is up to
hundred meters. The results show that private network-wire (LAN) is less than the private
network-wireless (LAN) in the fixed distance of twenty meters. All information are shown
as below about ping reply milliseconds (ms) for private wire and wireless network LAN in
table (2).
Table (2) describes all ping reply for private Wire and private Wireless network.
The table evaluation tests show a different between private Network-Wire LWAN
(PNWrL) and Private Network-Wireless LAN (PNWlessL). In PNWrL the suitable distance
limitation between client and server is hundred meters, but the distance limitation between
client and server approximately sixty five meters. The time response of PNWrL is much
better than PNWlessL because the signals easily reach more than hundred meters, but
PNWlessL signal difficultly reach to sixty five meters. This mean that after sixty meters the
signal will be weak and cause a communication problem but in wired after hundred meters
the signal is weak. Figure (7) show the ping result for PNWrL with distance twenty meters,
PNWlessL with distance twenty meters.
211
(ICoIT'17)
10thof April, 2017.
2 wire
0 wireless wireless
reply one wire

reply two
reply three
reply four
Figure (7) show the ping result for PNWrL with distance twenty meters PNWlessL
with distance twenty meters
All information are shown as below about ping reply (ms) for private wire and wireless
network LAN in table (3).
Table (3) show the PNWrL and PNWlessL results which are same in send, received and
lost for twenty meters distance.
Finally table (4) show the approximate round trip in milliseconds (ms) for PNWrL and
PNWlessL
Table (4) show the approximate round trip in (ms) for PNWrL
Round Trip approximation Round Trip approximation

Wired Wireless
Minimum 0 1
Maximum 2 4
Average 0 2
Figure (8) show the round trip minimum to maximum and average for PNWrL and
PNWlessL
212
(ICoIT'17)
10thof April, 2017.
5
4
3
2
1
0
0 0.5 1 1.5 2 2.5 3 3.5
Figure (8) shows the round trip minimum to maximum and average for PNWrL
and PNWlessL
All tests with twenty meters of distance between two points show the wired
communication medium is better and less than the wireless medium. The round trip for
wireless is not as good as wire round trip because of the table (4) declare that the limitation
approximation round trip for wireless is from (1-4)-brown upper line in Fig(8), but the
limitation approximation round trip for wire is from (0-2)-green bottom line in fig(8). It is an
obvious the (0-2) limitation is acceptable than (1-4) limitation.
4.2 FORTY METERS DISTANCE: As mentioned before in section (4.1) the distance
between two points is twenty meters. The same test will be repeated for forty meters distance
and sixty meter distance. Ping between two computers in an (peer to peer wired LAN), one
of them is server with the IP first computer is a server 192.168.1.1, and the other are clients
with IP second computer is a client 192.168, Figure(9a) test result of wire peer to peer LAN
in ping by prompt command.
Figure (9a) test result of wire peer to peer PNWrL
However, in Figure (9b) by command prompt for wireless communication between server
and client.
213
(ICoIT'17)
10thof April, 2017.
Figure (9b) test result of wireless peer to peer PNWlessL

All information are shown as below about ping reply milliseconds (ms) for private wire and
wireless network LAN in table (5).
Table (5) describes all ping reply for PNWlessL and PNWrL network.
Figure (10) show the ping result for PNWrL with distance forty meters, PNWlessL with
distance forty meters.
Figure (10) show the ping result for PNWrL with distance forty meters PNWlessL with
distance forty meters
The figure (11) shows the highest and lowest delay time for round trip of signals between
clients and server. The number in milliseconds delay time is very high if compared the
PNWlessL to PNWrL. round trip approximation is shows the better in response and where
to reactivate signals in order to avoid signal weakness. The weakness signals take it to
account as delay time and round trip maximum approximation.
214
(ICoIT'17)
10thof April, 2017.
Round Trip approximation Round Trip approximation
Wire Wireless
Minimum 0 3
Maximum 1 13
Average 0 7
When the round trip is higher than usual is cause weak signal and or delay. Hence, reactivate
the signals by placing repeater hardware. In ping statistics for forty meters distance there are
no lost packet in send and received and showed table (3) in section 4.1. Table (6) show the
approximate round trip in milliseconds (ms) for PNWrL and PNWlessL for forty meters.
Figure (11) show the round trip minimum to maximum and average for PNWrL
Figure (11) show the round trip minimum to minimum and average for PNWrL
and PNWless
4.3 SIXTY METERS DISTANCE:
As mentioned before in sections (4.1) and (4.2) for distance the twenty and forty meters test.
The same test will be repeated for sixty meters distance for both PNWrL and PNWlessL with
the IP first computer a server is 192.168.1.1, and the other are clients with IP second
computer is a client 192.168. Figure (12a & 12b) tests result of wire peer to peer LAN in
ping by prompt command
215
(ICoIT'17)
10thof April, 2017.
Figure (12a) test result of wireless peer to peer PNWrL for sixty meters distance
Fig (12b) test result of wireless peer to peer PNWlessL for sixty meters distance
All information are shown as below about ping reply milliseconds (ms) for private wire
and wireless network LAN in table (7). The results are very different.
Table (7) show the PNWrL and PNWlessL results which are different in send, received
and lost for sixty meters distance.
Figure (13) show the ping result for PNWrL with distance forty meters, PNWlessL with
distance forty meters. .
216
(ICoIT'17)
10thof April, 2017.
200
100 wire 1
wireless no wireless no
0
1 wire 1
2
3
Figure (13) show the ping result for PNWrL with distance sixty meters PNWlessL with
distance sixty meters
All results in wire media are reachable Distention but in wireless first reply is unreachable
distention. The approximate results between them are very clear and show that the PNWrL
much faster due to the distance. Table (8) show the approximate round trip in milliseconds
(ms) for PNWrL and PNWlessL up to sixty meter
The results of test for round trip from minimum to maximum are affected by the distance
and number of users. The higher number of users make the approximate round trip is higher,
also longer distance make the approximate round trip is higher too. The number of users is
effect of round trip in a proper LAN either PNWlesL or PNWrL. Nevertheless, the
consideration is on distance between two points which are clients and server. Factors are
seriously caused a signal fault or weak signal is destination between CS and Number of users
(Clients). The similarity of PNWrL and PNWlessL results in packet sent, packet received
and packet lost.
Figure (14) show the round trip minimum to maximum and average for PNWrL and
PNWlessL.
The figure (14) shows the highest and lowest delay time for round trip of signals between
clients and server. The number in milliseconds delay time is very high if compared the
PNWlessL to PNWrL. round trip approximation is shows the better in response and where
to reactivate signals in order to avoid signal weakness.
217
(ICoIT'17)
10thof April, 2017.
120
100 Maximum Manimum
80
60
40 Round Trip
20 approximatrion Wire
0
0 1 2 3 4
Figure (14) show the round trip minimum to maximum and average for PNWr and
PNWless
The weakness signals take it to account as delay time and round trip maximum
approximation. When the round trip is higher than usual is cause weak signal and or delay.
Hence, reactivate the signals by placing repeater hardware. In ping statistics for forty meters
distance there are no lost packet in send and received and showed table (3) in section 4.1.
5.Results and Discussion:
The proposed System evaluated by three seminar-evaluations; the first seminar-evaluation

took place in the main seminar hall at Lebanese French University for wired and wireless
communication with constant size of statics web side by hypertext mark-up language
(HTML) capacity of three Mbytes with distance of twenty meters, but the second seminar-
evaluation took place at computer science lecture with nearly fifty participated students for
forty meters wire and wireless and constant size capacity of three Mbytes. The third seminar-
evaluation took place at main corridor in information technology (IT) department is same as
the previous seminar-evaluation in size capacitybut sixty meters. In all seminar-evaluations,
the proposed system were calculate the results very accurate and rapidly. The Tables 4, 6 and
7 (Pages-14, 16 and 17) are test and illustrate the final calculation (result) in numeric for all
evaluations. It is obvious distance has a role for communication arrival signals from start
point to end point. The best result is wired communication with least distance. The sixty
meters distance signal start be weak and crash wireless, but for wire is hundred meters.
6.CONCLUSIONS
Since new adjustment is introduced in the modern communication, for example LAN-
wired and wireless. The ping is software that free to measure reply time, lost packet and
round trip. Round trip and reply time have a direct relation to number of users. Distance and
number of users are two important factors that affect the result test of round trip (Minimum
and maximum) and ping reply. The evaluation is achieved for a limited number of computers
in LAN-wire and wireless for windows server and clients. The accurate test result found that
numbers of users and distance are affected by using of ping.
218
(ICoIT'17)
10thof April, 2017.
7.REFERENCES
[1] Jerome Henry, WLAN Fundamentals, Cisco Press, 2012.

[2] Julia Andrusenko, Jack Burbank, Feng Ouyang, Commercial Wireless
Communications Trends and the Military, 2015
[3] Everett Murdock, The Computers the Easy Way, USA, 2014
[4] John McNamara, Local area networks: an introduction to the technology, Digital
Press,2014.
[5] Joshua Muscatello, Joshua Martin, Wireless Networks Security, IFMG 250, 2005
[6] B.Forouzan, S.Fergan, Data Communications and Networking, 5th Edition,
McGraw-HillScience/Engineering/Math,2012.
[7] Faraj Kamaran, Automated recruitment system, Sulamani University, 2010
[8] Tim Szigeti, Christina Hattingh, Kenneth Briley,End-to-End QoS Network Design:
Quality of Service for Rich-Media & Cloud Networks, Cisco Press 2013.
[9] Vikrant Kaulgud, IP Quality of Service: Theory and best practices, set labs, India,
2004.
[10] Steve Saed, Average Consensus In Wireless Sensor Networks With Probabilistic
Network Links, Purdue University, India, 2010
[11] W. Stevens. K. Fall, TCP/IP Illustrated Volume 1: The Protocols (Addison-Wesley
Professional Computing (Hardcover)),2011.
[12] W. Stevens. K. Fall, TCP/IP Illustrated Volume 1: The Protocols (Addison-Wesley
Professional Computing (Hardcover)),2011.
219
(ICoIT'17)
10thof April, 2017.
Distributed Data Aggregation protocol for improving lifetime of

Wireless Sensor Networks
Ali K. M. Al-Qurabat
Department of Software, College of Information Technology, University of Babylon - Iraq
alik.m.alqurabat@uobabylon.edu.iq.
Ali Kadhum Idrees
Department of Computer Science, College of Science for Women, University of Babylon - Iraq
ali.idrees@uobabylon.edu.iq.
ABSTRACT
ARTICLE INFO
In Wireless Sensor Networks (WSN), the deployed sensor
Article History: nodes can sense the same measures from the monitored
Received: 17 March 2017 area and forward these redundant measures to the sink
node. Although redundant measures provide better
accuracy but consume a lot of energy during the
Published: 10 April 2017 communication and the processing at the node and the sink,
and thus decrease the lifetime of the WSN. Therefore, the
Keywords: Wireless
Sensor Networks elimination of redundant measures and reducing the
(WSNs), Data communication cost are considered as essential
aggregation, APCA, characteristics during design the WSNs. In this article, a
Network Lifetime. Distributed Data Aggregation (DiDA) protocol for
prolonging the lifetime of WSNs is suggested. DiDA
protocol is an energy efficient approach for a clustered
network. DiDA works into cycles and each cycle
aggregates and reduces data dimensionality by using an
Adaptive Piecewise Constant Approximation (APCA)
method. DiDA was successfully evaluated using
OMNeT++ network simulator and based on sensed data of
a real sensor network. Percentage of Sent data to the Cluster
Head (CH), data accuracy, and energy consumption are the
performance metrics applied to assess the effectiveness of
the DiDA protocol. The conducted simulation results show
that the proposed DiDA protocol decreases the consumed
energy and extending the network lifetime, in comparison
with a method without using data aggregation technique,
whilst keeping the sensed data quality at the sink node.
220
(ICoIT'17)
10thof April, 2017.
1. INTRODUCTION
I
n the last years, a novel type of networks is emerged called Wireless Sensor
Networks (WSNs) due to the fast development in the wireless devices, electronics,
and the technology of embedded computing. A WSN consists of a large number of
tiny low-cost limited-energy devices that can sense, process, store, and transmit data
of surrounding environment with limited capabilities across the network to the sink
node. The most significant resource in the sensor node that impacts on the lifetime of
WSN is the energy provided by the battery which is difficult or impossible to replace
(or recharge) it especially in the remote or hostile environment. One of the biggest
challenges in WSN is the lifetime maximization of the battery [1, 2, 33, 34]. Therefore,
it is important to transmit as less volume of data as possible to the base station (sink).
The sensed data received by the sink node may be similar because there is more than
one sensor monitors the same region. In addition, the large volume of sensed data
generated by dense WSN leads to high processing load on the sink node. Therefore, it
is necessary to improve the energy efficiency of the sensor network to operate over a
long period of time. This can be done by using an energy efficient ways such as Data
aggregation methods [3]. The main objective of data aggregation algorithms is to collect
and accumulate sensed data with removing redundancy from sensed correlated data
generated by closely located neighboring sensors so as to save the energy thus extend
the network lifetime [3, 4].
Data gathering can be either triggered events (such as forest fire and gas or oil leaks
detection [5, 6]) or periodic triggering (such as habitat monitoring [7]). This paper
focuses on the periodic data gathering and aggregation in WSNs. In some specific WSN
applications, the accuracy of the observations is very critical for understanding the
underlying processes. Therefore, in order to design data aggregation algorithms for such
applications, it is very important to ensure the accuracy of the received sensed data by
the sink node.
This article provides the following contributions:
1- A new protocol named DiDA (Distributed Data Aggregation) is proposed to
aggregate the sensed data and prolong the network lifetime in WSNs. It uses an
energy efficient method for data aggregation for a clustered network. DiDA
protocol uses the Adaptive Piecewise Constant Approximation (APCA) method
to aggregate, and reduce the sensed data dimensionality.
2- DiDA protocol is evaluated by OMNeT++ network simulator using extensive
simulation experiments. DiDA has been compared to the results of the method
without using data aggregation technique.
The rest of this paper is organized as follows. Next section exhibits literature review.
Section 3 explains the description of (DiDA) protocol. Protocol evaluation is shown in
Section 4. Finally, we present the conclusion and future works in Section 5.
This section investigates some existing related works to data aggregation in WSNs. In
recent years, several proposed data management works focused on the data aggregation in
221
(ICoIT'17)
10thof April, 2017.
WSNs [8, 9, 10, 35, 36]. The principle objective of aggregating the sensed data is to eliminate
the redundancy in the sensed data and minimize the consumed energy thus extending the
lifetime of the network [11].
However, in order to reduce the data transferring, the works in [12]-[15] used simple
aggregation functions (such as max, min, avg, and sum) for aggregation. These methods do
not consider the correlation among the sensed data. Although they provide a high aggregation
performance but the accuracy of recovered data is badly poor. Therefore, these methods are
inappropriate for those applications that require a high data accuracy [16]. For example,
LEACH [17] protocol divides the network into several clusters. The cluster heads are chosen
during the setup phase whilst the data are aggregated using AVG method at each cluster head
in the steady phase so as to reduce the network data traffic [18].
HEED protocol [19] is the extended version of LEACH protocol. The cost of the intra-
cluster communication and the limits of communication range are included in original
LEACH. It elects the cluster head in a periodic way and based on two parameters: remaining
energy and the degree of the node. The authors in [12] presented EAST method for data
gathering based on spatial-temporal correlations. The nodes are spatially clustered by this
method into groups according to their positions in the region. The temporal correlation is
achieved by the cluster head to aggregate and transmit the sensed data to the sink only in the
case when they exceed a particular threshold. This method is suitable for event-driven data
aggregation [31]. It is not relevant for periodic data sending due to the processing and
communication overhead caused by the dynamic selection of the cluster head [18].
The works in [20]-[23] divided the sensor nodes into different clusters. One or several
nodes in each cluster are chosen as a representative set of nodes for data collecting and
sending whilst deactivating the other nodes in the same cluster. The nodes energy can be
saved significantly with these methods, but this can lead to important data loss due to a large
number of deactivated nodes.
The authors in [24] proposed a round-based clustering scheme that resolves the
transmission of redundant data in the network so as to improve network lifetime. Proposed
scheme works in four phases rounds: initialization, cluster-head selection, clustering, and
data aggregation. Proposed clustering scheme reduces energy consumption, thus increasing
network throughput by dealing with most of the redundant data. Several data aggregation
methods are proposed in [32].
This paper proposes a Distributed Data Aggregation (DiDA) protocol for lifetime
enhancement in WSNs. DiDA aggregates and decreases the data dimensionality using an
Adaptive Piecewise Constant Approximation (APCA) method. DiDA is assessed using
OMNet++ network simulator and based on real sensed data from a sensor network. The
results explain that the suggested DiDA protocol reduces the energy consumption and
prolonging the WSN lifetime compared to the results of the method without using data
aggregation technique while maintaining the quality of sensed data at the sink node.
3. DATA AGGREGATION STAGE USING APCA

The proposed protocol in this article is distributed on the sensor nodes. These nodes are
considered grouped into clusters so as to achieve energy efficient data aggregation with
reduced cost of communication. In this paper, DiDA protocol achieves the data aggregation
222
(ICoIT'17)
10thof April, 2017.
at the level of the sensor node periodically and then transmits the sensed data to the cluster
head. After that, the cluster head sends the data directly to the sink. Figure 1 illustrates the
flowchart of the proposed DiDA protocol. Table 1 explains some parameters used in this
paper.
Start
Initialize the parameters

SMPR ; .
Set periods counter Pc to 1
Is No Sensor is excluded from joining the

Residual Energy > 0 ? current cycle
Yes
Sample Capturing
End
S = Collect measures at SMPR
Segments Constructions Using SW

S W = Sliding Window ( S, )
Data Aggregation
S AP = APCA Transformation ( S W )
Update Residual Energy
Pc = Pc + 1
Figure 1. Flowchart of proposed DiDA protocol.
Table 1: Some parameters used in this paper.

Sampling rate =
Temperature readings series = 1 , . ,
Segment construction using SW(S, ), = 1 , . ,

APCA of , = 1 , . ,
Reconstruction error bound
Sensor id
Remaining energy of sensor n
The PSN consists of nodes(1 , 2 , . . . . , ), each node is responsible for sensing the
data measures of the dynamic physical environment such as humidity, temperature, or
pressure etc. In PSN, the periods are partitioned into time slots. Therefore, each sensor node
n captures the data reading periodically. Consequently, the time-ordered sequence of sensed
data constitutes a time series, = {1 , 2 , . . . , p1 , p }, where is the total number of
temperature readings generated by sensor node every T seconds. The redundant
temperature readings captured by the sensor node increase in two states: short time slot and
slowly variation of a monitored area of interest.
Often, the volume of temperature readings series is very huge. Therefore, it is not practical
to send all the collected raw readings from every sensor node back to the base station due to
223
(ICoIT'17)
10thof April, 2017.
the constrained bandwidth and energy consumption on data sending (radio transmission).
The dimensionality of temperature readings series (which is the number of observed
measures) have a direct proportionality relation with the communication cost. Thus, a smaller
can result in a significant reduction in the communication cost and hence, it will prolong
the lifetime of the sensor network [25]. In this stage, the DiDA protocol aims to perform
dimensionality reduction by removing the redundant data. It exploits the correlation nature
which is temporal among the sensed data of the sensor node efficiently by applying an
Adaptive Piecewise Constant Approximation (APCA) technique. DiDA protocol transforms
the temperature readings series S = {1 , 2 , . . . , p1 , p } that collected during the period to
an APCA representation in order to decrease the dimensionality of series. The APCA divides
the sorted temperature readings series into a set of constant value segments (with a bounded
reconstruction error ) of varying lengths based on data such that their individual
reconstruction errors are minimal. More formally,| (S ) | < , (S ) is the
reconstruction function, and is an error threshold. Long segments are used to represent data
regions of low activity, and short segments are used to represent regions of high activity [26].
The APCA representation of is given as follow:
S = {(dm1 , dr1 ), . . . , (dm , dr )}, dr0 = 0. (1)

The APCA approximates each segment S by a pair (dm , dr ) of two numbers, where
dm is the mean value of temperature readings in the segment which is defined as

=
1 +1
= (2)
1
Whilst dr is the right endpoint of the segment [27].

By using the standard form of APCA with a constant number of segments of varying
lengths can influence on the accuracy of temperature readings. Hence, the problem addressed
here is: for a given temperature readings series S and a given reconstruction error bound ,
find the number of segments to approximate the time series, such that the difference between
any approximation value and its actual value is less than . In our method, we make some
slight modifications on APCA. First, the number of segments will not be constant and
predetermined, but it will be adaptive based on the user specified reconstruction error . In
order to achieve this goal (i.e. making the number of segments adaptive), the sliding window
algorithm is utilized. The reason for making the number of segments adaptive is to increase
the accuracy of approximated measures by using a user-specified reconstruction error.
Second, we modified to represent the length of the segments rather than record the
locations of their right endpoints.
At the end of each period, DiDA protocol will apply the sliding window algorithm on the
collected readings to produce a different number of segments with varying lengths. The
Sliding Window approach is used because it is simple, online, and intuitive [28]. Algorithm
1 represents the process of segment construction using sliding window algorithm.
Algorithm 1. Segments Construction using Sliding Window
Input: ( );
:
Output: S
Process:
1: S (S) //Sorting temperature readings in descending order.
224
(ICoIT'17)
10thof April, 2017.
2: 1 // Starting point.
3: 1 // Number of Segments.
4: ( < )
5: 2
6: (_(S[: + ]) < )
7: +1
8:
9: S [] _ (S[[: + 1]
10: +
11: + 1
12:
13:
After segmenting the temperature readings series using sliding window algorithm, the
produced set of segments S = (1 , 2 , . ,

) is used by algorithm 2 to produce the
APCA representation for temperature readings series S. Algorithm 2 illustrates the process
of dimensionality reduction using APCA.
Algorithm 2. Aggregation Stage using APCA.

Input: S .
Output: S .
Process:
1: 1
2: S
3: 0
4: 0
5: 1 ()
6: + []
7: + 1
8:
9:

10:

11: S _( , )
12:
13: SAP
In the aggregation stage and at the end of every period, each sensor will have a set of
segments S with subsets(1 , 2 , . ,
), and two numbers per segment
( , ) that meet the reconstruction error bound , with no redundant measures.
The problem mentioned above is solved by constructing a set of segments with m subsets
that meet the reconstruction error bound .
4. PROTOCOL EVALUATION
In this section, we provided the framework of simulation and conducted a series of
simulations to assess the effectiveness and the relevance of the proposed protocol.
4.1. Simulation framework
In order to evaluate DiDA protocol, extensive simulations experiments are performed
with discrete event simulator OMNeT++ [29] and based on real sensor data. In these
simulations, we consider N sensors deployed in the lab. Sensors periodically capture local
readings (e.g., temperature) at a specified rate. We assume there is a single cluster head
225
(ICoIT'17)
10thof April, 2017.
located at the center of the lab. The cluster head receives sensed data readings from each
sensor node in the lab periodically via a single hop. DiDA protocol is distributed at each
sensor node and it is based on the dataset of Intel Berkeley Research Lab [30]. PSN in this
Lab includes 54 Mica2Dot sensors.
The sensed data of the weather (such as temperature, humidity, and light) are periodically
collected by these sensors once each 31 seconds. In our simulation, the sensor nodes use a
log file contains about 2.3 million readings collected previously by Mica2Dot sensor nodes
in the Lab. This article uses only one measure of sensor node measurements: temperature4.
There are 7 sensor nodes didnt used in our simulation because its data may be missed or
truncated. Therefore, the results are the average of 47 sensor nodes. Table 2 gives the selected
parameters settings.
Table 2: Simulation Parameters for PSN initialization.

Parameter Value
PSN size 47 nodes
20, 50 and 100 readings
0.03, 0.05, 0.07 Reconstruction error bound
E elec 50 nJ/bit
amp 100 pJ/bit/ m 2
In the experimental simulations, some performance metrics are applied to assess the
effectiveness of the DiDA protocol such as Percentage of Sent data to the CH, data accuracy,
and energy consumption. DiDA protocol uses the same energy consumption model discussed
in [30]. Energy consumed by the sensor node is caused by the communication unit (data
transmission and reception). Therefore, the cost of transmission is calculated for a m bits
message and for a distance d as follow
ETX (m,d) = Eelec m + amp m d 2. (7)
The energy consumption required for reception m bits is calculated as follow

ERX(m,d) = Eelec m. (8)
These experiment simulations consider the length of data reading m equal to 64. In fact,
we set the value of m to 64 bits because we consider each data reading as a double data type
that takes 8 bytes. In the case of transmission, 64 bits are added to m bits message which
corresponds to the frequency of data reading m. The length of the transmitted data packet is
calculated as follow DataPacket length = (Number of readings in the data set 2) 64 bits.
Hence, the packet length is the number of reading in the sensed data set with their frequencies
multiplying by 64 bits.
4.2. Performance comparison and analysis

Several experiments are achieved in this section to show the performance of DiDA
protocol. DiDA is distributed at each sensor node in the PSN. Every node reads real
temperature readings periodically and aggregate them using APCA. Furthermore, DiDA
protocol is compared to the results of the method without using data aggregation technique.
4
The other is done by the same manner.
226
(ICoIT'17)
10thof April, 2017.
1) Percentage of Sent data to CH after applying aggregation stage: The result of the
aggregation in this stage depends on the chosen reconstruction error bound , the number of
the collected measures in the period, and the changes in the monitored region. Figure 2
illustrates the Percentage of Sent data to the CH without and with applying aggregation stage
at the end of simulation by every sensor node using DiDA protocol compared with the results
of the method without using data aggregation technique. The results show a maximum of
25% of the data sent to CH after applying the aggregation stage by DiDA protocol at each
period, whilst the rate is equal to 100% without applying the aggregation step. Therefore,
DiDA protocol decreases the volume of sensed data transmitted to the CH by removing the
duplicated measures at every period successfully. It can be seen at the step of aggregation,
when the or increases, the sent data are decreases. The reason behind this is remove a
larger amount of similar data by using APCA method.
120
DiDA ( =20 ) DiDA ( =100 )
DiDA ( =50 ) WithoutAggregation
100
80
60
40
20
0
0.03 0.05 0.07
Threshold
Figure 2. Percentage of Sent data to the CH.
2) Data Accuracy: In this experiment, the data accuracy is considered as an essential

performance factor in WSNs. In this paper, it represents the percentage of data loss after
applying the aggregation inside the sensor node. It has a significant impact on the final
decision that will be taken by the end user. It can be considered as the error of aggregation.
In this section, the proposed DiDA protocol is compared with the results of the method
without using data aggregation technique. Table 3 shows the results of data accuracy without
and with using our technique DiDA. It can be seen that our protocol provides good results
from the data accuracy point of view. In the worst case, the percentage of data which are not
received by the sink are almost 0.298 % (i.e. =100). This percentage is not important in
comparison with the received data by the base station. Therefore, it can be noted that our
protocol is able to get rid of the redundant data while maintaining the accuracy of received
data by the end user. Furthermore, the data loss percentage minimizes when and decreases
because of using efficient method for data reduction.
227
(ICoIT'17)
10thof April, 2017.
Table 3: Data Accuracy (percentage of data loss after applying the aggregation inside the sensor).
Percentage of Lost Measures (%)
Without
Threshold DiDA DiDA DiDA Aggregation
(p=20) (p=50) (p=100)
0.03 0.0013 0.004442553 0.298617021 0

0.05 0.0013 0.004442553 0.298617021 0
0.07 0.0013 0.004442553 0.298617021 0
3) Energy consumption: The energy consumption is another performance factor for

evaluating our protocol in comparison with the results of the method without using data
aggregation technique. Figure 3 shows the energy consumption at each sensor node. The
energy consumption minimized when the transmitted data to the cluster head minimized.
DiDA protocol decreases the consumed energy while maintaining the integrity of
information by applying an energy efficient data aggregation approach. Figure 6 display the
energy consumption without and with using our technique DiDA for various and values.
The conducted simulation results explain that DiDA is the better from the energy
consumption point of view. DiDA saves more energy when both or increase.
Figure 3. Energy consumption at each sensor node.
5. CONCLUSION AND FUTURE WORKS

The increased use of the sensor networks in several applications led to provide a huge
amount of data which are transmitted across the network. This is greatly contributed in
decreasing the network lifetime due to the high communication cost. Therefore, the energy
efficient data aggregation approaches are very necessary for eliminating the replicated data
in the network. In this article, we propose a protocol named DiDA (Distributed Data
Aggregation) to extend the lifetime of the WSNs. This protocol uses Adaptive Piecewise
Constant Approximation (APCA) as an energy efficient data aggregation, thus improving the
network lifetime. The simulation results that based on real data of the sensor network using
OMNet++ network simulator show that DiDA protocol outperforms the method without
using data aggregation technique in terms of data reduction percentage that will be
228
(ICoIT'17)
10thof April, 2017.
transmitted to the cluster head, energy consumption, and acceptable data accuracy. In future,
we plan to apply the data aggregation into two levels: sensor node level and aggregator node
level (cluster heads). The first level is responsible for temporal correlation among the data
inside the sensor node whilst the second level deals with data correlation among neighboring
nodes. In addition, we plan to compare our proposed method with two existing methods in
the literature [35] and [36].
REFERENCES
[1] S. Misra, I. Zhang, and S. C. Misra, Guide to wireless sensor networks. Springer Science
& Business Media, 2009.
[2] A. K. Idrees, K. Deschinkel, M. Salomon, and R. Couturier, Distributed lifetime
coverage optimization protocol in wireless sensor networks, The Journal of
Supercomputing, vol. 71, no. 12, pp. 4578 4593, 2015.
[3] X. Zhai and T. Vladimirova, Data aggregation in wireless sensor networks for lunar
exploration, in 2015 Sixth International Conference on Emerging Security Technologies
(EST). IEEE, 2015, pp. 3037.
[4] S. Pino-Povedano, R. Arroyo-Valles, and J. Cid-Sueiro, Selective forwarding for
energy-efficient target tracking in sensor networks, Signal Processing, vol. 94, pp. 557
569, 2014.
[5] L. Yu, N. Wang, and X. Meng, Real-time forest fire detection with wireless sensor
networks, in Proceedings. 2005 International Conference on Wireless Communications,
Networking and Mobile Computing, 2005., vol. 2. IEEE, 2005, pp. 12141217.
[6] M. Dalbro, E. Eikeland, A. J. int Veld, S. Gjessing, T. S. Lande, H. K. Riis, and O.
Srasen, Wireless sensor networks for off-shore oil and gas installations, in Sensor
Technologies and Applications, 2008. SENSORCOMM08. Second International
Conference on. IEEE, 2008, pp. 258263.
[7] A. Mainwaring, D. Culler, J. Polastre, R. Szewczyk, and J. Anderson, Wireless sensor
networks for habitat monitoring, in Proceedings of the 1st ACM international workshop
on Wireless sensor networks and applications. ACM, 2002, pp. 8897.
[8] B. Yu, J. Li, and Y. Li, Distributed data aggregation scheduling in wireless sensor
networks, in INFOCOM 2009, IEEE. IEEE, 2009 , pp. 21592167.
[9] Y. Zheng, K. Chen, and W. Qiu, Building representative-based data aggregation tree in
wireless sensor networks, Mathematical Problems in Engineering, vol. 2010, 2010.
[10] J. M. Bahi, A. Makhoul, and M. Medlej, Data aggregation for periodic sensor
networks using sets similarity functions, in 2011 7th International Wireless
Communications and Mobile Computing Conference. IEEE, 2011, pp. 559564.
[11] M. A. Sharaf, J. Beaver, A. Labrinidis, and P. K. Chrysanthis, Tina: a scheme for
temporal coherency-aware in-network aggregation, in Proceedings of the 3rd ACM
international workshop on Data engineering for wireless and mobile access. ACM, 2003,
pp. 6976.
[12] F. Ren, J. Zhang, Y. Wu, T. He, C. Chen, and C. Lin, Attribute-aware data
aggregation using potential-based dynamic routing in wireless sensor networks, IEEE
transactions on parallel and distributed systems, vol. 24, no. 5, pp. 881892, 2013.
229
(ICoIT'17)
10thof April, 2017.
[13] H. Luo, H. Tao, H. Ma, and S. K. Das, Data fusion with desired reliability in wireless
sensor networks, IEEE Transactions on Parallel and Distributed Systems, vol. 22, no.
3, pp. 501513, 2011.
[14] X. Xu, X.-Y. Li, P.-J. Wan, and S. Tang, Efficient scheduling for periodic
aggregation queries in multihop sensor networks, IEEE/ACM Transactions on
Networking (TON), vol. 20, no. 3, pp. 690698, 2012.
[15] W. B. Heinzelman, Application-specific protocol architectures for wireless
networks, Ph.D. dissertation, Massachusetts Institute of Technology, 2000.
[16] C. Liu, J. Luo, and Y. Song, Correlation-model-based data aggregation in wireless
sensor networks, in Fuzzy Systems and Knowledge Discovery (FSKD), 2015 12th
International Conference on. IEEE, 2015, pp. 822827.
[17] W. R. Heinzelman, A. Chandrakasan, and H. Balakrishnan, Energyefficient
communication protocol for wireless microsensor networks, in System sciences, 2000.
Proceedings of the 33rd annual Hawaii international conference on. IEEE, 2000, pp. 10
pp.
[18] I. H. Brahmi, S. Djahel, D. Magoni, and J. Murphy, A spatial correlation aware
scheme for efficient data aggregation in wireless sensor networks, in Local Computer
Networks Conference Workshops (LCN Workshops), 2015 IEEE 40th. IEEE, 2015, pp.
847854.
[19] O. Younis and S. Fahmy, Heed: a hybrid, energy-efficient, distributed clustering
approach for ad hoc sensor networks, IEEE Transactions on mobile computing, vol. 3,
no. 4, pp. 366379, 2004.
[20] J. Yuan and H. Chen, The optimized clustering technique based on spatial-
correlation in wireless sensor networks, in Information, Computing and
Telecommunication, 2009. YC-ICT09. IEEE Youth Conference on. IEEE, 2009, pp. 411
414.
[21] Z. Liu, W. Xing, Y. Wang, and D. Lu, An energy-efficient data collection scheme
for wireless sensor networks, in Advanced Communication Technology (ICACT), 2013
15th International Conference on. IEEE, 2013, pp. 6065.
[22] L. A. Villas, A. Boukerche, H. A. De Oliveira, R. B. De Araujo, and A. A. Loureiro,
A spatial correlation aware algorithm to perform efficient data collection in wireless
sensor networks, Ad Hoc Networks, vol. 12, pp. 6985, 2014.
[23] T. D. Le, N. D. Pham, and H. Choo, Towards a distributed clustering scheme based
on spatial correlation in wsns, in 2008 International Wireless Communications and
Mobile Computing Conference. IEEE, 2008, pp. 529534.
[24] K. T.-M. Tran and S.-H. Oh, Uwsns: A round-based clustering scheme for data
redundancy resolve, International Journal of Distributed Sensor Networks, vol. 2014,
2014.
[25] C. Liu, K. Wu, and J. Pei, An energy-efficient data collection framework for wireless
sensor networks by exploiting spatiotemporal correlation, IEEE Transactions on
Parallel and Distributed Systems, vol. 18, no. 7, pp. 10101023, 2007.
[26] A. Zifan, M. H. Moradi, S. Saberi, and F. Towhidkhah, Automated segmentation of
ecg signals using piecewise derivative dynamic time warping, International Journal of
Biological and Life Sciences, vol. 1 , pp. 181185, 2007.
230
(ICoIT'17)
10thof April, 2017.
[27] Y. Wang, P. Wang, J. Pei, W. Wang, and S. Huang, A data-adaptive and dynamic
segmentation index for whole matching on time series, Proceedings of the VLDB
Endowment, vol. 6, no. 10, pp. 793804 , 2013.
[28] Y. B. Yahmed, A. A. Bakar, A. R. Hamdan, A. Ahmed, and S. M. S. Abdullah,
Adaptive sliding window algorithm for weather data segmentation, Journal of
Theoretical and Applied Information Technology, vol. 80, no. 2, p. 322, 2015.
[29] Omnet++ discrete event simulator. 2016, [Online; accessed 23August-2016].
[30] S. Madden, Intel berkeley research lab, 2004, [Online; accessed 25-August-2016].
[Online]. Available: http://db.csail.mit.edu/labdata/ labdata.html.
[31] M. Kamarei, M. Hajimohammadi, A. Patooghy, and M. Fazeli, An Efficient Data
Aggregation Method for Event-Driven WSNs: A Modeling and Evaluation Approach,
Wireless Personal Communications, vol. 84, no. 1, pp. 745764, 2015.
[32] Jesus, P., Baquero, C. and Almeida, P.S., A survey of distributed data aggregation
algorithms, IEEE Communications Surveys & Tutorials, vol. 17, no. 1, pp. 381-404,
2015.
[33] Idrees AK, Deschinkel K, Salomon M, Couturier R. Coverage and lifetime
optimization in heterogeneous energy wireless sensor networks. ICN 2014. 2014 Feb
23:60.
[34] Idrees AK, Deschinkel K, Salomon M, Couturier R., Perimeter-based coverage
optimization to improve lifetime in wireless sensor networks, Engineering
Optimization, vol. 48, no. 11, pp. 1951-72, 2016.
[35] Bahi, J.M., Makhoul, A., Medlej, M.: A two tiers data aggregation scheme for
periodic sensor networks. Adhoc & Sensor Wireless Networks, vol. 21, no. 1, 2014.
[36] Harb, H., Makhoul, A., Couturier, R. and Medlej, M., ATP: An Aggregation and
Transmission Protocol for Conserving Energy in Periodic Sensor Networks, In IEEE
24th International Conference on Enabling Technologies: Infrastructure for
Collaborative Enterprises (WETICE), pp. 134-139, 2015.
231
(ICoIT'17)
10thof April, 2017.
DSP for Mobile Wireless LTE System: AReview

Ashwaq Q. Hameed
Electrical Engineering Dept.
University of technology- Iraq
50058@uotechnology.edu.iq
Salman goli
Computer Engineering Dept.
University ofKashan- Iran
salmangoli@kashanu.ac.ir
Mohammed K. shoubith
Communication Engineering Dept.,
University of Kashan- Iran
Mohammed.jorany@gmail.com

Long term evolution (LTE) is the last step towards the 4G of
Article History: radio technologies designed to increase the capacity and
Received: 17 March 2017 speed of cellular networks. With the appearing of 4G LTE
networks, there has been increasing interests and special
attention is paid to the performance and power characteristics
Published: 10 April 2017 as compared with 3G networks. This paper provides detailed
information about mobile wireless MIMO-LTE system. It is
Keywords: LTE, MIMO, also presents a brief explanation of the characteristics,
OFDM, OFDMA, smart performance and state of art techniques used within mobile
antenna technology, wireless MIMO-LTE system. There are several types of
wireless network. commercial digital signal processing (DSP) chips available in
marks which can be used with MIMO- LTE system.
Therefore, this work will be focuses on the applications of
commercial DSPs in MIMO-LTE system and taking into
consideration types and characteristics where can enables
efficient baseband processing for 3G,4G and MIMO-LTE
wireless system equipment with its scalable, high-
performance architecture.
1. INTRODUCTION.
T
he new advanced networks always depends on many technological and
economic factors such as limited frequency resource availability, the cost and
operational conditions of developed radio equipment and optimization process.
Efficient analytical techniques are required for planning and analyzing the wireless
networks[1]. In regards to the third generation (3G),it has several problems such as,
232
(ICoIT'17)
10thof April, 2017.
technology is unripe, and resulting there is a big problems still exist from the base station
(BS) to the mobile station (MS). The speed of transmission is lowand instability. Speed
theoretically ranging from 384k / s to 480k / s. But the average speed in practicein
download from Internet only about 10K. Network coverage is bad due to the equipment
of base station (single station about 2m, 500kg-800kg), therefore, the coverage of large
areas will bring significant obstacles. The roaming is not international. For example,the
mobile brought to America or Europe, cant use 3G. The most extensive Global roaming
in 3G network is WCDMA, including the more developed regions in Europe and
America, Australia, etc. It is due to the many disadvantages of 3G, people have come
to expect 4G to solve these problems, which can provide higher data rates, greater
capacity and bandwidth [2]. LTE system also depends on many factors and is affected
by many factors which we spell out in this view. 4G LTE is the newtechnologyin the
mobile communicationsof cellular network that Equip us anexcellent speedin data
services. 4G technology is the acronym which refers to fourth generation mobile
communication technology. 4G LTEenabled us to transmit HD video and big image in
addition to other features[3]. Speed transmission theoretically Up to 100 Mbps. The
advantage of the bandwidth in 4G network is much higher than the bandwidth of 3G
network systems, Where approximately 100 MHz spectrum, That means roughly equal
20 times the W-CDMA 3G networks [4]. LTE can increase capacity by using standard
antenna techniques, widespread deployment and reliance on MIMO (Multiple-Input
Multiple-Output) antenna techniques[5]. Due to much increased data rates and reduced
latencies, LTE is pushing DSP based architectures to the maximum can be. The use of
higher order antenna systems are driving algorithmic demands forward at an exponential
rate [6].With the other side, using the kind optimal and appropriateof DSP chip in the
system will advance the capabilities of wireless broadband equipment, performance and
power savings, and other. There are many kinds of DSP chips can be integratedand
programmed with our work in LTE MIMO system to get the best results. This is due to
any types we'll use, and any types commensurate with the system. In this paper we'll
take some of the details about Texas Instruments DSP chip and Controllers
TMS320CS6, Qualcomm Snapdragon So Cs 835 and MSC8156 High-Performance Six-
Core DSP
1.1 background of LTE.
Mobile communications have a major development over the past years. Theintegration in
performance of technologies are rapidly moving toward. Reality ofMobile broadband leads
us togrow quicklyin internet generation to having broadband access in everywhere. Estimates
show1.8 billion people have broadband since 2012, some two-thirds will be mobile
broadband consumers and most of these services offered by HSPA (High Speed Packet
Access) and LTE (Long Term Evolution)[7]. The main sign of this project is to improve the
Universal Mobile Telecommunications System (UMTS). Scientists are busy in establishing
233
(ICoIT'17)
10thof April, 2017.
a mobile broadband highwaywith more features that will be that suits and supports the
growing demand in the future, like efficiency improvement,enhancement ofall service,
reduce costs significantly and better integrationin (technical and performance). In December
1998(Third Generation partnership project) established. The members of 3GPP project are
ARIB/TTC (Japan), China communication standard association, Telecommunications
industry association (North America) and telecommunication technology associate (South
Korea). Because of the failure to meet the requirements fully by 3G, The need necessary to
build up the fourth generation 4G [8]. It is worth mentioning that MIMO technology as
shown in figure 1is the important basics to building LTE technology, where it is
characterized by:
Figure 1: MIMO - Multiple Input Multiple Output [9]
1. Multiple antenna can be used to get rid of the detrimental effects of multi-path.
2. Superior Data Rates.
3. Increases throughput and range [10].
1.2LTE System Upgrade Path.
From the theoretical side, 4G is much faster than 3G. But before we start study the
difference between 4G speed and 3G speed, we must know first versions of 4G. When talking
about 4G, We will face some confusing things. Basically, there is 4G and 4G LTE. A lot of
people seem to think LTE is true 4G technology. Generally, if the 4G network described
without mentioning LTE, It most likely talking about a High Speed Packet Access (HSPA)
network. A faster version of the 3G GSM network is the HSPA network. But it's not faster
than LTE network [4]. LTE offers a major growth in cellular technology and it provides high
speed data and media transport as well as high capacity voice support well currently. Actually
234
(ICoIT'17)
10thof April, 2017.
LTE provides way to develop UMTS network towards fourth generation (4G) mobile
networks with today's 2G and 3G networks [11]. Transitional from 3G to 4G it will take a
long time, as is the case with 2G to 3G. Whereas, the mobile operators must look for the best
and fastest solutionsthat raise the performance and efficiency 3G networks while meeting the
requirements of the fourth generation development without upgrading all equipment [12].
And we can follow the evolution of LTE by embracing GSM (Global System for Mobile
Communication), GPRS (General Packet Radio Service) and EDGE (Enhanced Data rates
for GSM Environment) as well as WCDMA (Wide Band Code Division Multiple Access)
and now HSPA [9]. LTE It is an excellent design to providea very high bandwidth up to
several megabits, wherever it is more efficient use of the radio spectrum, reduce latency, and
enhancement mobility. The aim of this development that subscriber's interaction significantly
with the network with the (increase in demand and the speed of response) for mobile
multimedia services. In this,it became possible for subscribers accessreadily to their Internet
services, such as online television, video streaming, and interactive gaming, etc. [13]. Figure
2 describes the 3GPP group technical specifications.
Figure 2: LTE upgrade path. [10]
Table 1: Comparison of 2G, 3G, 3.5G and 4G [11]
G Technology Speed Features
1G AMPS,NMT, 14.4 Kbps This generation is used to transmit voice only

TACS (peak)
235
(ICoIT'17)
10thof April, 2017.
2G 9.6/14.4 Kbps
TDMA,
In one single channel, allowing multiple users by using
CDMA multiplexing. also during 2G, the send data are used in
phones side by side voice
3G CDMA 2000 3.1 Mbps (peak)

(1xRTT, 500-700 Kbps
EVDO) In this generation appeared services of multimedia in
UMTS, EDGE addition to previous service. Also 3G supports different
types of device such as (Telephones, PDAs, etc.)
3.5G HSPA 14.4 Mbps This generation supports increased throughput, larger
(peak) than the previous. In order to support the growing needs
1-3 Mbps of subscribers to data.
4G WiMax 100-300 Mbps With fourth generation find high Speed, high definition
LTE (peak) streaming, higher capacities and higher performance, In
addition to the increase of cover spaces.
1.4 LTE System Key Features.

What is the reason for the need to 4G LTE? To answer this question we must be
understand several limitations of 3G [4], such as below:
It is difficult in 3G provide higher data rates due to CDMA technology.

Multimedia need increase data rate and high bandwidth to meet its requirements.
Allocation of spectrum is limited.
Limited coverage due to access methods in 3G.
We can enumerate key features of LTE as follows:
1. Improvement the air interface allows increased data rates: LTE provides greater
Enhancement in performance of the system by using (OFDM) and (OFDMA) technology
rather than (WCDMA).
2. High Spectral Efficiency: LTE it gives a significant improvement in spectral efficiency
and it gives a field to mobile operators to increase number of subscribers within existing
spectrum with a detraction cost of delivery [14].
3. Reduced Latency: LTE technology has succeeded in reduces round trip times to 10ms or
less, which makes LTE provide more interactive and real time to services[15].
4. Flexible Radio planning: LTE can achieve optimal performance in a cell size of up to 5
km as an effective and successful than performance in cell size of up to 30 km radius but
performance is limited in cell sizes up to 100 km radius [16].
5. Scope of Services: smooth seamless service Available in LTE and users can access basic
data services Even if they are outside the coverage areas[12].
236
(ICoIT'17)
10thof April, 2017.
2. MIMO LTE DSP System Model.
The basic idea of the research is writing LTE system as program and apply it to an efficiently
DSP chip, later added noise and work to improve the system to adapts and raise his
performance. There are many kinds of DSP chip can be used it in practical model
construction, we will take idea to some of it:
A. Texas Instruments DSP chip and Controllers TMS320CS6, as shown in figure 3 :
Figure 3:Texas Instruments DSP chip and Controllers TMS320CS6 [17]
DSP, DSC Multicore Fixed and Floating.

(Clock Frequency) as Maximum: 1000 MHz, 1200 MHz.
(Memory Size) of Program: 256 kB.
Size of (Data RAM): 6 MB.
(Supply Voltage) in Operating: 900 mV to 1.1 V.
Number of I/O: 16 I/O.
B. Qualcomm Snapdragon SoCs 835, as shown in figure 4 :
Figure 4: Qualcomm Snapdragon SoCs 835 [17]
237
(ICoIT'17)
10thof April, 2017.
CPU 4* Kryo 280 performance 2.45GHz.

Memory 2x 32-bit @ 1866MHz 29.9GB/s.
Integrated Modem Snapdragon X16 LTE (Category 16/13) DL = 1000Mbp.
C. MSC8156 High-Performance Six-Core DSP, as shown in figure 5 :
Figure 5: MSC8156 High-Performance Six-Core DSP [17]
Six Star Core subsystemswith (DSP) core, each onereach to 1 GHz and 512 KB.cache
is unified (L2 /M2) memory.
Two DDR 64-bit,reach to (800 MHz data rate).
Voltage: (1-volt core, 2.5, 1.8/1.5-volt I/O).
D. Texas Instruments (TI) - multi-core DSP architecture, as shown in figure 6:
Figure 6: TMS320C66x (C66x) [17]
TMS320C66x (C66x).
(Clock speeds) is available from (1.0) to (1.25 GHz).
There are additional accelerators for (3G chip-rate) and (4G bit-rate) processing.
238
(ICoIT'17)
10thof April, 2017.
2.1 Comparison between DSP Chips
We can clarification some differences between the DSP chips based on CPU core, Clock
Frequency, Memory Size, and Supply Voltage as shown in table 1:
Table 1: Comparison between DSP Chips
Texas Qualcomm MSC8156 Texas

Instruments Snapdragon High- Instruments -
DSP chip and SoCs 835 Performance multi-core
Controllers Six-Core DSP DSP
TMS320CS6 architecture
CPU Core Multicore Fixed CPU 4* Kryo Six-Core up to 8 cores

and Floating
Clock 1000 - 1200 1900 2400 800 MHz 1000 1250

Frequency MHz MHz MHz
supports 2 unified 512 KB shared with all
channel L2 cache cores,memory
Memory Size 256 kB
LPDDR4X- memory size 4MB
1866 memory
voltage and
frequency 85 mV - 1.8 V
Supply Voltage 900 mV to 1.1 V 1-volt core, 2.5,
Varying to
1.8/1.5-volt I/O
reduce power of
CPU
3Mobile WirelessLTE System over DSP.

DSP Code optimized for memory or speed,or both.We will discuss ways and benefits for
optimization for speed. This can be done in many methods and different levels as shown in
figure 7:
1-Design level: In highest level, the most appropriate design may be best to benefit the most
as possible of the available resources. Selection the efficient algorithms will benefit in the
implementation of this design. The architectural design effects on the performance of a
239
(ICoIT'17)
10thof April, 2017.
system. The choice of algorithm it is the largest impact on the design. So, the first thing that
must be done choice the appropriate and most efficient algorithm. A "fully optimized"
program it is the most difficult in terms of understanding,thereforemay contain faults larger
than small programs [18].
2-Source code level: To improve performance, poor quality coding should be avoiding. After
that, some improvement in encoding can be actually reduce maintenance times. Some, but
not all.Taking into consideration, nowadays improvement compilers can improve the system
[19].
3-Build level: Between the compile and source level, directives can affect performance
options in a positive way in thecompiler and the sources of codes, such as disables the
unneeded codes (software features) by using preprocessor, or improvement for processor
models or increasecapabilities of hardware [19].
4-Compile level: Use the best and most efficient compiler leads to ensure that the executable
program Instead of the loss of time in the proceedings unnecessary [18].
5-Assembly level: At lowest level, assembly language using in writing codes, designed for a
particular hardware platform can be the best and most efficient in several ways. Many
operating systems used on embedded systems better to write in the assembler language due
to this reason. Programs (What is meant here is not programs that characterized it small) are
very rare written from the beginning to end in assembly due to it takes time and increase the
cost [18].
6-Run time: Run time optimization can be done by compilers and assembler, In addition to
modifying of code by itself as response to increase the efficiency of the system. Some designs
of CPU can perform many optimizations at runtime [19].
Figure 7: PUSCH (Physical Uplink Shared Channel) [18]
240
(ICoIT'17)
10thof April, 2017.
Figure 8: Processing blocks at PUSCH receive[18]
4 LTE DSP System Model.
The high performance can be obtained by implement the communication system using DSP
chip. This requires several important steps for the new system to be effective and integrated
in terms of construction. And in order to clarify the new system in a simplified formatas
shown in figure 9.
Figure 9: Block Diagram of DSP Model
241
(ICoIT'17)
10thof April, 2017.
LTE - DSP System Model consists of several major parts, the most important are: LTE
program, it includes a multiple transmitter andreceivers (MIMO), It could be a mathematical
as follows:
For MIMO channel model, described by matrix-type equation [20]:

y(a) = =0 ()( ) (1)
Y = H*X + n (2)
11 12 13
H= { 21 22 23
31 32 (, )
} (3)
Where y represent the received signal, H matrix represent channel, Xrepresent transmit
signal, nrepresent noise andh, k represent the location of the downlink beam from base station
(BS) to mobile station (MS), and a represent the number of users who connect with the
system.And drepresent subcarrier number, erepresent transmit antenna number [21].
And the noise can be as:

n= 2 1 (4)
Where2 the variance of additive, and Ir1 is an identity matrix of the size r1.MIMO channel
capacity using Shannons mutualinformation modifying of all input contrast matrices where
BA isthe bandwidth of the system.And capacity expression becomes:

C= BA log2 det[ Iz + H ] (5)
2
Channel transfer matrixH can be diagonal as:

H = UD (6)
WhereD= diag( 1, 2, 3, , 0,0 )
andU and V are unitary matrices [4].
The received signal at the downlink for each user a can be represented as the following
equation:
Ri,a = Fi,aXi,a + =0 Fj, a Xj, a + Na(7)
Where the j=0 C , C = number of neighboring cell and j i, the pi , pj is the transmit
power ofdonor BS and neighboring BSs, also, the Fi,aFj,a is the fading channel gain for
donor and neighboring cellrespectively and Na isAdditive white Gaussian noise(AWGN) for
each a- user.
When the massages are sent over channel:
242
(ICoIT'17)
10thof April, 2017.
Tm+1= Tmarrival_time(Tm)(8)
packet_no(Tm+1) = 1 (9)
WhereTm+1is the process time of sending massage over channel [22].
The second important part is conversion LTE program to a C++ language with everything
contained in the program from equations and algorithms etc. Then insert the converter
program to DSP chip after choosing the appropriate type, and this is the third part. Fourth
part is the source of signal and noise,which is provided system by Signal, and at the same
timeprovides the system by noise for the purposes of the experiment and development. Let
it be, for example, a laptop. So we will need to useoscilloscope in final stage to measure the
impact ofnoise on the system. And then try to reduce noise effect, which mean that
decreasingbit error rate (BER) and increasesignal-to-noise ratio (SNR)tomake the system
more receptive and less affected. Where SNR can be a form mathematical as:
|,|2
SNi,j= (10)
+ =0 |,|2
Where the SNi,jis signal to noise ratio [23].
5Mobile Wireless LTE- MIMO system Applications.
There are a lot of applications and features of LTE MIMO, we can see some of them below:
a.In (Smart Home): LTE MIMO system it will be of high speed network, a send information
which does not need for cables, 4G communication allows more Seamless connectivity
between people. Based on 4G mobile technology. All things are going to be joined on the
Internet, and this is called internet of thinks.[24]
b. In (live TV): 4G way has many advantages, like long distance transmission, easy to use,
low in cost of equipment. So television media news that require a direct transfer.Alsousing
4G way will facilitate (power supply conditions) in harsh terrain.
c. In (field of Security): The (network monitoring) has become one of the main things in
environment of security. The Monitoring Currently becoming very obvious than the
previous, it trending HD. Who is looking for security want to develop cloud storage due to
large amount of data generated by security products program, supported large bandwidth and
stability of wireless speed are the first important,such as LTE MIMO system. [25]
d. Other Applications: Health care and emergency services has become easier, cheaper and
faster with LTE MIMO network, and other important services such as fight against floods,
earthquakes, flight and Airports [26].
Conclusion
243
(ICoIT'17)
10thof April, 2017.
some new technologies has been added to 4G wireless networks , Thistechnologies makes
4G more efficiency and speed than the 3G network. Covering the world and seamless service
are the basic goals of 4G,and it will achieve its goals through the optimal use of spectrum,
dynamic and capacity of broadband, etc. Where we can say that a full update for 3G. 4G can
use broadband more efficient to improve the(speed of transmission) and the (flow of
information) within the time unit, and the (e-commerce). Taking into consideration
Concurrent application of digital signal processors (DSP) in wireless base station designs to
continue as an effective design approach. This will ensure final products that are not only
scalable and cost-effective, but flexible and reconfigurable across multiple evolving
standards.
REFERENCES
[1] RimvydasAleksiejunas, "Simulation Framework for MIMO LTE Network Performance
Analysis," Innovative Infotechnolojies for science, Vol. 2(15), 2013, pp. 9-13.
[2] Z. Guennoun, and M. Jaloun, Wireless Mobile Evolution to 4G Network, Wireless

Sensor Network, vol.02, pp. 309-317, 2010.
[3] J. Huang, F. Qian, A. Gerber, Z. M. Mao, S. Sen, and O. Spatscheck. "A Close
Examination of Performance and Power Characteristics of 4G LTE Networks", In
MobiSys, 2012.
[4] L. Hanzo, J. Akhtman, L. Wang, and M. Jiang. MIMO-OFDM for LTE, Wi-Fi, and
WiMAX: coherent versus non-coherent and cooperative turbo-transceivers.
Chichester, West Sussex, U.K. ; Hoboken, N.J: Wiley, 2011.
[5] Brian Stetler, F.C.C. Chairman: We Need to Auction Off More Spectrum, Gadgetwise
(blog), New York Times, January 7, 2011
[6] S. Sesia. I. Toufik, M. Baker, LTE The UMTS Long Term Evolution: A Pocket
Dictionary of Acronyms, Wiley, 2009
[7] Sanjay Kumar Sarkar, A Long Term Evolution (LTE) Downlink (DL) inspired channel
simulator using the SUI 3 channel model, Blekinge Institute of Technology, 3,15,
August, 2009
[8] L. Hanzo, J. Akhtman, L. Wang, and M. Jiang. MIMO-OFDM for LTE, Wi-Fi, and
WiMAX: coherent versus non-coherent and cooperative turbo-transceivers.
Chichester, West Sussex, U.K. ; Hoboken, N.J: Wiley, 2011.
[9] J. G. Andrews, W. Choi, and R. W. Heath. Overcoming interference in spatial

multiplexing MIMO cellular networks. Wireless Communications, IEEE 14(6)
(2007) 95104.
244
(ICoIT'17)
10thof April, 2017.
[10] Jim Zyren, Dr. Wes McCoy, "Overview of the 3GPP LTE physical layer", freescale TM,
3, 2007.
[11] http://eeecommunity.blogspot.com .
[12] A. N. Kulkarni, S. K. Sharman, A band Antenna with MIMO Implementation for U4G
LTE Wireless Devices, In: Proceedings of the IEEE APSURSI, vol.3, pp. 2215-2218,
2011.
[13] S. Sesia, I. Toufik, M. Baker (eds), LTE The UMTS Long Term Evolution: From
Theory to Practice, Wiley, 2009
[14] Kalis, A., A. G. Kanatas, and C. B. Papadias, Parasitic Antenna Arrays for Wireless
MIMO Systems, Springer, New York; Heidelberg; Dordrecht; London, 2014
[15] M. Chang, Z. Abichar, and C.-Y. Hsu, Wimax or lte: Who will lead the broadband
mobile internet? IT Professional, vol. 12, no. 3, pp. 26 32, 2010.
[16] R. Nossenson, Long-term evolution network architecture, in Microwaves,

Communications, Antennas and Electronics Systems, 2009.
[17] Wikipedia the free encyclopedia, 4G Available: Wikipedia online:

http://en.wikipedia.org/wiki/4G [Accessed: Nov 11, 2008]
[18] R. Aleksiejunas, J. Aleksandravicius, J. Krivochiza, and K. Svirskas. Analysis of radar

interference effects on MIMO LTE downlink performance. // Submitted to EMC
Europe 2014.
[19] Z. Zhuang, T.-Y. Chang, R. Sivakumar, and A. Velayutham. "A3: Application-Aware

Acceleration for Wireless Data Networks" . In MOBICOM, 2006.
[20] G. Su, Research on 4G mobile Communications Network Applications of TD-LTE

standard, Information and Communications, vol.7, pp. 221-222, 2013.
[21] IEEE, IEEE Standard for local and metropolitan area networks / Part 16: Air interface
for broadband wireless access Systems, New York, 2009.
[22] Baber Aslam, IEEE 802.11 wireless network simulator using Verilog, 11th WSEAS
International Conference on communications, AgiosNikolaos, Crete Island, Greece,
July 26-28, 2007.
[23] E. Kaerginskis, L. Narbutait, Mathematical model based software design for

simulation of 4G wireless protocols" , INTERNATIONAL JOURNAL OF
COMPUTERS AND COMMUNICATIONS, Issue 2, Volume 7, 2013
245
(ICoIT'17)
10thof April, 2017.
[24] J. Huang, Q. Xu, B. Tiwana, Z. M. Mao, M. Zhang, and P. Bahl. Anatomizing

Application Performance Differences on Smartphones. In MobiSys, 2010.
[25] Sanjit K Mitra, "Digital Signal Processing Applications", Wiley, 2010.
[26] Sedaghat, M., R. Muller, and G. Fisher, "A novel single-RF transmitter for massive
MIMO," Proc. ITG Workshop on Smart Antennas, 1-8, Mar. 2014
246
(ICoIT'17)
10thof April, 2017.
Effect of Doppler Shift frequency on the performance of

2x2 OSTBC-OFDM System
Diana Hayder Hussein
Information System Engineering Department, Technical Engineering College, Polytechnic
University, Kurdistan-Iraq
Diana.Hayder@yahoo.com
Thuraya Mahmood Qaradaghi
Electrical Engineering Department, College of Engineering, Salahaddin University, Kurdistan-Iraq
Thuraya.Alqaradaghi@su.edu.krd

In this paper the effect of the Doppler shift frequency on the
Article History: performance of the 2x2 MIMO system has been studied and
Received: 12 March 2017 presented. A system with two transmits antennas and two
receive antennas has been used. Accurate and efficient
channel estimation which plays a key role in MIMO-OFDM
Published: 10 April 2017 communication system has been implemented using pilots
and training sequences. To guarantee a much more reliable
and robust transmission over a hostile wireless channel, a
Keywords: MIMO, convolutional codes have been added to the 2x2 MIMO-
Doppler shift, OFDM, OSTBC-OFDM system as a Forward Error Correction (FEC)
OSTBC, Rayleigh
codes. Two types of convolutional code rate have been used
Fading Channel,
(1/2 and 2/3) depending on the type of modulation. To make
Convolutional code
this code more robustness against channel impairments;
interleaving has been used with the convolutional code. The
system has been evaluated for four types of QAM modulation
(4-QAM, 8-QAM, 16-QAM, and 64-QAM) as a baseband
modulation. The system has been evaluated for three different
values of Doppler Shift frequency (5, 50, and 100) Hz. The
simulation results demonstrated the performance of the
systems over Flat and multi-path Frequency-Selective fading
channels, assuming the channels are with No Line Of Sight
(NLOS), so the channels are Rayleigh fading channels. The
results show that when the Doppler Shift Frequency increased
the BER also increases and when Doppler Shift Frequency is
5Hz, the performance of the system is better for four types of
QAM modulation and in the two cases of Flat and multi-path
Frequency-Selective fading.
247
(ICoIT'17)
10thof April, 2017.
1. INTRODUCTION
he Future wireless mobile systems will aim to support high quality of services
T
and high data rates by employing techniques that can enhance channel capacity
[1]. The transmitted data is interfered by channel noise, Co-Channel Interference
(CCI) when transmitted, and these noises and CCI affect randomly and suddenly
on all transmission bits. Diversity techniques, including spatial, frequency, and
time domain diversity, have been suggested to decrease the channel fading
effect. There are many techniques and algorithms to mitigate the CCI effect also. Sufficiently
spaced antennas are an attractive source of diversity since they do not typically incur in
bandwidth expansion as in frequency division diversity, and does not incur delays as in time
diversity. Though spatial diversity is available at transmitter and receiver, it may not be
possible to get much diversity gain at mobile terminal because of the limitations in space and
power [2]. Space Time Coding (STC) systems process transmit and receive signal waveforms
in temporal, spatial and coding dimensions to deliver high data rates with diversity and
coding gains, and MIMO technology has attracted attention in wireless communications,
because it offers significant increases in data throughput and link range without additional
bandwidth or increasing transmitted power. It achieves this goal by spreading the same total
transmit power over the antennas to achieve an array gain that improves the spectral
efficiency (more bits per second per hertz of bandwidth) and/or to achieve a diversity gain
that improves the link reliability (reduced fading). Because of these properties, MIMO is an
important part of modern wireless communication standards such as IEEE 802.11n (Wi-Fi),
4G, WiMAX etc. [3].
Orthogonal Frequency Division Multiplexing (OFDM) is the vehicle that drives modern
digital communications. Its high spectral efficiency and resilience toward multipath
distortion has made possible the increased data rates required in many devices. It is used in
many systems including DSL, IEEE802.11a, g, n (Wi-Fi), digital radio, digital TV,
IEEE802.16 (WiMAX), and 3G, 4G, and Long Term Evaluation (LTE) cell services [4].
MIMO with OFDM reduces the equalization complexities by transmitting different data on
different frequency levels to gain spectral efficiency and error recovery features, which will
offer high spatial rate by transmitting data on multiple antennas and transmission in Non-
Line-of sight (NLOS). Thus the MIMO-OFDM technique is used to achieve diversity. It
utilizes the three basic parameters that is frequency (OFDM), time (STC) and MIMO in
spatial. The MIMO-OFDM is the reproductive and highly famous services for wireless broad
band access. The combination of MIMO and OFDM accumulates the purpose of each and
every scheme that provides the high throughput [5].
2. Space Time Coding (STC)

Spacetime coding is a coding technique used in wireless communications to transmit
multiple copies of a data stream (generally Quadrature Amplitude Modulation (QAM)
symbols) across a number of antennas to exploit the various received versions of the data to
improve the reliability of the communication. In fact, STC combines all the copies of the
received signal in an optimal way to extract as much information from each of them as
possible. The spacetime block codes (STBCs) achieve significant error rate improvements
over single-antenna (SISO) systems [6]. The STC technique is essentially a two-dimensional
space and time processing method. While multiple antennas both for transmission and
248
(ICoIT'17)
10thof April, 2017.
reception are used to improve wireless communication systems capacity and data rate in
space-domain. In time-domain, different signals can be transmitted at different time slots
using the same antenna at the same time [7].
2.1 Alamouti code
The Alamouti space time coding scheme can be used to achieve diversity at the transmitter
and receiver if number of antennas are involved at the transmitter and receiver. The
approach as outlined by Alamouti is shown in Figure 1. The information bits are first
modulated using an M-ary modulation scheme. The encoder then takes a block of two
modulated symbols S1 and S2 in each encoding operation and gives it to the transmit
antennas according to the code matrix,
S= (1)
In equation (1), the first row represents the first transmission period and the second row
the second transmission period.
Figure 1: Alamouti Space-Time Encoder block diagram.

During the first symbol period, the first antenna transmits S1 and the second antenna
transmits S2. During the second symbol period, the first antenna transmits S2*and the
second antenna transmits S1*.
At the receiver the signals after passing through the channel can be expressed as,
= + + (2)
=- + + (3)
Where, n1, n2 are independent complex variables with zero mean and unit variance,
representing Additive White Gaussian Noise (AWGN) samples at time t = 1 and t = 2,
respectively, h1 is the path gain between the first transmitted antenna and the received
antenna h2 is the path gain between the second transmitting antenna and the received
antenna and r1, r2 are the received signals at the two time slots [8]. And the estimated
symbols are as shown below:
249
(ICoIT'17)
10thof April, 2017.
= + (4)
= - (5)
3. Convolutional Code
Convolutional codes are linear codes, many telecommunications applications have used
convolutional codes because of their ability to deliver good coding gains on the AWGN
channel for target bit error rates around 105 [9]. So it is a powerful and widely used class of
codes, which are used in a variety of systems including todays popular wireless standards
(such as 802.11) and in satellite communications.
Convolutional codes are often preferred in practice over block codes, because they provide
excellent performance when compared with block codes of comparable encode/decode
complexity. Whereas block codes take discrete blocks of K symbols and produce there from
blocks of N symbols that depend only on the k input symbols, convolutional codes are
frequently viewed as stream codes, in that they often operate on continuous streams of
symbols not partitioned into discrete message blocks [10].
When the encoded information is transmitted over the channel, it is distorted; the
convolutional decoder regenerated the information by estimating the most likely path of state
transition in the trellis. The receiver, of course, does not have direct knowledge of the
transmitters state transitions. It only sees the received sequence of parity bits, with possible
corruptions. Its task is to determine the best possible sequence of transmitter states that could
have produced the parity bit sequence. This task is called decoding, a decoder that is able to
infer the most likely sequence is also called a maximum likelihood decoder. The Viterbi
decoder finds a maximum likelihood path through the Trellis [11]. In this paper two types of
convolution code rate have been used depending on the type of modulation as in the Table
1.
Table1: Convolution code rates for different types of modulation
Modulation Convolutional Code

rate (Rc)
4-QAM 1/2
8-QAM 2/3
16-QAM 1/2
64-QAM 1/2
4. Interleaving
Interleaving plays a vital role in improving the performance of Forward Correction
(FEC) codes in terms of Bit Error Rate (BER). Interleaving is the process to rearrange code
symbols so as to spread burst of errors into random like errors and thereafter FEC techniques
could be applied to correct them. In conventional block interleaver, the bits received from
the encoder are stored row wise in the interleavers memory and read column wise as shown
in Figure 2, WiMAX uses a special type of block interleaver in which the Interleaver Depth
(ID) and pattern vary depending on the code rate and modulation type [12].
250
(ICoIT'17)
10thof April, 2017.
Figure 2: The operation of Interleaver

The Interleaver dimension has been choosing with the following parameters as in WiMAX
system:
= 12 (6)
= (7)
5. Doppler Shift
The Doppler Effect is the change in frequency of a wave that is perceived by an observer
moving towards or away from the source of the waves. It is well-known that Doppler effects
generated by high speed mobility are the major reason for the reduction of data rates in
cellular systems. The Doppler Effect may occur from either motion of the source or motion
of the receiver as in Figure 3.
Figure 3: a) Stationary source b) Moving source
251
(ICoIT'17)
10thof April, 2017.
It is important to comprehend that the frequency of the signal that the source emits does not
actually change, but the wavelength () does; consequently, the perceived frequency is also
affected. When the receiving end moves towards the base station the receiving frequency
becomes higher and when is receding the receiving frequency becomes lower (see Figure
(3)). The Doppler shift in frequency depends on the velocity between the source and the
receiver and on the speed of propagation of the signal. Doppler frequency is given by the
formula:
f * (8)
Where f is the change in frequency, is carrier frequency, v is the speed difference

between the source and the receiver, and c is the speed of light in vacuum which is 3* m/s
[13]. For wireless communication, when electromagnetic wave is traveling towards or away
from the receiver, the carrier frequency will be shifted, causing Doppler shift. It is noticed
that Doppler shift is usually prominent when the transmit antenna is far from receive antenna
[14].
6. System model
The block diagram of the wireless communication system with two transmit antennas
and two receive antennas is shown in Figure 4. As it can be seen from this Figure, the
generated random binary data is first encoded using binary convolutional encoder, and then
these encoded data are further interleaved to avoid error bursts and increase the efficiency of
FEC. M-QAM have been used for mapping the coded and interleaved data (bits) into
constellation symbols. The mapped data is inserted to the IFFT Input Packing block, in this
block, 8-Pilot symbols are added to the mapped data, these pilots can be used to perform
frequency offset compensation at the receiver, additionally they can be used for channel
estimation in fast time-varying channels. After IFFT Input Packing block, the mapped data
is passed to STBC encoder to be modulated using OFDM modulation, and then transmitted
by two transmitting antennas. The data passed through the channel which is affected by the
presence of Additive White Gaussian Noise (AWGN) and Rayleigh fading. At the receiver
side, the data is received by two receiving antennas, the data is demodulated using OFDM
demodulator and linear combining is performed in STBC decoder, and then demapped by
QAM demapper. The Deinterleaver rearranges the symbols and returns it back to its original
form as before the interleaving process, finally Viterbi decoder uses decoding algorithm to
get the transmitted data.
252
(ICoIT'17)
10thof April, 2017.
Figure 4: Block diagram of 2x2 MIMO-OSTBC-OFDM system

7. Results and Discussion
The system has been simulated using MATLAB. The performances of the proposed system
has been introduced depending on Bit error rate (BER) versus signal to noise (SNR) ratio
plots. The system has been evaluated for four types of QAM (4,8,16,and 64-QAM) in two
cases of channels, Flat-Fading and Frequency-Selective Fading using Convolutional Code
and interleaver in the transmitting side, deinterleaver, and Viterbi decoder in the receiving
side. From the figures, to achieve 10-5 BER, the SNR ratio should be approximately 17.5dB
when the Doppler frequency is 100Hz, the SNR should be 16dB when the Doppler frequency
is 50 Hz, and it should be 15dB when the Doppler frequency is 5 Hz.
253
(ICoIT'17)
10thof April, 2017.
0
2x2 MIMO-OSTBC-OFDM system,with Convolutional Code, MDS=5Hz
10
4-QAM,Flat-fading
8-QAM,Flat-fading
16-QAM,Flat-Fading
-1
10 64-QAM,Flat-Fading
4-QAM,Freq.-Selective
-2 16-QAM,Freq.-Selective
10
BER
-3
10
-4
10
-5
10
0 5 10 15 20 25 30
SNR(dB)
Figure 5: Performance of the coded 2x2 MIMO-OSTBC-OFDM system and

MDS=5Hz
0
2x2 MIMO-OSTBC-OFDM system,with Convolutional Code, MDS=50Hz
10
4-QAM,Flat-fading
8-QAM,Flat-Fading
16-QAM,Flat-Fading
-1
8-QAM,Freq.Selective
10
BER
-3
10
-4
10
-5
10
0 5 10 15 20 25 30
SNR(dB)

MDS=50Hz
254
(ICoIT'17)
10thof April, 2017.
0
2x2 MIMO-OSTBC-OFDM system, with Convolutional Code, MDS=100Hz
10
4-QAM,Flat-Fading
8-QAM,Flat-Fading
16-QAM,Flat-Fading
-1
10
BER
-3
10
-4
10
-5
10
0 5 10 15 20 25 30
SNR(dB)

MDS=100Hz
8. Conclusion
The Figures (5-7) which represent the performance of 2x2 system with FEC show better
performance when MDS=5Hz than the other two values 50, and 100Hz. For example by
comparing Figure 5 to Figure 7, the value of SNR in Figure 5 is 15dB for 64-QAM
Frequency-Selective Fading, but in Figure 7, the value of SNR=17.5dB for 64-QAM
Frequency-Selective Fading, so it can be concluded that increasing the Doppler frequency
by ten times (from 5 Hz to 50 Hz) just extra 1dB is needed to achieve same SNR, and when
increasing the Doppler frequency to 100 Hz extra 2.5dB is needed.
9. Reference:
[1] F.Delestre and Y.Sun, Pilot Aided Channel Estimation for MIMO-OFDM Systems,
University of Hertfordshire, London Communications Symposium 2009.
[2] Subhankar Chakrabarti, Study The Effect of Co-Channel Interference in STC MIMO-
OFDM System and Mitigation of CCI Using Beamforming Technique, MSc thesis,
National Institute of Technology, Rourkela, 2011-2013.
[3] Prof. Shivendra Singh, Prof. Vikas Gupta, and Shiv kumar Yadav, Performance Analysis
of MIMO-OFDM (802.11n System) for WLAN Channel Model, International Journal
of Engineering Research and Applications (IJERA) Vol. 3, Issue 3, pp. 946-949, May-
Jun 2013.
[4] Steven Charles Hemple, Analysis and Simulation of Wireless OFDM
255
(ICoIT'17)
10thof April, 2017.
Communications, MSc thesis, San Diego State University, summer 2012.

[5] Ruchin Mangla, and Maninder Singh, Performance Comparison of MIMO-OFDM
Transceiver Wireless Communication System Using QAM and QPSK Modulation
Schemes, Haryana, India, International Journal of Advances in Engineering Science and
Technology, Vol.1, No.2, pp. 66-72, December 2012.
[6] Dinesh Divakaran, OWSS and MIMO-STC-OFDM: Signaling Systems for The Next
Generation of High Speed Wireless LANs, PhD.thesis, Department of Electrical
Engineering College of Engineering University of South Florida, 2008.
[7] Wang Yi Gu linfeng, An Investigation of Peak-to-Average Power Reduction in MIMO-
OFDM Systems, MSc thesis, Blekinge Institute of Technology, October 2009.
[8] Abhishek Sharma, and Anil Garg, BER Analysis Based on Transmit and Receive
Diversity Techniques in MIMO-OFDM System, International Journal of Electronics &
Communication Technology, Haryana, India, Vol. 3, Issue 1, pp.58-60, January - March
2012.
[9] Peter Sweeney, Error Control Coding from Theory to Practice, John Wiley and Sons
Inc., 2002
[10] Todd K. Moon, Error Correction Coding, John Wiley and Sons Inc., 2005.
[11] MIT 6.02 Draft Lecture Notes, Convolutional Coding, October 4, 2010.
[12] Bijoy Kumar Upadhyaya, and Salil Kumar Sanyal, Novel design of WiMAX multimode
interleaver for efficient FPGA implementation using finite state machine based address
generator, International Journal of Communications, Vol. 6, pp. 27-36, 2012.
[13] Ioannis D.Erotokritou, " Space-Time Block Coding for Multiple Transmit Antennas
over Time-Selective Fading Channels", Louisiana State University , May 2006.
[14] Jing Xiao,"Analysis of OFDM MIMO", San Diego StatebUniversity, Summer 2010.
256
(ICoIT'17)
10thof April, 2017.
Energy-Efficient Sensor Activity Scheduling Protocol for

Wireless Sensor Networks
Wesam Huseein
Dept. of Information Networks, College of Information Technology, University of Babylon - Iraq
wesamhusain4@gmail.com
Ali Kadhum Idrees
Dept. of Computer Science, College of Science for Women, University of Babylon - Iraq
ali.idrees@uobabylon.edu.iq

The power depletion is one of the critical issues in Wireless
Article History: Sensor Networks (WSNs) since most sensor nodes are
Received: 19 March 2017 provided with limited power irreplaceable / rechargeable
batteries. One fundamental way to minimize the power
consumption is to exploit the spatial data correlation among
Published: 10 April 2017 the sensor nodes by reducing the number of active nodes
especially in the dense sensor network. In this paper, we
Keywords: wireless propose Energy-efficient Sensor Activity Scheduling (ESAS)
sensor networks, protocol for WSNs that periodically choose the best
scheduling, optimization, representative set of sensors that will be activated during the
power saving , special sensing phase in the current period. ESAS works into periods.
correlation. Each period is composed of three phases: data collection,
decision, and sensing. ESAS measures the similarity degree
among the sensor data that collected in the first phase and
makes a decision of which sensors stay active during the
sensing phase in each period and put the other nodes into low-
power sleep mode with maximum level of accuracy. The
simulation results are conducted based on real world datasets
provided by Intel Berkeley Research Lab using OMNeT++
network simulator show that the effectiveness of the proposed
ESAS protocol in energy saving and extending the network
lifetime.
1. INTRODUCTION
W
ireless Sensor Network (WSN) composed of hundreds or thousands of nodes,
which deployed in high density manner in the area of interest for different
purposes such as battle field surveillance, disaster observation, smart homes and
many other applications [1]. Sensor nodes data reporting can be divided into two general
categories event-based data reporting and continuous data reporting. In event-based data
reporting, the sensor nodes send the data to the sink when an event triggered. On the other
257
(ICoIT'17)
10thof April, 2017.
hand, in continuous data reporting the sensor nodes continuously send their readings to the
sink such as humidity and temperature monitoring [2].
The high density deployment of sensor nodes in area of interest causes the data
redundancy at the sink node and leads to an unnecessary power consumption since some
nodes remain active whilst they sense redundant data. Therefore, we need a protocol for
scheduling the sensor activity by making sleep/active schedule for the sensor nodes that
minimize the data redundancy at the sink and eventually save the power since the power is
the most important constraint in the sensor nodes that must be considered when design any
protocol for the WSN.
The main contributions in this paper are as follow.
i. We propose a protocol, called the Energy-efficient Sensor Activity Scheduling
(ESAS) protocol, which exploits the correlation among the sensor nodes for
scheduling the sensors to minimize the data redundancy at the sink and maintain
the data accuracy and improve the network lifetime. ESAS works into periods.
Every period is composed of three stages: data collection, decision and sensing.
In data collection stage, all the nodes will be activated and send the data to the
cluster head to find the correlation among the sensors reading. The decision stage
executes a memetic algorithm at the sink node (or cluster head) to choose the best
set of active sensors for the current period and deactivate the other nodes. The
active nodes in the sensing stage continue transmitting the sensed data to the sink
node until the end of the current period.
ii. A modified optimization model that uses the data correlation among the sensor
nodes so as to decide which sensors stay active and which sensors will turn off.
iii. Extensive simulation experiments using OMNeT++ simulator are conducted to
demonstrate the efficiency of ESAS protocol. . Simulation results based on
multiple criteria (Number of active sensors, energy consumption, data loss ratio,
transmitted data ratio) illustrate that the proposed ESAS protocol can minimize
the energy consumption and transmitted data with suitable data loss ratio.
The reset of the paper is organized as the follow. Section 2 presents the related works on
various scheduling protocols in WSNs. Section 3 describes the proposed ESAS protocol. The
results are presented in Section 4. Section 5 demonstrates the conclusion and future works.
2. RELATED WORK
Energy-efficiency is one of the important requirements in WSN protocol design since
most of WSN nodes are battery powered. Several scheduling approaches are proposed in the
literature. The authors in [3, 4, 5] proposed scheduling algorithm for maintaining the
coverage and improving the network lifetime of WSNs. They proposed optimization models
based on both primary points and perimeter coverage model to optimize both the coverage
and the lifetime of WSNs. Some other existing approaches exploit the spatial correlation
among the sensor data in WSN for making sleep/active schedule in order to increase the
network lifetime [6, 7, 8].
The authors in [6] proposed a method in which the base station divides the sensors into n
sets by using closeness factor. This method supposes the close nodes can sense the same
data. Depending on shift base while the previous set of nodes exhaust its battery the other set
258
(ICoIT'17)
10thof April, 2017.
of nodes will be activated. The work in [7] is presented structure fidelity data collection
(SFDC) framework which consists of two phases: learning and data collection. In the
learning phase, the framework utilizes the spatial correlation of the collected data. It
introduces cluster formation approach that includes cluster head selection and active node
selection in each cluster based on similarity index in sensor readings. In data collection phase,
all the active nodes continue to send the data to the cluster head. Three algorithms are
proposed in [8] for active node selection and less energy consumption. The first algorithm is
Cover Sets Balance algorithm (CSB) to select a set of active nodes that have wide data
coverage ranges and high energy levels. It depends on the tuple (data coverage range, residual
energy) to find an initial active node set and then balance the size of the cover sets in order
to replace low-energy nodes. The second algorithm is Correlated Node Set Computing
algorithm (CNSC) to calculate the correlated node set with minimum set size and the
maximum geometric mean of residual energy of each node in the sensor network to use the
correlation results in the next algorithm. The third algorithm is High Residual Energy First
algorithm (HREF) to further reduce the number of active nodes selected by CSB algorithm.
Xu et al. [9] suggested a data-coverage method in which the active sensor can recover the
missing data of sleeping nodes. This method is based on the data correlation between the
sensor nodes. A Greedy Reduction Algorithm (GRA) to select the set of active sensors is
proposed. The missed data of sleeping nodes can be returned using the sensed data of active
nodes as well as the correlation between the sensor nodes.
3. THE PROPOSED ESAS protocol

In this section, we describe the proposed ESAS protocol for minimizing the energy
consumption in wireless sensor network, and decreasing the data redundancy at the sink
node.
Figure 1. Proposed ESAS Protocol
ESAS protocol mainly divides the network lifetime into P periods. In each period, there
are three stages data collection, decision, and sensing as described in the Figure 1. In data
collection phase all the sensor nodes wakeup and send their sensed data to the sink for making
a decision using a metaheuristic optimization algorithm to provide the best representative set
of sensor nodes which are responsible for monitoring during the sensing stage. In the sensing
stage the active sensor nodes are responsible for sensing and sensing the sensed data to the
sink. ESAS protocol uses two types of packets: DataPacket for sending the sensed data of
each sensor node to the sink, and the StatusPacket used by the sink after achieving the
optimization for informing the all the sensor nodes with their new status (active or sleep).
259
(ICoIT'17)
10thof April, 2017.
The DataPacket is 328 bits length (8 bits for header and 320 bits for 10 data reading). The
StatusPacket is 16 bits length (8 bits for header and 8 bits for the status of the sensor, 0 is
sleep, 1 is active).
In this paper, the main objective of ESAS protocol is to minimize the active sensor nodes
and maximize the network lifetime with maintaining an acceptable level of data accuracy.
Our mathematical optimization model is inspired from [10] with some modifications. Let S
= {s1, s2, , si, , sN} the set of the sensor nodes in the network of size N. Let Xij represents
a binary indicator for the correlation between the sensor i and the sensor j. In the decision
phase, we exploit the correlation matrix that extracted at the first stage to decide which
sensor node will remain active in the next stage of the period (sensing) and which node go
to the sleep mode until the next period for saving the power.
1 If the sensed data of the sensor i correlated with sensed data of the sensor j or
i=j
= (1)
0 Otherwise.
The data correlation between two sensor nodes is calculated by applying the Jaccard
similarity between their two data sets A and B as follow
|| ||
(, ) = || = ||+ || || (2)
The data are sorted in A and B and a difference function is used to compare the similarity
between each opposite elements inside A and B. The Difference function can be defined as
in (3), where is a threshold fixed by the application.
1 if | |
( , ) = (3)
0 Otherwise.
1 If the sensor i is active

= (4)
0 Otherwise.
(
=1 ) 1 The number of correlations with other sensors
= (5)
1 Otherwise.
The optimization problem can be formulated as follow.
Minimize =1 (6)
Subject to
=1 = 1 (7)
We propose to use a metahuristic search algorithm for solving our optimization problem.
A memetic algorithm is suggested to find the best sensor set S after applying the optimization
model. The proposed memetic algorithm is a combination between a differential evolution
with local search (Hill climbing) to recombine the results and preventing the local optima.
260
(ICoIT'17)
10thof April, 2017.
Figure 2 shows the pseudo code of the proposed activity scheduling based memetic
algorithm.
Figure 2. Proposed activity scheduling based memetic algorithm.
At the end of the decision stage, the sink node transmits a StatusPacket for every sensor
node based on the values of vector S that provided by the optimization algorithm based on
the memetic approach. The active nodes stay sending the data to the sink until the end of the
current period.
4. RESULTS
In this section, we conducted several experiments based on real sensor data from [11]
and using OMNeT++ network simulator [12]. In order to evaluate our ESAS protocol, some
performance metrics are used during this simulation such as Average of active sensor nodes,
Data loss ratio, Transmitted data ratio, and energy consumption. ESAS protocol applied the
energy consumption model that proposed by Heinzelman [13].
261
(ICoIT'17)
10thof April, 2017.
4.1 Average of Active Sensor Nodes

In this experiment, the average number of active sensor nodes versus is studied.
Figure 3 shows the average number of active nodes for the deployed nodes in Intel Berkeley
Research Lab. As shown in Figure 3, when the value of the threshold of the Difference
function increases, the number of active sensor nodes decreases. This is due to the high
correlation between the data sets of the sensor nodes when the value of increased.
Figure 3. Average number of active nodes for the deployed nodes in Intel Berkeley Research
Lab..
4.2 Transmitted Data Ratio to Sink

In this section, the percentage of the transmitted data to the sink node is investigated.
Figure 4 shows the transmitted data ratio to the sink node. The results show that the sink
node receives as less number of data as possible when the threshold increases. This is due
to the small number of active nodes provided by the ESAS protocol after performing the
memetic optimization algorithm based on the correlated data of the sensor nodes.
262
(ICoIT'17)
10thof April, 2017.
Figure 4. Transmitted data ratio to the sink node.
4.3 Data Loss Ratio

This experiment studies data loss ratio after applying the scheduling algorithm in the
network. Figure 5. Shows the data loss ratio after receiving the data at the sink node. As
shown in this figure, ESAS protocol loses a lot of data when the threshold increases because
it activates a smaller number of sensor nodes due to the high correlation between the sensor
nodes.
Figure 5. Data loss ratio.
263
(ICoIT'17)
10thof April, 2017.
4.4. Energy Consumption

In this study, the performance of the ESAS protocol from the energy consumption
point of view is investigated. Figure 5 illustrates the consumed energy by the sensor nodes
during applying our ESAS protocol. The results show that ESAS protocol can consume less
energy when the value of increases due to increase the correlation between the sensor nodes
and thus decreases the number of active sensor nodes.
Figure 5. Energy consumption.
5. CONCLUSIONS AND FUTURE WORKS

One of the essential problems in WSNs is to provide a best sensor activity schedule that
save the energy of the sensor node while ensuring a suitable level of accuracy for the received
data at the sink. One solution is to exploit the data correlation between the sensor nodes so
as to decrease the energy consumption in the WSN. In this paper, we propose Energy-
efficient Sensor Activity Scheduling (ESAS) protocol for WSNs. It is periodic and it chooses
the best representative set of sensor nodes which are activated during the sensing phase in
the period. Each period in ESAS is composed of three phases: data collection, decision, and
sensing. Experimental results demonstrate the effectiveness of ESAS protocol in terms
energy efficiency, data loss ratio, number of active nodes, and the transmitted data ratio to
the sink.
In the future, we will include the energy inside the optimization model in order to prevent
selecting the low power nodes and to improve the balancing in selecting the sensor nodes in
each period. In addition, we plan to compare our protocol with some recent existing method
to show the superiority of our protocol.
6. REFERENCES
[1] I.F. Akyildiz, W. Su, Y. Sankarasubramaniam, E. Cayirci, Wireless Sensor Networks:
A Survey, Computer Networks, pp. 393-422, Aug.2002.
264
(ICoIT'17)
10thof April, 2017.
[2] Makhoul, A., Laiymani, D., Harb, H. and Bahi, J.M. An adaptive scheme for data
collection and aggregation in periodic sensor networks. International journal of sensor
networks, 18(1-2), pp.62-74, 2015.
[3] Idrees AK, Deschinkel K, Salomon M, Couturier R. Coverage and lifetime optimization
in heterogeneous energy wireless sensor networks. ICN 2014. 2014 Feb 23:60.
[4]Idrees AK, Deschinkel K, Salomon M, Couturier R. Distributed lifetime coverage
optimization protocol in wireless sensor networks. The Journal of Supercomputing. 2015
Dec 1;71(12):4578-93.
[5]Idrees AK, Deschinkel K, Salomon M, Couturier R. Perimeter-based coverage
optimization to improve lifetime in wireless sensor networks. Engineering Optimization.
2016 Nov 1;48(11):1951-72.
[6] Mrs. Surekha K.B., Mr. Raghunandan V, Dr. Mohan K.G., Efficient Sleep Scheduling
Strategy to enhance the Network Lifetime of WSN International Journal of Engineering
Research,2015.
[7] Mou Wu 1, Liansheng Tan 1 and Naixue Xiong, A Structure Fidelity Approach for Big
Data Collection in Wireless Sensor Networks, Sensors 2015.
[8] Hongju Cheng, Zhihuang Su,Daqiang Zhang,Jaime Lloret,and Zhiyong Yu ,Energy-
Efficient Node Selection Algorithms with Correlation Optimization in Wireless Sensor
Networks, Hindawi Publishing Corporation International Journal of Distributed Sensor
Networks ,2014.
[9] Xu X, Hu YH, Liu W, Bi J. Data-coverage sleep scheduling in wireless sensor networks.
InGrid and Cooperative Computing, 2008. GCC'08. Seventh International Conference on
2008 Oct 24 (pp. 342-348). IEEE.
[10] F. Pedraza, A. L. Medaglia, and A. Garcia. Efficient coverage algorithms for wireless
sensor networks. In Proceedings of the 2006 Systems and Information Engineering Design
Symposium, pages 7883, 2006.
[11] Intel Lab Data. Available: http://db.csail.mit.edu/labdata/labdata.html.
[12] A. Varga. Omnet++ discrete event simulation system. Available:
http://www.omnetpp.org, 2003.
[13] Heinzelman, W.B., Chandrakasan, A.P. and Balakrishnan, H., 2002. An application-
specific protocol architecture for wireless microsensor networks. IEEE Transactions on
wireless communications, 1(4), pp.660-670.
265
(ICoIT'17)
10thof April, 2017.
E-Service Quality Factors Impacting Customers

Purchase Retention in E-Retailing in Malaysia
Azwar Mehdi Qassim Agha
Department of Business Administration, Collage of Administrative and Economics, Lebanese
French University, Erbil - Kurdistan Region /Iraq
Azwar.mehdi@lfu.edu.krd

Many organizations have faced problems and challenges on
Article History: furnishing online service quality to satisfy the customers. In
Received: 20 March 2017
Accepted: 1 April 2017 Malaysia e-retailing is still a new business area. All
Published: 10 April 2017 Malaysians are not exposed to the online transaction over the
internet and most often they are still hesitate to make online
Keywords: E-service purchasing due to reasons of security, reliability and threat of
quality, relationship
fraud on transactions. Education of internet use for reliable
quality, online purchase
decision, Customer information search and internet purchasing is also one of the
retention, e-retailing. main factors required to go for e-retailing. Without this
technical literacy it would be very difficult for the consumers
to go for online purchasing. This study designed and collected
data from 148 customers who purchase online.
Findings of this study indicate that Information Quality,
Internet Security, Service reliability are the key
determinants of E-Service Quality among online buyers in
Malaysia.
The implication for the owner-managers of the businesses in
Malaysia is that they must be aware of the significance of the
key E-Services Quality indicators while designing their
businesses, to attract and retain customers.
1. INTRODUCTION
T
he growth of e-commerce in the global and emerging markets has created an
important and special interest on the measurement of the e-service quality and
also on the dimension of the e-services that influences on the customers
satisfaction and their experience on the shopping through e-commerce or e-
retailing (J. Yang, G. Miao, 2005).
With the rapid development of information and communication technology, Internet
and World Wide Web have become important tools in business which has a major
impact on business world. (Li, Hongxiu, Liu, Yong, Suomi, Reima, 2009) With the
increase of e-service adoption in business field, the importance of measuring and
monitoring e-service quality in the virtual world has been recognized, and e-service
266
(ICoIT'17)
10thof April, 2017.
quality posits to be a hot topic in research field. Some academic researchers have
already been conducted to develop e-service quality measurement. From the existing
literature, it is evident that these studies have been conducted mainly in the domains
of online retailing service, online shopping website quality, and e-service quality.
There has been limited attention on online service sector.
The activities of online retailers are focusing on the measurement and the
improvement of e-service quality that is provided to their potential customers. E-
service quality can be defined as the extent to which the web site promotes efficient
and effective shopping, purchasing, and delivering of products and services. The e-
commerce refers to generally all forms of transactions relating to commercial
activities including both organizations and individuals that are based on the
processing and transmission of the digitalized data. Since it was found by many
researchers that service quality is the main aspects of the e-commerce, there have
been many investigations done on how to improve the online services to attract the
online customers and also how to retain current customers.
This study makes an attempt to investigate the factors of the e-service quality which
have a drastic effect on the customers retention to go for e-retailing.
Web service or online service is related to various attributes of an online retailing
web site. According to (Renata Mekovec, 2007) the activities of online retailers are
focusing on the measurement and the improvement of e-service quality that is
provided to their potential customers. E-service quality can be defined as the extent
to which the web site promotes efficient and effective shopping, purchasing, and
delivering of products and services. The quality of service is one of the essential
elements of competitive advantage of online retailers whose objectives are to convert
web site visitors into customers and also to retain the customers in the competitive
market where other vendors are just a click away and are offering similar products at
a comparable price (S. Ba, J. Stallaert, Z. Zhang, 2005).
Since the Internet was discovered, Web sites have become an important channel for
companies to provide services to the customers. Businesses around the world now
operate more than 400,000 publicly accessible Web sites. In recent years, the
applications of information technology have provided a powerful platform for e-
service. One noticeable trend of e-service is the increasing use of self-service
technologies. According to (Rust and Kannan, 2003), the implementation of SSTs in
e-service extends more choices, more options, and, ultimately, more power to
customers in their transactions with businesses.
In Malaysia, internet business is not so advanced and is a new business area. It is
because not all the Malaysians are exposed to the online transaction through internet
267
(ICoIT'17)
10thof April, 2017.
and also they hesitate to make online purchasing due to various problems like
security, reliability, threats of frauds etc. (Noor HabibahArshad et al, 2008).
Furthermore, research on the determinants of service quality can be evaluated with
the help of two streams i.e. The first concentrated on the quality of the Web sites
interface design and hardware performance which was not able to fully capture the
service quality of a technical support Web site due to the neglecting of the quality of
the Web content for customers. And the second focused on service quality during the
order fulfillment process of transaction-based Web sites. (Po-Young, Chia-Yi, Yu-
Ling and Wei-Chen, 2009).
The growth of e-commerce in the global and regional markets creates a special
interest in the measuring of e-service quality and also in the investigation of the
dimensions of e-service as one of the important factors that influence customer
satisfaction and shopping experience of the customers of e-commerce web sites.
E-service quality covers various aspects of the e-commerce web site and diverse
elements of customer's experience. It includes all phases of the customers'
interactions with a web site and refers to the extent in which the e-commerce oriented
web site supports shopping behavior, as well as the processes of buying and delivery.
The rapid increase in the number of e-commerce web sites initiated research studies
that were focused on the measurement and evaluation of online service quality (A.
Parasuraman, V.A. Zeithaml, A. Malhotra, 2005).
2.1. E-retailing
E-retailers are major players in the field of electronic commerce and their success
would seem to depend on service quality, because they are selling the same products
that traditional retailers sell. E-retailing is the largest marketing activity in the rapidly
growing field of electronic commerce and, logically, perceived service quality would
seem to be the key success factor that lifts this new form of retailing above traditional
retailing because the products the two types of retailers sell are the same. However,
e-retailing service quality must be validly measured if its actual role is to be
demonstrated empirically. Moreover, a valid measure of service quality at each stage
of e-retailing is required if e-retailers seek to improve their service quality. (John R.
Rossiter, 2007).
Successful Web-based information systems (WIS) are critical for electronic retailers
to attract and retain consumers and deliver business functions and strategy. Previous
researches show that the gap in Internet usage has been declining between men and
women, and among different age groups. However, online customers still represent a
distinct group with higher income and education. In making their decisions, they
search the Web for information and explore lower prices and higher value. They have
more access to information and thus possess more market knowledge. They actively
exchange information with other customers. They may evaluate products and
purchase online or offline. Online customers may also have different decision-
268
(ICoIT'17)
10thof April, 2017.
making processes using different evaluation criteria that may rely on electronic tools.
(Fang Wang and Milena M. Head, 2001).
The Internet as a direct selling technique has attracted the attention of established
retailers and potential entrants. Although the concept of direct selling is not a novelty
for the sector, the scope that the Internet offers is much broader than the scope of
traditional direct selling techniques like catalogs and TV shopping. This new
communication and distribution tool has questioned the necessity of traditional
distribution channels in areas like software and music, and it is increasingly becoming
established in markets. In order to understand the behavior of retailers, both
incumbent and new, attention should be paid to the characteristics and the importance
of the intermediation function they perform between manufacturers and consumers.
(Veneta Andonova, 2003).
The internet together with the implementation of innovative business practices plays
a fundamental role in the development of e-commerce. However, at the early stage,
it was not uncommon that many virtual business models cannot achieve expected
outcomes because of the misunderstanding of consumer behavior and the
inconsistency with a particular market and social environment. Some service
providers intensively promoted the benefit of e-commerce and assumed that
individual consumers might be willing to use the services. The retail market in
Malaysia is characterized by the availability of a variety of different goods and
services in the metropolitan area. Many consumers enjoy shopping individually or
with friends and relatives. Individual style, taste and shopping behavior can be easily
influenced by fashion, in-group opinions and shared assumption in such a collective
society. Thus, subjective norm might affect consumer perception and shopping
behavior. The issue of social context being involved in the process of psychological
adoption of technological innovation should not be overlooked in certain
circumstances. (Ziqi Liao and Xinping Shi, 2009).
2.2.1 Information Quality
The Internet is an innovative form of information technology, yet most commercial
Web sites function as well-defined information systems. Information system quality
can be divided into system and information quality. (Zhilin Yang, 2004) System
quality refers to software development caliber, while information quality embraces
accuracy, timeliness, currency and reliability of information. The principal goal of
information systems service is to enable customers to function independently and to
conduct numerous transactions on their own. In addition, as end users, consumers
often seek desired product and service information through Web sites.
Information quality means to the correct information quality in the website. In e-
service, the accurate information quality makes customers always accessible to the
online service offered by online companies, which can help customers to have a good
image of online companies. If customers cannot use the online system when they
269
(ICoIT'17)
10thof April, 2017.
need online service, they will switch to some other online companies. Dissemination
of information not related to price favors a decision based on quality, thereby helping
the consumer to select the product which best suits his desires and needs. Greater
qualitative information transparency may mitigate the importance of price, thereby
affording greater profitability to the electronic retailer. With respect to the emphasis
on price information consumers tend to underestimate the frequency of more regular
promotions and overestimate the frequency of promotions. In other words, those who
offer promotions sporadically are doubly benefited, both by the mistaken perception
of consumers and by the expected higher profit margins obtained over the course of
time. (Luiz Antonio Joia and Paulo Sanz, 2005).
With the rapid development of information and communication technology, Internet
and World Wide Web (WWW) have become important tools in business, which has
a major impact on business world. In the virtual environment of e-service, for
customers website is the main access to online organizations and to a successful
purchase process. The deficiency of website design can result in a negative
impression of the website quality to the customers, and customers may exit the
purchase process. Website is the starting point for customers to gain confidence.
Website design can influences customers perceived image of company, and attract
customers to conduct purchasing online easily with good navigation and useful
information on the website. Website should provide appropriate information and
multiple functions for customers. (Li Hongxiu, Liu Yong, Suomi, 2009).
Since the main purpose of technical support Web sites is to provide information and
after-sales service, they do not emphasize online transactions. A review 3of functions
provided by real-world technical support Web sites shows that many do not even
support online transactions. The transaction-based approach, therefore, may also not
adequately measure the service quality of a technical support Web site. To date, few
studies have explored factors that determine customers service quality perceptions
of technical support Web sites. The present study, therefore, aims to fill this research
gap. (Po-Young Chu, Chia-Yi Chen, Yu-Ling Lin, Wei-Chen Wu, 2009)
This current study examines the combined effects of perceived information
availability and perceived e-service quality on customers retention value of e-
retailing.
H1: there is a positive effect of Information Quality of e-Service quality on the
customer retention towards e-retailing
2.2.2 Website Security
Security refers to the freedom from danger, risks or doubts during the service
process. This dimension holds an important position in e-service. Customers
perceive significant risks in the virtual environment of e-service stemming from the
possibility of improper use of their financial data and personal data, which is an
270
(ICoIT'17)
10thof April, 2017.
important barrier to online customers to purchase online. The following are the
attributes of security dimension of e-service quality.
- Protect the financial data of customers.
- Protect the personal data of customers.
- Terms on payment and delivery.
- Good reputation.
Research on service quality on the Internet has been mainly based on the
SERVQUAL model. Recent studies have also tried to model and measure e-Service
Quality. Research in the e-retailing context (Zeithaml, 2002) showed that e-SQ has
seven dimensions that form two scales: including efficiency (ability, easiness to get
to and navigate a Web site), fulfillment (having products in stock and delivering
them on time), reliability (technical functioning of a Web site), and privacy
(assurance regarding data sharing and security); and a recovery e-Service Quality
scale including responsiveness (provide appropriate data when problems occur,
online guarantees and mechanisms for handling returns), compensation and contact.
(Marianna Sigala&Odysseas Sakellaridis, 2004).
In e-service, interactivity between customers and companies offers opportunities for
companies to obtain information about customer, such as purchasing habits, needs,
preferences and so on, which makes it possible to offer customer personalized
service. Personalized service can improve customers satisfaction, and customers will
be reluctant to try other companies. Not only services and products, but also payment
and delivery in the service process, can be personalized, which is aimed at meeting
all the requirements and demands of customer. Personalization constitutes a vehicle
to retain the customer in e-service. This research examines the combined effects of
Website Security and perceived e-service quality on customers retention value of e-
retailing.
H2: There is a positive effect of Website Security of e-Service quality on the
customer retention towards e-retailing.
2.2.3 Service Reliability

Reliability on e-service quality is defined as ability to perform the promised service
dependably and accurately. Reliability is associated with the technical functioning
of the site, particularly the extent to which it is available and functioning properly.
The privacy dimension includes assurance that shopping behavior data are not
shared and that credit card information is secure. It is important to stress that the
solution space only contains feasible solutions. In the example above, if a specific
interconnection mix violates the requirements of the Internet Service Providers with
respect to other criterias like security or reliability it is not considered feasible and
therefore not part of the solution space.
271
(ICoIT'17)
10thof April, 2017.
The assessment of service quality is largely based upon Parasuraman et al.s (1988)
original development of the measure of perceived service quality. Considerable
empirical examination of the reliability and validity of the scale has resulted in a 22-
item scale. Reliability is the dominant dimension in traditional service quality. In the
context of the web, reliability is defined as the proper functioning of the website and
the ability to perform the promised services dependably and accurately. (Lee & Lin,
2005) Moreover, the reliability of the website can be reflected in the reliability of
the information that is captured by information quality. Information quality should
meet certain criteria, namely: accuracy, timeliness, relevance and understandability.
The presence of high quality information increases satisfaction with both the
experience and product purchase.
The research on the reliability towards service quality has passed more than twenty
years which says that it is one of the most important dimensions for all the services
provided. (Santos, 2003) quoted to reliability as the ability to perform the promised
service accurately and consistently, including frequency of updating the web site,
prompt reply to customer enquiries, and accuracy of online purchasing and billing.
Reliability specifically involves consistency of performance and dependability.
This research investigates on the factors of e-service quality influencing the
customers retention on e-retailing. Thus it is very important to find the contribution
of e-service quality which can change the customers retention in e-retailing in
Malaysia.
H3: Reliability of the Website is having positive influence on the customer retention
in e-retailing.
3. RESEARCH METHOD
3.1 Research Problem
The main problem for the e retailing companies is of convincing surfers to advance
beyond monitor shopping'' to make purchases and, more importantly, to become
repeat customers has proven to be difficult. On average, only 5 percent of unique
visitors to sites become customers and only 1.6 percent of all visits result in purchases
(Boston Consulting Group, 1998). The problem is that consumers develop entrenched
shopping habits exhibiting complex and interrelated behaviors and responding to
various motivators to make their purchase decisions. Successful e-commerce retailers
tap into these motivators and present new motivators to encourage consumers to alter
their normal pattern of behavior and purchase over the Internet.
(Brandt, 1997) quoted that the primary reasons for these difficulties reside in service
providers lack of experience in streamlining operations of this fairly recent channel
and their limited understanding of online customers usage behaviors. Moreover, as
online firms often possess limited resources, priorities must therefore be determined
among alternative service attributes when making investment decisions related to
272
(ICoIT'17)
10thof April, 2017.
achievement of superior services. As such, it is desirable for online service providers

to uncover what attributes consumers utilize in their assessment of overall service
quality and satisfaction and which attributes are more important (Zhilin Yang &
Xiang Fang, 2004).
3.2 Research Questions
What is the impact of Information Quality of the service quality on the
customer retention for e-retailing in Malaysia?
How the Internet Security have high impact on the customer retention for e-
retailing in Malaysia?
How Service reliability of the service quality have impact on the customer
retention for e-retailing in Malaysia?
3.3. Research Hypothesis

Hypothesis 1
H0: There is no influence of information quality of the e-service quality on the
customer retention for e-retailing in Malaysia.
H1: There is a positive influence of information quality of the e-service quality
on the customer retention for e-retailing in Malaysia.
Hypothesis 2
H0: There is no influence of Internet Security of the e-service quality on the
H1: There is a positive influence of Internet Security of the e-service quality on
the customer retention for e-retailing in Malaysia.
Hypothesis 3
H0: There is no influence of Service Reliability of the e-service quality on the
H1: There is a positive influence of Service Reliability of the e-service quality
on the customer retention for e-retailing in Malaysia.
4. DATA ANALYSIS
4.1. Descriptive Statistics
Descriptive statistics are used to describe the basic features of the data in a study.
They provide simple summaries about the sample and the measures. The respondents
chosen for data collection are located in the areas located in the Selangor and Kuala
Lumpur of Malaysia. The major respondents targeted were the high level and highly
educated customers having experience of online purchasing in past. A set of
questionnaires was developed and sent to more than 200 respondents.
Data gathering proceeded by telephone interview with a follow up mail out of the full
instrument, inclusive of responses made to the subset of questions. Validation testing
273
(ICoIT'17)
10thof April, 2017.
was performed with a cross tabulation of non-respondents and respondents. From this
testing it was found that 148 respondents performed positively whereas the rest of 52
were considered as a non-respondent. However, the research was further extended by
investigating on the customers experience on e-retailing before. This investigation
would seek to know the confidence level of the customers.The data of the respondents
are described with the frequency distributions and depending on the particular
variables all the datas have been represented and then they are grouped into
categories first. Overall in total around 148 questionnaires were considered as 100%
contributed and were included as a final data. And as outlined below are the
demographic profiles of the respondents Table 1:
Table 1: Demographic profile of respondents
Demographics Frequency Percentage

Age:
18 to 27 41 27.70%
28 to 37 64 43.20%
38 to 47 34 23.00%
48 and above 9 6.10%
Gender:
Male 95 64.20%
Female 53 35.80%
Qualification
Diploma 24 16.20%
Degree 51 34.50%
Master 35 23.60%
Professional/Doctorate 29 19.60%
others 9 6.10%
internet access
Yes 125 84.50%
no 23 15.50%
Online usage
less than one hour 10 6.80%
1 to 5 hours 70 47.30%
more than 5 hours 68 45.90%
E-retailing
Yes 50 33.80%
no 71 48%
In future 27 18.20%
E-retailing experience
less than 1 year 9 6.10%
274
(ICoIT'17)
10thof April, 2017.
1 to 5 year s 19 12.80%
more than 5 years 22 14.90%
no experience 98 66.20%
From the demographic results it was found that most of the respondents were at
the range of 28 to 37 with 43.2% of the respondents, whereas less respondents
were from the age of above 47 years with 6.3%. It is concluded from the age as a
demographic profile of respondents that the middle aged peoples are more
attracted towards the online purchasing.
Gender was also taken into account as a medium of profile to investigate on the
taste of the customers on the e-retailing. From the findings it was noted that around
64.2% of the respondent were masculine were as around 35.8% of the respondents
were feminine.
Qualification or education are very important to know whether the respondent is
properly educated to seek the knowledge on the usage of online purchasing. From
the analysis it was known that the respondents with degree as their prime
qualification were around 34.5%, whereas the master degree holders were found
to be of 23.6% which they are the bigger number. Thus from this finding it can be
said that the respondents were having a good understanding on the research topic.
The next question asked to the respondents was regarding the internet access
availability. Almost more than 80% of the respondents were having access to
internet via hand phone or through internet connection or broadband.
After the access of internet the next question that was asked to the respondents
was regarding the internet usage in their daily activities. The response was quite
positivistic as majority of the respondents use internet of more than 5 hours daily.
This shows the importance of internet in their daily life.
The next important question that was asked to the respondent was regarding the
online purchasing intention. This question was quite negative to the research as
the majority of the respondents were against e-retailing. From the finding it was
noted that around 48% of the respondents have never performed online purchasing
in their life, whereas around 18% shows their intention to do e-retailing in future.
This shows that the belief and trust on the e-retailing. Thus it can be said that the
antecedent variables taken in this research could increase the trust on the e-retailed
customers. This approach would increase the usage on technology in this 21st
century of modern era.
4.1.Correlation Test
Correlation analysis is used to know how the variables are correlated with each
other. For the collection of the datas the symmetric quantitative variables for the
Pearsons correlation coefficient is used. Before calculating the correlation
275
(ICoIT'17)
10thof April, 2017.
coefficients we need to screen the datas for the outlier. Pearsons correlation
coefficient assumes that each pair of the variables is the bivariate normal and it is
a measure of the linear associations. For the Pearsons correlation coefficient to
be efficient the variables need to be linear and accurate. For example there must
be a linear relationship between the dependent and the independent variables. And
as outlined in Table 2:
Table 2: Correlation analysis of the data
The significance level of each correlation coefficient is stated in the correlation

table. The significance level or the p-value is the probability of obtaining results
as extreme as the one observed. If the significance level is small i.e. less than 0.01
then we can say that the two variables are linearly correlated. If the significance
level is large than 0.01, then we can say that the two variables are not significant
and also they are not linearly related. In the above figure we can see the significant
level of all the three independent variables is having high value of Pearson
correlation and have good significant of less than 0.01 so we can say that the
variables may have positive relationship with each other. The figure also shows
that there is a high correlation between information quality and e-retailing from
all the other independent variables. There is a strong relation between the
independent variables.
4.2.Reliability Test
The reliability test is having the main efforts in identifying and knowing the exact
result of the research because we can know from this test whether the test is
reliable or not. This result will also give us an idea whether the test conducted was
correct or not. Only one item question cannot be used to measure the reliability
test. For the reliability test there must be more item questions so that the result
could be accurate. Here the mean of the items are taken into consideration for the
test. Mean is taken to know the exact reliability.
276
(ICoIT'17)
10thof April, 2017.
Variable Name No of Mean (Std Cronbachs Comments

Items deviation) alpha
Information 6 4.31 (0.370) 0.789 Strong
Quality
Website 3 4.28 (0.375) 0.734 Strong
security
Service 6 4.29 (0.337) 0.782 Strong
Reliability
E-retailing 5 4.60 (0.434) 0.706 Moderate
Overall 4 0.927 Strong
Table 3: Reliability test of the variables
as outlined in Table 3 which shows the Reliability test of the variables . the
reliability test says that the Cronbachs alpha of overall is 92.7 % which indicates
us that the test is reliable and can proceed further. The closer Cronbachs alpha
coefficient is to 1.0 the greater the internal consistency of the items in the scale.
George and Mallory (2003) provide the following rules of thumb_ > .9
Excellent, _ > .8 Good, _ > .7 Acceptable, _ > .6 Questionable, _ > .5 Poor,
and _ < .5 Unacceptable. As our result comes under the range of excellent, this
says that there is a deep influence of the independent variable on the dependent
variable.
Also the other independent variables information quality, Webiste security and
service reliability are having good reliability and covers the rule of thumb criteria.
Thus all the variables are exactly fit for the study and can proceed for further
analysis. Also the reliability test talks positively on the accuracy of the
questionnaire.
4.3. Linear Regression Analysis
Regression analysis is a statistical tool for the investigation of relationships
between variables. Usually, the investigator seeks to ascertain the causal effect of
one variable upon another. It shows the influence of one variable on the other.
To explore such issues, the investigator assembles data on the underlying variables
of interest and employs regression to estimate the quantitative effect of the causal
variables upon the variable that they influence. The investigator also typically
assesses the statistical significance of the estimated relationships, that is, the
degree of confidence that the true relationship is close to the estimated
relationship.
Thus after the linear regression analysis done individually between the dependent
and independent variables it was found that the information quality was having
277
(ICoIT'17)
10thof April, 2017.
influence of 77.7% on the customer retention of E-retailing in Malaysia, and it

outlined in table 4.
Although all the variables are having high influence they cannot be neglected as
they all are significant.
Variables R square Significant level
Information Quality E-retailing 77.7% 0.000
Website Security E-Retailing 67.7% 0.000
Service reliability E-Retailing 65.3% 0.000
Table 4: Linear regression analysis of the data
4.4.Coefficient Model
In multiple regressions the model takes the form of an equation that contains the
coefficient of the independent variables.
From the above table the b value indicates us the individual contribution of the
independent variables on the dependent variable.
The b value also tells us the relationship between the predictor and the outcome.
Thus from the table the following equation is obtained:
Y = a + b1X1 + b2X2 +b3X3 + e
Y = -.850 + 0.526 (Information quality) +
0.346 (website security) +
0.395 (service reliability) +e
From the above estimated equation we can say that all the factors are positively
related with the dependent variable customer retention on e-retailing in Malaysia.
The equation above shows the coefficients of the independent variables. From the
coefficient analysis we can say that the information quality and the website
security play a significant role in enhancing the retention of the customers to adopt
e-retailing and make easy online transactions. From the findings it seems that the
information on the quality of the products and also the transaction information are
very important in order to increase the online purchase increase and marketing by
the producers.
278
(ICoIT'17)
10thof April, 2017.
Table 5: Coefficient model of the variable
4.5.Multiple Regression Analysis

Multiple regression analysis is a technique that allows additional factors to enter
the analysis separately so that the effect of each can be estimated. Multiple
regressions are essential when the investigator is only interested on the overall
influence of the dependent variable due to the independent variables.
Table 6: Multiple regression of the variable
Here we have the R Square value as 0.889, which states that approximately 88.9%
of the dependent variable is influenced by the independent variable. The result
was not able to feel the total gap of the research. Although the results are relevant
to investigate on the dependent variable cannot say that the variables taken for the
research was enough to investigate on the wide topic of customer retention on e-
retailing.
The significant level is also satisfied and acceptable. To confirm that our analysis
is acceptable we also need to see the difference between the R Square and
Adjusted R Square. If there is a wide difference in their value than the model is
not acceptable.
Thus overall the model is fit as it satisfies all the criteria for the model fit in the
regression analysis.
5. CONCLUSION AND RECOMMENDATION

5.1.Findings and Conclusions
From the recent evidence it is clear that the factors involved in the e-retailing
intention by the customers have great impact on it and also for an effective growth
of the retail producers in Malaysia. From the analysis it was found that all the
factors are having positive relationship with the customer intention of e-retailing.
That means all the factors are playing there part for enhancing the retail business
in Malaysia with an influence of having great impact on the minds of customers.
From the analysis the main contribution was fulfilled by the Information quality
factor and Service reliability factor. This means there is a positive relationship
between Information qualities with the Customer retention on e-retailing in
Malaysia. Findings show that when there is an increase of one unit in the
279
(ICoIT'17)
10thof April, 2017.
information quality, then there is an increase of around 52.6% in the customer

retention on e-retailing in Malaysia.
From the practical point of view we can say that there is a significant effect of
information quality to have impact on the customer retention on adoption of e-
retailing in Malaysia.
Hypothesis Findings
H1: From the analysis it was found there is a positive relationship between
information quality and customer retention on e-retailing in Malaysia. There is a
total influence of 52.6% of the independent variable information quality on the
dependent variable customer retention on e-retailing in Malaysia.
H2: From the findings we can say that there is a positive relationship between
Internet security and the customer retention on e-retailing in Malaysia. The
analysis found that there was in total 34.6% of influence on the customer retention
for e-retailing in Malaysia.
H3: The finding says that there is a positive relationship between the service
reliability and the customer retention on e-retailing in Malaysia. In total there was
39.5% of influence of service reliability on the customer retention for e-retailing
in Malaysia.
By using the empirical model through regression analysis we have noticed that the
information quality on the retail product is the main factor to watch for an increase
in the customer retention on e-retailing. The quality of information availability
needs to be speculated and demonstrated at every moment which will help the
customer to make a decision on online purchasing.
5.2.Recommendation of the Study
Previous research has shown that the widespread consumer experiences of
inadequate e-service quality stall the growth of the emerging economy. Therefore,
it is important to develop an appropriate instrument to measure e-service quality.
The researcher able to find three dimension representation of e-service quality. In
terms of actual performance in relation to the provision of e-service quality on
their e-retailing sites, the banks appear to be performing well. The theoretical
implications of this research are manifold. First, the support for and importance
placed upon all service quality dimensions suggests that different aspects of e-
service quality.
Customers prefer multiple channels to make transactions and to access online
service providers, including phone, email, and physical access, even if they intend
to execute most transactions over the Web site. Our results suggest that customers
are concerned with three key factors closely related to information systems
quality, i.e. ease of use, timeliness of information, and security. Additionally,
280
(ICoIT'17)
10thof April, 2017.
adequate navigation functions such as site search engines and clear menus are
critical factors in enhancing the usability of a Web site.
The variables were not enough to indicate the total gap of the research. Thus it is
necessary to analyze the dependent variable in future research with different variables.
5.3.Limitation of the Study

As this research was very widely popular then also the information based on the
literature review was not easily available.
There was not much time for the completion of such a wide sector of e-retailing
and the customer retention to analyze which was a major limitation of the research.
There are various other factors affecting the customer retention on e-retailing in
Malaysia which were not included in this research. Thus only 80% of the influence
was explained by the variables and the rest of 20% were still remains to evaluate.
The number of respondents for conducting the research was not sufficient to go
on major decisions. As the sample is derived from one region of Malaysia,
generalizability of the results is limited. The items are drawn from existing
measures and although tests for reliability and validity provide initial support for
the adopted measures, there remains a possibility that not all e-service quality
dimensions are captured.
Finally, although this research considers customer retention as an overall outcome,
future qualitative research may provide a better understanding of the factors
influencing lower levels of customer satisfaction with current e-retailing
approaches and uncover aspects that are missing from current banking online
service provision.
5.4.Suggestions for the Further Research
It was known from the current research that there are various other factors which
affect the customers retention in Malaysia, thus it is highly suggested to include
these variables in the further research.
It is also highly suggested to know the key areas that have high effect on customer
retention on e-retailing have to look to increase the customers importance on the
retail products in Malaysia.
There must be involvement of the other factors like user friendliness, Efficiency,
Site organization and several others as a variable in the further research.
Retail firms should look upon the improvement on the products and service quality
to increase the value of the product in customers retention.
References
1) Brandt, R. (1997), Satisfaction studies must measure what a customer wants and
expects, Marketing News, Vol. 31, p. 17. Carmel Herington and Scott Weaven,
281
(ICoIT'17)
10thof April, 2017.
2009, E-retailing by banks: e-service quality and its importance to customer

satisfaction, European Journal of Marketing Vol. 43 No. 9/10, pp. 1220-1231
2) Fang Wang and Milena M. Head, 2001, A model for Web-based information
systems in e-retailing, Internet Research: Electronic Networking Applications and
Policy Volume 11. Number 4 pp. 310321
3) J. Yang, G. Miao, 2005, the estimates and forecasts of worldwide e-commerce. In:
Q. Li and T.-P. Liang (Eds.), Proceedings of the 7th international conference on
Electronic commerce (ICEC 2005), Xi'an, China, pp. 52-56.
4) John R. Rossiter, 2007, Toward a Valid Measure of E-Retailing Service Quality,
Journal of Theoretical and Applied Electronic Commerce Research, VOL 2 / ISSUE
3 / 36 48
5) Lee, G. & Lin, H. (2005). Customer perceptions of e-service quality in online
shopping. International Journal of Retail & Distribution Management, 33(2), 161-
176
6) Li Hongxiu, Liu Yong, Suomi Reima, 2009, Measurement of Eservice Quality: An
Empirical Study on Online Travel Service, 17th European Conference on
Information Systems, pg 1-13
7) Li, Hongxiu, Liu, Yong, Suomi, Reima, 2009, Measurement of Eservice Quality:
An Empirical Study on Online Travel Service, www.ecis2009.it/papers/ecis2009-
0449.pdf
8) Luiz Antonio Joia and Paulo Sanz, 2005, The hidden value of sporadic customers in
e-retailing An empirical investigation, Internet Research Vol. 15 No. 1, pp. 7-20
9) Marianna Sigala & Odysseas Sakellaridis, 2004, Web Users Cultural Profiles and
E-Service Quality: Internationalization Implications for Tourism Web Sites,
Information Technology & Tourism, Vol. 7 pp. 1322
10) Noor Habibah Arshad et al, 2008, Empirical Study on Electronic Service Quality,
Public Sector ICT Management Review, Vol. 2, No 1.
11) Parasuraman, A., Zeithaml, V.A. and Berry, L.L. (1988), SERVQUAL: a multiple
item scale for measuring customer perceptions of service quality, Journal of
Retailing, Vol. 64 No. 1, pp. 12-40.
12) Parasuraman, V.A. Zeithaml, A. Malhotra, 2005 . E-S-QUAL, a multiple-item scale
for assessing electronic service quality. Journal of Service Research, Vol. 7, No. 3,
pp. 213-233.
13) Po-Young Chu, Chia-Yi Chen, Yu-Ling Lin and Wei-Chen Wu, 2009, Determinants
of the Service Quality of Technical Support Web Sites: An Empirical Study of IT
Companies in Taiwan, International Journal of Business and Information, Volume
4, Number 1
14) Renata Mekovec, 2007, A Method for Improvement of Objectivity of E-Service
Quality Evaluation, Original scientific paper, UDK 330.322:336.76
282
(ICoIT'17)
10thof April, 2017.
15) Rust, R.T., and Kannan, P.K. 2003. E-service: A new paradigm for business in the
electronic environment, Communications of the ACM 46(6), 36-42.
16) S. Ba, J. Stallaert, Z. Zhang, 2005, Brand recognition, service quality and price
competition. In: Q. Li and T.-P. Liang (Eds.), Proceedings of the 7th International
Conference on Electronic Commerce (ICEC 2005), Xi'an, China, 2005, pp. 350-354.
17) Santos, J. (2003), E-service quality: a model of virtual service quality dimensions,
Management Service Quality, Vol. 13 No. 3, pp. 233-46.
18) Veneta Andonova, 2003, Differences in the Behavior of Traditional Retailers in
Adopting E-Commerce, Management Research, vol. 1, no. 3, pp. 279290.
19) Zeithaml, V.A., Parasuraman, A. and Malhotra, A. (2002), Service quality delivery
through web sites: a critical review of extant knowledge, Journal of the Academy
of Marketing Science, Vol. 30 No. 4, pp. 362-75.
20) Zhilin Yang, 2004, Online service quality dimensions and their relationships with
satisfaction, International Journal of Service Industry Management, Vol. 15 No. 3,
pp. 302-326
21) Ziqi Liao and Xinping Shi, 2009, Consumer perceptions of internet-based e-
retailing: an empirical research in Hong Kong, Journal of Services Marketing 23/1,
2430
283
(ICoIT'17)
10thof April, 2017.
Establishing Educational Robotics Labs in Iraqi Universities

Dr.Ghada Saleem
Educational Robotics Lab., College of Education for Pure Science, University of Baghdad - Iraq
ghadasaleem@gmail.com

The Iraqi universities are seeking to develop the education
Article History: and catching up with the scientific progress and technological
Received: 19 March 2017 advances in the industrialized countries. One of the most
important goals of education development is to graduate
Accepted: 1 April 2017 students capable of competition in the labour market, and
Published: 10 April 2017 participate in the society development. The Iraqi universities
aimed to give the students level of experience depending on
the basis of scientific knowledge to help students solve the
Keywords: STEM
real world problems. The universities also seek to provide
Education, Practical
education linking the individual with the environment and
Robotics, Open Source
Community, Raspberry Pi, economic requirement. The mission of the Educational
Arduino, 3D Printing, Robotics Lab. at the University of Baghdad is to provide
GitHub, COIL Courses. students a rigorous, interdisciplinary learning environment
focused on Science, Technology, Engineering, Mathematics
(STEM), to foster the joy of discovery, and to promote a
collaborative culture of innovative problem solving. STEM
education plays a big role as the catalyst to meet the
challenges and demands of our present and future economy.
1. INTRODUCTION
S
TEM education integrates concepts that are usually taught as separate subjects
in different classes and emphasizes the application of knowledge to real-life
situations. A lesson or unit in a STEM class is typically based around finding
a solution to a real-world problem and tend to emphasize project-based learning [1].
Many STEM lessons involve building models and simulating situations. A good STEM
lesson ensures that students understand the connection to the real world according to
[2]:
- Science: - discover and describe a better understanding of life (What is).
- Technology: - Invent and innovate by improving the natural world.
- Engineering: - control, modify, or design materials, processes, and systems (What
could be).
- Mathematics: - symbolic language for representing reality.
284
(ICoIT'17)
10thof April, 2017.
The educational robotics lab provides a unique opportunity for teachers to place
engineering design, scientific processes, technological literacy, and mathematics in
context that students find engaging and understanding.
The Lab. Includes both physical devices, and virtual simulation software that enable
faculty, students, technical staff, and researchers to improve their skills, and update their
knowledge in the direction of innovation and transfer it as educators to indirect impact
to Iraqi high schools.
2. The Educational Robotics Lab. Goals

The goals of University of Baghdad for establishing the educational robotics lab are:
Bolster the innovative capacity of Iraq workforce, which is falling behind other
countries that are creating higher numbers of STEM-trained individuals each year
compared to Iraq.
Boost the proficiency of all students in basic STEM knowledge. This goal is
designed to improve the ability of students and workers to assess problems,
employ STEM concepts, and apply creative solutions in their daily lives.
Build an Iraq robotics team from different Iraqi schools in the ages K9-K18 to
participate in the Arabic and international competitions in robotics field.
Together, these goals are intended to enhance the global competitiveness of Iraq economy
and help individuals achieve economic security in their careers.
The Lab. in its initial undertaken was looking at three venues to disseminate the idea
behind STEM through:
- Firstly: by training undergraduate students in the college of education for pure
science /University of Baghdad, because they are the future teachers in the Iraqi high
schools. The college is responsible for the generation of teachers in the ministry of
education so those will be the ambassadors of new technologies and STEM
education.
- Secondly: The educational robotics lab convenes workshops for the teachers and
high schools students to gain acquaintance in STEM fields.
- Thirdly: The lab has a plan to gain more stakeholders acceptance of these ideas in
order to flourish the concept of STEM and gain support for the work in order to
extend to other customers.
3. The Educational Robotics Framework

During the work in the educational lab, there are many achievements occurred,
which reflects the Iraqi STEM education framework shown in FIGURE1, these are:
285
(ICoIT'17)
10thof April, 2017.
Understand and focus

on STEM education
Assess
Assess data related to goals Create relevant STEM
& adapt plans as needed goals and STEM plan
Support Faculty and staff Access resources &

through coaching and professional development to
mentoring support STEM Goals
FIGURE 1. The Iraqi STEM Education Framework
- University of Baghdad established and developed educational robotics lab in the

college of education for pure science to be the first lab in the Iraqi universities that
accomplish STEM education.
- University of Baghdad faculty learned and practiced STEM skills and knowledge
and applied them to current courses and curriculum.
- University of Baghdad/Iraq, Gannon University/ USA, and the American
University of Madaba/Jordan developed an international and interdisciplinary
collaborative team thereby establishing a regional partner.
- University of Baghdad/Iraq, Gannon University/ USA, and the American University
of Madaba/Jordan developed and teaching a Collaborative Online International
Learning(COIL) course-Practical Robotics. This course engages faculty and
students from the three universities in collaborative learning and projects.
4. The Lab Activities

The lab is established to accomplish different activates that enable the college of
education for pure science to be the first college implementing the STEM education
concepts. These activities are:
1- The lab is stuffed with gradated students and faculty of the college of
education to coordinate the different workshops, training, and researches in STEM
and educational robotics.
286
(ICoIT'17)
10thof April, 2017.
2- The engagement in open source community through the curriculum and

training courses in practical robotics such as Arduino, Rraspberry Pi and 3D
printing.
3- Design and execute a learning platform dedicated to STEM education called
the Iraqi STEM Platform as shown in FIGURE 2. This is a tool that enables
students and faculty to interact actively in online learning environment, and
increase the community awareness in these new concept and technologies. The
structure of the platform is shown in FIGURE 3.
FIGURE 2. The main page of the Iraqi STEM Platform
287
(ICoIT'17)
10thof April, 2017.
FIGURE 3. The Structure of the Iraqi STEM Platform
4- Building a team from different colleges in Baghdad University to participate

in the COIL course, the team gains the skills of building robots and practices the
STEM concepts through the engagement with different international universities
online and in real-time environment.
4.1. The Educational Robotics Curriculum

The curriculum of the STEM courses was designed to provide the knowledge
and skills in robotics field, this includes:
The 3D printing topics: open source community delivers many H/W and S/W
platform tools for 3D printing. The lab equipped with the latest technology in 3D
printing which is LulzBot printers, it is one of the RipRap printers [3] that uses
the fused filament fabrication (FFF) [4] technology as shown in FIGURE4. The
software program used to design the 3D models is sketchup[5], and the 3D
printing program is Cura[6] ,as shown in FIGURE5.
288
(ICoIT'17)
10thof April, 2017.
FIGURE 4. Fused Filament Fabrication
FIGURE 5. 3D printing Process
The Arduino platform topics: It is one of the open source H/W and S/W
platform which enables the students to gain skills in electronics components,
electrical components like power supply and motors, sensors, actuators, and S/W
C++ IDE programming skills [7], as shown in FIGURE 6.
289
(ICoIT'17)
10thof April, 2017.
FIGURE 6. Arduino Platform
The Raspberry Pi platform: It is another new open source platform that

enables the students to gain skills in communication protocols, talking to sensors,
writing code in different languages such as Python, and Java[8], FIGURE 7,
illustrates different types of educational robotics implemented using Raspberry
Pi 2 & 3 Model B.
290
(ICoIT'17)
10thof April, 2017.
FIGURE 7. Raspberry Pi Robots
The Software Development Platform GitHub: It is the most popular

repository of the Open-Source Communitys programming contributions. Even
when open-source projects include non-code elements such as electrical
components and/or 3D printed models, the associated program is most often
uploaded to GitHub. GitHub also offers an arsenal of project management tools
that allow programmers to collaborate on a project as shown in FIGURE 8.
Programmers often download code from GitHub improve it and then upload the
improved code back to the original code [9].
FIGURE 8. The GitHub Platform
291
(ICoIT'17)
10thof April, 2017.
Practical Robotics COIL course: COIL is a teaching and learning methodology

that provides innovative cost-effective internationalization strategies, fosters staff
and student interaction with peers abroad, co-taught multicultural online and
blended learning environments, and emphasizes experiential student
collaboration.
The practical robotics coil course is a pragmatic approach to the technical skills
that fall between disciplines. This course spans topics in electrical, mechanical,
computer, and software engineering and networking. Students will learn how to
utilize the open-source community to find solutions. Students will learn to write
programs to employ and control typical robot components including motors,
servos, sensors, and cameras. Students will learn to create circuits and power
systems to deploy and operate these components. Students will learn the skills to
connect and communicate with robotic components using multiple
communication protocols such as I2C, SPI, and UART. Students will learn IOT
concepts and utilize internet protocols such as TCP, UDP, and SSH to connect
and coordinate robots in close proximity and across the world. This is a
Collaborative Online International Learning (COIL) Course as shown in FIGURE
9; the following table (TABLE 1) illustrates the syllabus of this course
FIGURE 9. The COIL Course
TABLE 1: The Syllabus of the practical Robotics Course
292
(ICoIT'17)
10thof April, 2017.
Lecture Theme Topic

1 Syllabus Overview of Course
2 3D Printing Introduction
3 Calibrating
4 GCODE Commands and Format of GCode file
5 3D Modeling Sketch Up
6 Designing& Printing Printing Projects
7 EXAM 1
8 Arduino Quick intro + Arduino
9 Software Dev Software Development Raspberry Pi
10 Long Range Communication Protocol
11 Communications Connecting Across Globe
Raspberry Pi, Netbeans, Java, Remote
12 Workshop Platforms + Communication Protocols
13
Videos
14 Introductory videos in each classroom
15 Constructing Power Circuits
Power Systems
16 Power Regulation
17 Controlling Motors Controlling Motors Pres
18 GitHub/HowTo How to contribute to OS community
19 Teaching Motors to each other
Controlling Motors
20 Teaching Motors to each other
21 Register Maps and Variables
Advanced Code Reading
22
23 Control Structures and Manual Control
Algorithmic Problem
24 Solving Artificial Control
25 Math and Balancing Balancing Algorithm and PID/PIPD
26
27 Math and Balancing
28 Collaboration and Competition
Final Exam
5. CONCLUSION
The educational robotics lab has a vision for an innovative future that stimulates action
among key stakeholder groups and that helps identify what we know currently, what needs
to be discovered, and what needs to be developed to achieve the ultimate goalcreating
equity of opportunity in STEM to promote lifelong learning among the Iraqi youth.
Certainly, challenges remain, but lessons can be learned and ideas for moving forward can
be developed. In this process of discovery and knowledge propagation, Iraqi universities can
adjust and refine the future of STEM education as new experiences and new evidence are
293
(ICoIT'17)
10thof April, 2017.
gained, particularly about what approaches work best in certain contexts and to serve diverse
learners.
The Educational Robotics lab established to be a prototype for other Iraqi universities to
follow and even it is equipped with simple and open source components, this represents the
importance of engaging in the open source community and enabling the students to build
their own robots from scratch with low cost materials aiming to teach them the STEM
concepts and enhancing their skills and abilities in this important field.
One of the important activities in the lab is the participation in the collaborative online
international learning course in practical robotics with Gannon University and the American
University in Madaba/Jordan. It is the first time of the lab to engage in the course which
introduced also for the first time at Gannon University.
Educational robotics is a pragmatic approach to the technical skills that fall between
disciplines. This will provide a unique opportunity for the universities in Iraq to reform and
modernize the higher education process.
6. REFERENCES
[1] Robert M.Capraro, STEM Project-Based Learning, 2d. ed., SENSE PUBLISHER,
2013.
[2] STEM 2026 A Vision for Innovation in STEM Education, U.S. Department of
Education (the Department), 2016, the report is available on the Departments
website at https://innovation.ed.gov/what-we-do/stem/.
[3] Lulzbot 3D printers, https://www.lulzbot.com/education.
[4] RepRap 3D printers, http://reprap.org/wiki/Fused_filament_fabrication.
[5] SketchUp: 3D modeling for everyone, https://www.sketchup.com/.
[6] Cura 3D Printing Slicing, https://ultimaker.com/en/products/cura-software.
[7] Arduino Development platform, https://www.arduino.cc/.
[8] Raspberry pi platform, https://www.raspberrypi.org/.
[9] The software development platform Github, https://github.com/
294
(ICoIT'17)
10thof April, 2017.
Evolving fuzzy neural network equalization of channel

impulse response in optical mode division multiplexing
Anahed M. Kareem
Scientific Information and Technology Transfer Center, Ministry of Science & Technology,
Baghdad, Iraq
anaheedfareed@gmail.com
Assoc. Prof. Dr. Angela Amphawan
Head of Optical Computing & Technology Research Laboratory, School of Computing, Universiti
Utara Malaysia
angelaoptics@gmail.com

Optical mode division multiplexing (MDM) systems
Article History: suffer from the inter-symbol interference (ISI) issues due to
Received: 19 March 2017 nonlinear channel impairments in multimode fiber from
mode coupling and modal dispersion. Existing equalization
algorithms in MDM systems such as least mean square
Published: 10 April 2017 (LMS) and recursive least squares (RLS) operate linearly and
Keywords: evolving thus far, are unable to effectively mitigate the nonlinear
fuzzy neural network channel impairments. To address this issue, a nonlinear
(EFUNN), mode division evolving fuzzy neural network (EFuNN) equalization
multiplexing (MDM), scheme was developed in MATLAB to reshape the channel
multimode Fiber (MMF), impulse response of a MDM system comprising five
intersymbol interference Laguerre-Gaussian (LG) modes at the transmitter to five
(ISI). reference Gaussian pulses at separate time intervals at the
receiver in order to mitigate the effects of ISI from nonlinear
channel impairments.
1. INTRODUCTION
T
he remarkable growth of traffic volume caused by the exponential increase of the
number of Internet users and continual development of increased bandwidth has led to
a huge capacity of communication system [1] and the creation of various applications
[2]. One of the most well-established communication technologies is fiber optic, which
has easily become indispensable, standing prominently as the mainstay of the modern-day
information infrastructure [3]. The existing infrastructure in high-speed local area networks is
extensively based on the multi-mode fiber (MMF) due to its easy usage, good performance and its
provision of the required bandwidth for a short distance at a more reasonable cost than the
fundamental single-mode fiber (SMF) [4]. In order to meet the expansion in growth of data traffic,
the ability of SMF has been improved consistently last a few decades, and is converging to nonlinear
capacity limit [5]. For managing the future bandwidth predicament, in the last years, mode-division
295
(ICoIT'17)
10thof April, 2017.
multiplexed (MDM) transmission over MMF has garnered in much attention as an intelligent way to
deal with the capacity limits of SMF [6]. This is possible because the capacity of an MDM system
increases relative to the number of modes. In a practical MMF, nonlinear mode coupling is induced
as a result of the fiber manufacturing defects caused by non-circularity of the core, micro-bending,
macro-bending, or twists. Consequently, the signal on each fiber mode propagates down the fiber
with its own distinguished speed signal strength. Modes that take shorter paths arrive at the receiver
earlier than the ones taking longer paths. Mode coupling leads to linear and nonlinear modal
dispersion of the signal or spreading of the optical pulse in time. A short pulse becomes longer and
ultimately joins with the pulse behind, making it less probable for a reliable bit stream for recovery
due to inter-symbol interference (ISI) [7]. To date, compensation for MDM nonlinear channel
impairments, in particular, for de-multiplexing of the output signals and reshaping of the received
channel impulse response, has been realized by utilizing linear equalization strategies such as least
mean square (LMS) and recursive least squares (RLS) algorithms. The linear equalization strategies
are inaccurate considering the nonlinearity of the channel impairments. Accordingly, MDM
equalizers should be developed to be adaptive to time variation and nonlinear characteristics. Based
on nonlinear characteristics of the MDM channel impairments, it is of great interest to adapt the
EFuNN algorithm for nonlinear equalization of the channel impulse response at the receiver. EFuNN
is capable of modelling nonlinear system relationships and has been proven by Kasabov (1999) to be
a very powerful tool when used to data which is comparatively interference, uncertain, and
changeable. EFuNN which has been successfully applied in applications such as, adaptive learning,
knowledge-based systems and for prediction. Many studies have used either neural networks or fuzzy
logic in equalization communication systems. Most of these studies focus on the combination of NN
and fuzzy logic, in designing a system that uses the neural network for a self-learning characteristic.
This increases the data transmission accuracy, whereas, fuzzy logic allows the complexity of the data
to be reduced and it is able to represent knowledge in an interpretable manner for any uncertainty to
be dealt with. The fuzzy neural network evolves over time because of the nature of the constructional
development and the adaptation of the entire evolving connectionist system (ECOS) of which it has
become part. In this research, an equalization scheme for MDM base on EFuNN is developed to
reduce ISI which is caused by mode coupling by reshaping the channel impulse response and
measured the performance of the proposed equalization scheme in MDM based on root mean square
error (RMSE) and CPU-time. Furthermore, to obtain the distorted channel impulse response, an
MDM system over MMF is simulated in OptiSystem in conjunction with its Multimode Component
Library.
2. RELATED WORK
The existing equalization schemes for MDM system are based on several algorithms such as
LMS, RLS, and (RLSCMA).
o Title
N Author Algorithm Summery
MIMO Signal Processing It applies RLSCMA algorithm as the
1.
Zhao et al.
RLSCMA
For MDM With RLSCMA equalization scheme in MDM in FMF. The
(2014)
Algorithm [8]. received signal suffers from mode coupling
and mode delay, unless the limitations are
compensated by multi-input multi-output
296
(ICoIT'17)
10thof April, 2017.
(MIMO)-(digital signal processing) DSP

equalizers.
MIMO Signal Processing for The adaptation cost for RLS is high.
2.
Kahn et al.
LMS/RLS.
MDM. [6]. Adapting LMS suffers from higher asymptotic
(2014)
SERs and low cyclic prefix efficiency. It was
found that cyclic prefix transmission reduces
system throughput and average-power
efficiency and decreases complexity.
Table 1. Currently available equalization schemes in MDM
o Title
N
Author Algorithm Summery
EFuNN Based Load Amphawan EFuNN computational model was utilized

1. The EFuNN
Prediction Model In (2003) making prediction. It works through one pass
Dynamic Cell Sizing [10]. learning and automatically adapts the
parameter values as operating. The study found
that EFuNN is able to learn traffic sequences in
an adaptive way.
Neuro-Fuzzy Equalizers Raveendranathan, The study made use of a nonlinear, fuzzy

2. Neuro-
for Mobile Cellular (2013). adaptive equalizer. It was attached in mobile
fuzzy
Channels [11]. broadband and Ultra-Wide Band (UWB)
communication channels.
Table 2. Previous nonlinear equalization schemes in radio
substantial research effort has been made for several studies to investigate about the efficient
using of equalization techniques in wireless communication like artificial neural network (ANN),
fuzzy systems and their combinations. They have received attention in the various divisions of
wireless communication. This is because of their ability to mitigate the ISI as well as adaptation to
the nonlinear nature. It motivates the efforts to a potential of applying these techniques to the optical
fiber according to the similarity between them in terms of ISI. The optical system over MMF behaves
similarly with the radio system and shares the multipath nature of transmitted channel. It leads to an
overlap in the transmitted signal, causing random variations of the power of signals propagating in
different modes, which means that the transmitted signal arrives at the receiver in different times.
This phenomenon in a radio system is known as fading and it is called mode coupling in the optical
system over MMF. Both radio and optical systems have the same strategies to manage fading and
mode coupling. One of the techniques is equalization, widely used in radio networks. It can be used
in MDM system, in which RLS and LMS are the conventional algorithms. The aim of conventional
adaptive algorithms based channel equalizers is to reduce or mitigate the effects of ISI, which has
happened in the communication channel. These are gradient-based learning algorithms and hence, its
weights never reach their optimum values because of the mean square error (MSE) being trapped in
local minimum due to the possibility of training mode of the channel equalizer. This means that they
297
(ICoIT'17)
10thof April, 2017.
suffer from long training time, power consumption, and slow convergence. Channel non-linearity,
additive noise, and distortion, are issues that affect the performance of linear equalizer.
One of the robust alternative nonlinear constructions is the evolving fuzzy neural networks,
which offers superior performance than that of linear models due to its capability to represent
nonlinear functions, allowing the use of small number of parameters, is fast and easy trained
equalizer, learn fast from a large amount of data (using fast training), adaptive incrementally in an
online mode, has an open structure, and memorizes information that can be used at a later stage.
3. DESIGN THE PROPOSED ALGORITHM

In order to implement the EFuNN equalization scheme for MDM, the distorted data comprising
the channel impulse response from the receiver of the MDM system. It is required as an input into
the EFuNN equalization scheme. The distorted data embodies the nonlinear channel impairments
from mode coupling and modal dispersion in the MDM system as shown in FIGURE 1. The
transmitter consists of five lasers at a wavelength of 1550nm, each launching a specific Laguerre-
Gaussian (LG) mode into the MMF: Channel 1 launches LG 0 1 mode, Channel 2 launches LG 1 0
mode, Channel 3 launches LG 1 1 mode, Channel 4 launches LG 2 1 mode and Channel 5 launches
LG 1 2 mode. Each mode carries an independent data channel at 7 GB/s, thus achieving an aggregate
data rate of 35 GB/s for the 5 channels. The five independent data channels are propagated through
a 5km parabolic refractive index MMF. The presumed value for peak-index refractive index is 1.4,
attenuation is 0.25 dB/km, with consideration of modal coupling. The signal is then de-multiplexed
into five received data streams.
FIGURE 1. Simulation MDM system
FIGURE 2. Channel impulse responses of five channels FIGURE 3. Channel Impulse Response of Five Time-
Separated Gaussian Reference Signals
298
(ICoIT'17)
10thof April, 2017.
Results from the MDM system demonstrate that ISI is prevalent due to nonlinear channel effects
from mode coupling and modal dispersion, resulting in each mode at the transmitter to excite fifty-
five modes at end of the MMF and detected by the photodetector. The distorted channel impulse
response of fifty-five modes at the receiver for each channel is shown in FIGURE.2. This will be
used as the input to the EFuNN equalizer. After data collection, two data processing procedures will
be conducted in MATLAB to prepare for the training stage in the EFuNN equalization. These
procedures are:
3.1 Design of Gaussian Reference Signal
In any communication system, the transmitter sends the signals while the receiver detects and
receives the signals. The primary objective at the receiving end is to sample the received signal in a
manner that enhances the probability of reducing ISI. It is vital because real receivers sampling point
cannot be optimal always for every signal since any communication system will contain some
distortions and ISI [12]. The ISI lead to errors in the identified signals and causing information loss
in MDM systems, which is usually caused by the mode coupling. The signals should be-reshaped in
such a way that they avoid interfering each other at the optimal sampling point and should equalize
the input source signals well. FIGURE. 1. shows that the received channel impulse response of all
channels after the MDM system have broadened and overlap with one another due to mode coupling
and modal dispersion. The Gaussian reference signal is used to make some changes in the pulse shape
of the transmitted signals, so that it better suits its purpose for the communication channel, and avoids
ISI in the data transmission typically by limiting the effective bandwidth of the transmission. In this
research, MATLAB is used to simulate the Gaussian reference signal, which is based on Equation 1.
The Gaussian reference signal is described as:
1
e xmu /2 (1) [13]
2 2
2 2
Where sigma is the channel impulse width, mu is the mode at which the channel impulse response
is maximum, x is the mode number.
The performance of an EFuNN equalization scheme is the result of a training process to produce
an approximation as close as possible to the optimal signal with less RMSE and CPU-time. To come
up with a better performance of the equalization scheme, there is a need to design at least 5 Gaussian
reference signals for 5 distorted signals, = 5.5. Each one has a length of 11, adopting the proper
characterization of the Gaussian reference signals, as illustrated in Figure. 3.
3.2 Normalization process of the distorted signal, and the Gaussians reference signal
Normalization process of distorted signal is required because of the equalization scheme link and
learning with sequential and time-varying signals. It consists of a normalization of pre-processed data
to the networks working range from their natural range, therefore the normalized signal is highly
shaped to meet the reference signal of the equalization scheme and are adjusted to the nonlinearities
of the neurons, therefore their outputs should not beyond the capacity bounds. It can be fixed between
0 and 1. Regarding that, this study needs five channels and five Gaussian reference signals. Based
on Equation 2:
xni xi / max xi (2) [41]
Where xi = original level of the channel impulse response
299
(ICoIT'17)
10thof April, 2017.
3.3 Implementing the EFuNN equalization scheme
The distorted signals will be inserted along with the Gaussian reference signals after normalization
to implement the EFuNN equalization for MDM. The implementation was executed using MATLAB
R2014a (8.3.0.532) 64-bit (win64). A computer with 8GB memory (RAM) and due CPU core i5
(2.40) GHz has been utilized. Principally, the EFuNN is a five-layer neural network that is able to
incorporate expert knowledge (Kasabov & Woodford, 1999). The structure and parameters are driven
by the input variables. Koprinska and Kasabov (1999) showcase this five-layer structure as seen in
FIGURE. 4.
FIGURE. 4 Evolving fuzzy neural network (EFuNN)
Principally, the first layer of the network is characterized by the distorted signals and Gaussian
reference signals as input variables. Meanwhile, executives of the fuzzification for the inputs will be
in the second layer. Practically, this layer characterizes fuzzy quantization of the distorted signals and
Gaussian reference signals. During the process of learning, the system creates the (rules / knowledge),
it has learned in the form of if-then rules. In addition, the Membership Function node (MF) is used
in EFuNN to represent the fuzzy values. There are two neurons values represent fuzzy values of
different sizes, in which MF related with these neurons. In this layer, new neurons evolved together
with the input values. If the variable value does not belong to any of the existing MF, it means the
value is more than that of the membership threshold. Therefore, new fuzzy input neuron will be
created through the system by implementing the EFuNN steps. The main task of the fuzzy input nodes
is to transfer the distorted signal and Gaussian reference signal values into membership degrees to
which belong to the MF. The third layer consists of rule nodes that evolve through learning. Evolving
means all nodes in the third layer are created during training. Each one of the rule nodes is defined
by two connection vectors (weights); W1(r) from the fuzzy input layer to rule nodes and W2(r) from
rule nodes to the fuzzy output layer. These nodes represent the prototypes of input and output data
association. In this layer, a linear activation function or a Gaussian function will be used and the
weights will be adjusted. This layer represents the most important layer which determines the number
of rule nodes that will be created in order to make the output equalized signals as close as possible to
the Gaussian reference signals by minimizing the RMSE and lead to reduce ISI from the distorted
signals. The fourth layer is fuzzy output layer where each node is represented the fuzzy quantization
of the output variables. Finally, the output variable layer creates the defuzzification for fuzzy output
equalized signal. The output error (Err = || D D' || / Ni) should be less than an error threshold E,
where D is Gaussian reference signal, D' is shaped by EFuNN output equalized signal, Ni is the
number of outputs and E is the error threshold (tolerance) of the system for fuzzy output.EFuNN
300
(ICoIT'17)
10thof April, 2017.
equalizer parameters either be kept constant during training in equalization scheme, or adjusted
according to the subsequent incoming data. The determination of the network parameters like number
and types of MF for each input variable, sensitivity threshold, error threshold and the learning rates
were determined by the previous experiments. The parameter values for optimal performance have
been used as follows: Initial value for sensitivity threshold (radius of role nodes) S= 0.5, Error
threshold (The level of error tolerance for the output) E= 0.01, Number of membership functions= 5,
Rule extraction thresholds (weight of first layer) W1= 0.5, (weight of second layer) W2= 0.5, and the
consecutive data examples used= 55. The training process involves an iterative modification of the
biases of the network by reducing the (RMSE) of the distorted signals and Gaussian reference signals
in the EFuNN. FIGURE. 5 showcases the training for the five channels. Furthermore, in testing mode
in EFuNN equalization scheme in this research has tested the data EFuNN without Gaussian reference
signal and determined the bias according to the appropriate parameters that are set in the training
mode and was achieved the best desired signal with less ISI. The rules were extracted from the
training stage.
FIGURE 5. Training of five channels in EFuNN equalization in MDM
Results and Discussion

Referring to the collected data from the training process, the Performance of the EFuNN equalization
scheme (detailed in Table 1), is evaluated by the lower RMSE and CPU-time. The RMSE is
calculated using Equation (3).
1 n
RMSE
N i 1
( X i Yi )2
(3) [15]
X Y
In Equation (3), N is the number of data input samples, i is the distorted signal, and i is the
Gaussian reference signal. When the CPU-time is, the time taken for both the training and testing
stages of the equalization given pre-inserted distorted and measured channel impulse response
301
(ICoIT'17)
10thof April, 2017.
signals, measured in seconds. After analyzing the results, EFuNN equalization scheme illustrates for
fast adaptive learning and the lower RMSE values via the learning process. To evaluate the
performance of EFuNN equalization scheme, there is a need to compare the EFuNN equalization in
terms of RMSE and CPU-time with previous equalization schemes such as LMS and RLS which
have been used before in MDM system, (detailed in Tables 2 and 3), The goal of the comparison is
to show that the nonlinear equalization scheme can compensate the nonlinear problem of ISI in the
distorted signals in MDM system better than linear equalization schemes.
Table 2 RMSE and CPU-time for the EFuNN Equalization
Channel RMSE Before Training RMSE After Training CPU-Time/ s
Channel 1 - (LG 0 1) 0.3658 0.0002 0.000674
Channel 2 - (LG 1 0) 0.3246 0.0002 0.000671
Channel 3 - (LG 1 1) 0.2548 0.0605 0.000714
Channel 4 - (LG 2 1) 0.228 0.0002 0.000678
Channel 5 - (LG 1 2) 0.1949 0.0004 0.000734
Table 3 RMSE and CPU-time for LMS Equalization
Channel MSE MSE RMSE of MSE CPU-

time/s
Before equalization After equalization After equalization
Channel 1 - (LG 0 1) 0.1042 0.0350 0.1871 0.005545
Channel 2 - (LG 1 0) 0.0925 0.0481 0.2194 0.005428
Channel 3 - (LG 1 1) 0.1038 0.0572 0.2392 0.005012
Channel 4 - (LG 2 1) 0.0840 0.0262 0.1620 0.004627
Channel 5 - (LG 1 2) 0.1434 0.0073 0.0852 0.004585
302
(ICoIT'17)
10thof April, 2017.
Table 4 RMSE and CPU-time for RLS Equalization
Channel MSE before MSE after RMSE of MSE CPU-time

equalization equalization After equalization
Channel 1 - (LG 0 1) 0.1042 0.0616 0.2482 0.011242
Channel 2 - (LG 1 0) 0.0925 0.0666 0.2629 0.010120
Channel 3 - (LG 1 1) 0.1038 0.0669 0.2588 0.008881
Channel 4 - (LG 2 1) 0.0840 0.0367 0.1918 0.011353
Channel 5 - (LG 1 2) 0.1434 0.0501 0.2240 0.010503
4. Future work and Expected result

As coming to the end of this research, our expectation is to adopt a nonlinear fuzzy neural
network training algorithm utilized in MDM to overcome the problem of ISI for the first time.
This will overcome RLS, LMS equalization limitations.
REFERENCES
[1] S. R. Abdollahi, H. Al-Raweshidy, and T. Owens, Super-Broadband Wireless Access
Network: INTECH Open Access Publisher, 2012.
[2] C. V. N. I. (VNI). (2015). The Zettabyte EraTrends and Analysis. Available:

http://www.cisco.com/c/en/us/solutions/collateral/service-provider/visual-networking-
index-vni/VNI_Hyperconnectivity_WP.html
[3] N. Panchal, "Optical Fiber Communication System Theory or Analysis," International

Journal of Research, vol. 1, pp. 1878-1884, 2014.
[4] P. Prasad, "Investigation of Optical Multi Input Multi Output (OMIMO) transmission link
using Multimode Fiber," THAPAR UNIVERSITY, 1956.
[5] P. J. Winzer, "Scaling Optical Fiber Networks: Challenges and Solutions," Optics and
Photonics News, vol. 26, pp. 28-35, 2015.
[6] S. O. Arik, J. M. Kahn, and K.-P. Ho, "MIMO signal processing for mode-division
multiplexing: An overview of channel models and signal processing architectures," Signal
Processing Magazine, IEEE, vol. 31, pp. 25-34, 2014.
[7] Z. K. Adeyemo, A. A. Badrudeen, and R. O. Abolade, "Intersymbol Interference Distortion

Cancellation Using a Modified Maximal Ratio Combiner in Mobile Wireless
Communication," Journal of Information Engineering and Applications, vol. 3, pp. 16-25,
2013.
303
(ICoIT'17)
10thof April, 2017.
[8] H. Zhao, L. Zhang, B. Liu, Q. Zhang, Y. Wang, Q. Tian, et al., "MIMO signal processing
for mode
division multiplexing with RLSCMA algorithm," in Optical Communications and

Networks
(ICOCN), 2014 13th International Conference on, 2014, pp. 1-3.
[9] S. O. Arik, J. M. Kahn, and K.-P. Ho, "MIMO signal processing for mode-division
multiplexing: An
overview of channel models and signal processing architectures," Signal Processing

Magazine, IEEE,
vol. 31, pp. 25-34, 2014.
[10] A. Angela, "Evolving Fuzzy Neural Network Based Load Prediction Model In Dynamic
Cell Sizing,"
Multimedia University, 2003.
[11] K. Raveendranathan, Neuro-fuzzy equalizers for mobile cellular channels: CRC Press,
2013.
[12] P. Agrawal, R. Mehra, and M. Singh, "Implementation cost & performance analysis of
pulse shaping filter," in Green Computing and Internet of Things (ICGCIoT), 2015
International Conference on, 2015, pp. 1168-1172.
[13] A. Kang and V. Sharma, "Pulse Shape Filtering in Wireless Communication-A Critical
Analysis," IJACSA) International Journal of Advanced Computer Science and
Applications, vol. 2, 2011.
[11] S. Nayak, B. Misra, and H. Behera, "Impact of data normalization on stock index
forecasting," Int. J. Comp. Inf. Syst. Ind. Manag. Appl, vol. 6, pp. 357-369, 2014.
[15] R. R. Matlakunta, "Web traffic prediction for online advertising," Auckland University of
Technology, 2011.
304
(ICoIT'17)
10thof April, 2017.
Fast and Accurate Real Time Pedestrian Detection Using

Convolutional Neural Network
Hayder Albehadili
Department of Software , Kadhum College for Islamic Science University, Iraq
Sadqmo4@gmail.com
Laith Alzubaidi
Department of Systems and Applications, University of Information Technology &
Communications Baghdad, Iraq
Laithicci@gmail.com
Jabbar Rashed
Department of Electrical Engineering , Engineering College, University of Misan- Iraq
Dr.jabar72@gmail.com
Murtadha Al-Imam
Department of Software , Kadhum College for Islamic Science University, Iraq
Muam1987_computer@yahoo.com
Haider A. Alwzwazy
Department of Mathematics, Mathematics College, University of Misan- Iraq
Haiderabd83@gmail.com

Recently, pedestrian detection has become an important
Article History: problem of interest. Our work primarily depends on robust
Received: 20 March 2017 and fast deep neural network architectures. This paper used
very efficient and recent methods for pedestrian detection.
Recently, pedestrian detection has become an important
Published: 10 April 2017 problem of interest. This paper suggests robust convolutional
neural network models to solve this problem. We primarily
Keywords: evaluate accuracy and speed. Our work primarily depends on
Convolutional Neural robust and fast deep neural network architectures; substantial
Network, Pedestrian changes to those models achieve results that are competitive
Detection, Multiscale with prior state-of-the-art methods. As a result, we
input images outperformed all the prior state-of-the-art pedestrian
detection methods. We also overtook other models that use
extra information during testing and training. All experiments
used three pedestrian detection challenge benchmarks:
Caltech-USA, INRIA, and ETH.
1. INTRODUCTION
O bject detection is one of the most challenging tasks in computer vision. Recently,
pedestrian detection has become a topic of interest for many researchers [1, 2, 3, 4,
5, 6] because it is widely used in many real world applications such as safe driving
systems, security surveillance, and robotic navigation. A large number of difficulties such as
object appearance make pedestrian detection a challenging task. Different appearances can
305
(ICoIT'17)
10thof April, 2017.
result from different poses and environmental factors such as illumination. Pedestrian
detection becomes an even more challenging task when multiple persons are adjacent to each
other, leading to a bounding box that covers more than one pedestrian. Prior methods of
pedestrian detection can be categorized into two groups: hand-crafted or end-to-end feature
learning systems. For the first type, researchers first proposed different approaches such as
Integral Channel Features [7] and HoG [3] followed by trainable classifiers such as Support
Vector Machine SVM [8, 9], boosted classifiers [10], or random forests [11]. In contrast, a
convolutional neural network (CNN) is an example of an end-to-end feature extraction
system. Recently, CNNs have become a very effective and popular approach for pedestrian
detection [2, 12, 13]. CNNs have been used in variety of machine learning tasks including
image recognition [14,15, 16, 17, 18]. We summarize our contributions as follows: we adapt
contemporary CNN object detection architectures to pedestrian detection. We extensively
analyse and investigate two contemporary CNN architectures: the SPPnet architecture in [19]
and Fast Regions with CNN (F-RCNN) architecture. Our work principally depends on prior
work; however, we comprehensively study the suggested models and carefully choose their
parameters. Consequently, we consider the advantages of some prior existing methods and
propose a unified deep model. We achieve the best current performance on the Caltech-USA,
ETH, and INRIA benchmarks. We not only show that this model outperforms all existing
models but also demonstrate its performance as a real-time pedestrian detector. This
proposed CNN achieves results that are superior to all existing methods. Note that we always
follow the best parameter settings, as investigated in [20] (e.g. methods for generating
proposals such as Squares ChnFtrs, the size of input windows, and so on).
2. RELATED WORK
Because CNNs are used in our study, in this section, we briefly review all deep neural
network models used for pedestrian detection. A very efficient model was proposed in [23]
that achieves particularly good results. This model efficiently joins detection components
such as feature extraction, deformation, occlusion, and the final stage of classification. All
these parts are combined into one unified CNN and trained using back-propagation. The
model presented in [2] essentially focused on using ConvNet (proposed by [27]) with slight
changes, and the model is entirely trained in an unsupervised fashion. The main contribution
of ConvNet [12] is a classifier with hierarchical features. In this case, the classifier is fed
from two successive preceding layers simultaneously. This allows the classifier to use both
local and global features. One interesting work is JointDeep [23], which combines several
learning-stage pedestrian detection models to produce a unified deep model. All the stages
are trained simultaneously as a synergistic system. The system joins four main stages used
in pedestrian detection into one unified CNN: feature extraction, a PDM, occlusion, and
finally classification. A third implementation of the model was demonstrated in [16]. The
authors reviewed prior work and showed the effect of different factors on detection. In
addition, they showed the effect of combining different methods and achieved results that
were superior to previous work. In [15], a pedestrian detection model was proposed called
Switchable Deep Network [15], which achieves state-of-the-art results on the Caltech-USA
dataset with a 37.87% miss rate (MR). The main contribution of this system is the
hierarchical extraction of features. In other words, the model extracts features from different
parts of the body such as the head or shoulder. In [24], two methods to enhance the
performance of a CNN were proposed. The first method is a technique called Random
306
(ICoIT'17)
10thof April, 2017.
Dropout. In this method, the authors replaced the conventional Dropout method used by
CNN. Instead of using a fixed rate of dropout, they used a random dropout. Their second
method was named Ensemble Inference Network, which uses several fully connect layers
that each have different network topologies. Both methods were evaluated on the Caltech
Pedestrian Dataset and Daimler Mono datasets, achieving reasonable results. Recently, the
authors of [25] proposed different methods to generate foreground and background windows
to feed a CNN using Aggregated Channel Features. In [26], a multistage deep network
classifier called MultiSDP was proposed. This model has more than one stage of classifier.
Each stage is fed with contextual information. The model requires unsupervised pre-training
to avoid overfitting resulting from the multi-stage classifiers. Finally, the most recent work
of interest was proposed in [20]. The authors introduced a type of ConvNet called a vanilla
network, and performed extensive experiments and tested different parameters that can
influence the performance of a CNN. The authors showed that big ConvNets such as AlexNet
or small ConvNets such as CifarNet can achieve results that are competitive with prior
methods on the widely used Caltech-USA dataset. Although both SpatialPooling+ [21] and
Katamari [22] represent the state-of-the-art, these methods require additional information at
testing and training time, and the vanilla network achieved competitive results. Fig. 1 shows
the results of all abovementioned methods on both the ETH and Caltech-USA datasets. It is
clear that the vanilla network achieves results that are superior to all existing deep learning
methods on the Caltech-USA dataset. In addition, JointDeep [23] surpasses the other
methods on the ETH dataset.
Vanilla networks [20] achieve the state-of-the-art results over all existing deep neural
network approaches. Consequently, in this work, we enhance the performance of ConvNet
[12] and address the weakness inherited from vanilla networks by using a very recent
powerful deep neural network model. Different network architectures are demonstrated to
enhance pedestrian detection performance.
307
(ICoIT'17)
10thof April, 2017.
77
80 70 62.94
70 56.27
58.96 60
60 47.1 48
50 45
50 45
39 37.77 40
40
30
30 22.49 21.6
20 20
10 10
0 0
[39] [101] [104] [105] [106] [100] [107] [39] [101] [104] [106] [100]
FIGURE 1. Results of pedestrian detection using deep neural networks
In addition, we further show the MR results of all other prior strategies in Fig. 2. Note that
a lower MR is better. Optical flow is used by pedestrian detection strategies as additional
information during training and testing [21, 22].
The CNN state-of-the-art pedestrian detector [20] and most existing models that use CNN
follow the same training procedure, which is similar that of [18], which was originally used
for object detection. The steps for detection can be summarized as follows:
Warp or crop the proposals generated by the first step and feed them into the CNN
because CNNs require a fixed input size, as demonstrated in [18].
Train the CNN using candidate windows.
After training, use the features extracted from one of the last layers of the CNN to train
the SVM, which is used later for classifying extracted features from the last or
preceding layers of the CNN.
Note that the best performance achieved using a CNN follows the procedure above [20]. This
procedure has some drawbacks, as demonstrated in [27]. First, the training and testing phases
are very slow because it is necessary to extract features for both training and testing. Second,
because the input to the CNN is fixed, the model is trained with only a single input image
size [20]. Finally, the systems still have three dependent steps: proposal creation, training the
deep model, and classification using SVM, as shown in Fig. 3. These steps reduce both the
performance and speed. In addition, the CNNs used in [20, 18] accept a fixed image size of
227 227. Therefore, all candidate proposal windows must be cropped before they are fed
to the CNN. To summarize the limitations of the vanilla networks described in
308
(ICoIT'17)
10thof April, 2017.
Katamari [104] 0.224898
ACF-Caltech+ [29] 0.297591
LDCF [137] 0.247976
SpatialPooling+ [123] 0.218875
SpatialPooling [123] 0.292374
InformedHaar [43] 0.34598
WordChannels [38] 0.423007
RandForest [30] 0.511729
SDN [41] 0.378714
MultiSDP [37] 0.453934
JointDeep [40] 0.393187
Franken [32] 0.4868
pAUCBoost [24] 0.59656
ACF+SDt [42] 0.37341
ACF-Caltech [29] 0.442191
ACF [29] 0.513573
MOCO [36] 0.455319
MultiResC+2Ped [35] 0.434163
DBN-Mut [34] 0.482178
DBN-Isol [28] 0.53144
MT-DPM+Context [39] 0.376419
MT-DPM [39] 0.405447
MLS [21] 0.610336
AFS+Geo [18] 0.66759
AFS [18] 0.65382
ConvNet [13] 0.772001
MultiResC [33] 0.484515
FeatSynth [23] 0.601556
FPDW [25] 0.573991
ChnFtrs [26] 0.563396
LatSvm-V1 [12] 0.797807
HogLbp [17] 0.677692
HikSvm [15] 0.733935
MultiFtr+CSS [22] 0.60891
MultiFtr [16] 0.68263
Shapelet [139] 0.913718
FtrMine [14] 0.744201
HOG [1] 0.684556
VJ [9] 0.947337
a b
c
FIGURE.2 Results of pedestrian detection for all prior works
1. Only a fixed input image size is allowed.

2. Several exhaustive separate stages are required: proposal generation, deep model
training, and classification using SVM.
3. Only a single resolution is allowed for input images.
4. A large amount of time is consumed because of limitations (13).
309
(ICoIT'17)
10thof April, 2017.
We endeavor to enhance the performance and overcome the limitations inherent in vanilla
networks. Very efficient and fast deep models proposed in [19, 27] are used, and the potential
parameters that can influence performance are further investigated.
3. DEPLOYMENT OF DIFFERENT CNN ARCHITECTURES FOR

PEDESTRIAN DETECTION
Because our experiments use SPPnet and F-RCNN proposed in both [19] and [27],
respectively, they have crucial intrinsic differences. SPPnet is very similar to the methods
used in [20, 18]; however, SPPnet accepts different image sizes and keeps the same training
and testing procedures as in [20]. F-RCNN has more qualifications than SPPnet because it
has a unified CNN that shortens all three main steps mentioned earlier further details are in
Section 4.1
3.1. PIPELINE STEPS OF SPPNET FOR PEDESTRIAN DETECTION

As described earlier, we essentially use two predominant CNN paradigms. First, we use
SPPnet as described in [19], which is similar to the vanilla networks in [20]. The only
difference is that SPPnet is flexible and can take multi-scale input images for training or
testing. In this section, we explore the leverage of multi-scale input images. The pedestrian
detection pipeline shown in Fig. 3 consists of three main parts. In the first part, multi-scale
input images are prepared. In the next step, a large CNN is used to extract the features from
the input proposals. Finally, SVM is used on top of the CNN for classification. These steps
match most pedestrian detection CNN approaches, specifically the vanilla network, although
here, multi-scale input images are used.
a b c
Multi-size input CNN SVM classification

image
FIGURE 3. Pedestrian detection pipeline: a) multi-scale input images, b) the CNN, and c) final, fully connected layer
and SVM.
It is clear that this model is similar to [20] except for the first stage of multi-scale images.
Note that SPPnet can process whole input images or we can generate candidate proposals. In
this work, we use the second option. The following steps describe Fig. 3.
Step 1. Generated candidate windows. There are a variety of methods for generating
proposal windows. For fair comparison with [20], we followed the
methodsuggested by the authors for generating proposals.
Step 2. The generated category-independent windows are passed into the CNN for
training and testing purposes. Using the robust CNN architecture described by
Krizhevsky et al. [12], 4096-dimensional feature vectors are extracted from the
last layers using Caffe [28]. In contrast to [20], which uses RCNN [18], our CNN
can accept any size of input image because we use the SPPnet architecture [19].
310
(ICoIT'17)
10thof April, 2017.
In addition, the need for crop and warp proposal windows is eliminated using the
SPPnet approach [19]. Cropping or warping images is considered one of the main
drawbacks of RCNN because it takes a long time to warp images and loses
information. However, SPPnet can process entire images in a single step.
Step 3. Features are extracted from the CNN layers and the SVM is trained for use later
for testing.
4. EXPERIMENTAL SETUP
The model above was evaluated on three challenge benchmarks, the INRIA, Caltech-
USA, and ETH datasets. The next section briefly describes these datasets. For the
evaluation, we follow the conventions proposed by Dollar and Zitnick. [1].
4.1. CALTECH-USA DATASET
One of the most challenging datasets that we used is the Caltech-USA Pedestrian dataset
[1]. It contains numerous video clips (set00 to set11). The dataset has 250,000 frames.
All image frames are accompanied by annotations. The annotations number 350,000
bounding boxes for 2,300 unique pedestrians. All videos were taken by a driven vehicle. All
images have the same resolution of 640 480 pixels. The image quality is very poor
compared with the INIRA dataset. The general setting for the dataset is that the first six
datasets are used for the training and the other five sets are used for testing. In the standard
setting, sampling is chosen as one of thirty frames in both the training and testing sets.
However, it is possible to increase the training and testing samples by choosing different
numbers of frames to skip. We increased the training and testing sample rate by tenfold, as
described in [20].
4.2. INRIA DATASET
The INRIA dataset [3] contains very high resolution images. However, it is a small dataset
and has some incorrectly labeled images. The dataset has 614 positive frames and 1,218
negative frames for training. In addition, it has 288 negative images and 453 positive images
for testing. Overall, it has around 1,774 pedestrian annotations.
4.3. ETH dataset
In this dataset, all the data frames have a resolution of 640 480 pixels. This dataset has a different
density than the other datasets. We followed the author-recommended setting in [28].
5. COMPARISON OF THE ACCURACY OF THE PROPOSED METHOD AND VANILLA

NETWORKS
To evaluate the advantages of using multi-scale images for the deep model over the
traditional method used in vanilla networks [20], we used the AlexNet [12] model,
which is the same model as used in [20]. However, a SPP-net [19] was exploited to use
multi-scale input images. The experiments were conducted using the three benchmarks
listed above. After building the model described in 5.1, we conducted our experiments
without substantial changes to the model. We not only achieved results that were
competitive with all former CNN models, but the results were reasonable with respect
to other approaches that use additional information during testing such as optical flow.
Figs. 4(a), (b), and (c) show the results on the Caltech-USA, ETH, and INRIA datasets,
respectively. It is clear that SpatialPooling+ [21] and Katamari [22] are state-of-the-art
and outperform all the prior models. Furthermore, SpatialPooling and Katamari overtake
the vanilla network, which is the state-of-the-rt for pedestrian detection using deep
311
(ICoIT'17)
10thof April, 2017.
neural networks. The highest reported result on the Caltech-USA dataset using the
vanilla network is 23.3% MR [20].
FIGURE 4.SPPnet results using multi-scale input images: a) Caltech-USA, b) ETH, and c) INRIA
Using the SPPnet [19] technique without any significant changes, we achieved
accuracies of 22.21%, 40.49%, and 11.1% MR on the Caltech-USA, ETH, and INRIA
datasets, respectively. Note that we significantly outperformed vanilla networks. Thus,
we next investigated a more powerful CNN model called F-RCNN [18] for pedestrian
detection.
6. WEAKNESSES OF SPPNET
Although SPPnet achieves competitive results to the state-of-the-art pedestrian
detection using multi-scale input images, it still has the following shortcomings: 1) we
still need to perform the four training and testing stages described in Section 4 and 2)
fine-tuning can be accomplished only for the last two fully connected layers [27]. Thus,
to alleviate these limitations, a more sophisticated model is used. In this section, we use
the F-RCNN model proposed in [27] for pedestrian detection. As stated earlier, F-RCNN
has a more powerful architecture than all the CNNs mentioned in [18, 19, 20] for the
following reasons:
It is a unified deep model that enfolds more steps in a single CNN and shortens all
four stages of CNN training and testing. For instance, feature extraction, cropping
candidates, and SVM training and testing are all encapsulated in a single unified
CNN. Moreover, it is much faster than all previously mentioned models because
F-RCNN does not extract features and store them in a database; instead, all
proposals are fed into the CNN with the original images so they can project the
proposals into the images. F-RCNN can also propagate error back-to-back, in
contrast to SPPnet, which can propagate error only to the last max-pooling layer.
Therefore, all layers of the CNN are tuneable, which gives it a greater ability to
tune the model for pedestrian detection.
There are more features recognized in F-RCNN than in other models.
Experiments were also conducted on the three pedestrian detection datasets. Again
without any substantial changes, F-RCNN achieved 21.83%, 38.18%, and 10.9% MR
on the Caltech-USA, ETH, and INRIA datasets, respectively as shown in figure 5,6,7.
312
(ICoIT'17)
10thof April, 2017.
The results are shown in Fig. 5. Intuitively, F-RCNN is clearly superior to SPPnet. Thus,
we designed a more effective model based on F-RCNN.
FIGURE5: F-RCCN results using multi-scale input images,Caltech-USA dataset
FIGURE 6: F-RCCN results using multi-scale input images,ETH dataset
313
(ICoIT'17)
10thof April, 2017.
FIGURE 7: F-RCCN results using multi-scale input images, INRIA dataset
7. CONCLUSION
In this work, we used very efficient and recent methods for pedestrian detection. We
evaluated both SPPnet and F-RCNN, which were originally used for object detection.
We showed that SPPnet performs better than conventional CNNs, which are used in
prior pedestrian detection methods [20]. Using SPPnet without substantial changes, we
achieved better accuracy than the best-designed model using deep neutral networks. In
addition, we accomplished results that are competitive with other models that do not use
CNN. Motivated by the results from SPPnet, we used a leveraging CNN called F-RCNN
that is more efficient than SPPnet. As a result, we outperformed all the prior state-of-
the-art pedestrian detection methods. We also overtook other models that use extra
information during testing and training.
8. REFERENCES
[1] [1] P. Dollar, C. Wojek, B. Schiele, and P. Perona. Pedestrian detection: an
evaluation of the state of the art. TPAMI, 34(4):743761, 2012.
[2] [2] P. Sermanet, K. Kavukcuoglu, S. C. Pedestrian, and Y. Le-Cun.Pedestrian
detection with unsupervised multi-stage feature learning.CVPR, 2013.
[3] [3] N. Dalal and B. Triggs. Histograms of oriented gradients for human
detection.In CVPR, 2005. 1, 2, 5, 7
[4] [4] O. Tuzel, F. Porikli, and P. Meer. Pedestrian detection via classifcation on
riemannian manifolds. IEEE Trans. PAMI, 30(10):17131727, Oct. 2008. 1, 2
[5] [5] C. Mikolajczyk, C. Schmid, and A. Zisserman. Human detection based on a
probabilistic assembly of robust part detectors. In ECCV, 2004.
314
(ICoIT'17)
10thof April, 2017.
[6] [6] P. Felzenszwalb, R. Girshick, D. McAllester, and D. Ramanan. Object

detection with discriminatively trained part based models. In PAMI 2010. 1
[7] [7] P. Dollar, Z. Tu, P. Perona, and S. Belongie, Integral Channel Features Proc.
British Machine Vision Conf., 2009.
[8] [ 8] M. Bertozzi, A. Broggi, M. Del Rose, M. Felisa, A. Rakotomamonjy and F.
Suard A Pedestrian Detector Using Histograms of Oriented Gradients and a
Support Vector Machine Classifier Proceedings of the 2007 IEEE Intelligent
Transportation Systems Conference Seattle, WA, USA, Sept. 30 - Oct. 3, 2007
[9] [ 9] Reecha P. Yadav, Vinuchackravarthy, Sunita P. Ugale Implementation of
Robust HOG-SVM based Pedestrian Classification International Journal of
Computer Applications (0975 8887) Volume 114 No. 19, March 2015
[10] [10] JiaolongXu, David Vazquez, Sebastian Ramos, Antonio M. Lopez , and
Daniel Ponsa Adapting a Pedestrian Detector by Boosting LDA Exemplar
Classifiers CVPR 2013
[11] [11] [1] J.Gall and V.Lempitsky. Class-specific hough forests for object detection.
In CVPR, 2009.
[12] [ 12] A. Krizhevsky, I. Sutskever, and G. E. Hinton. Imagenet classification with
deep convolutional neural networks.In NIPS 2012: Neural Information Processing
Systems. 1
[13] [ 13] P. Luo, Y. Tian, X. Wang, and X. Tang. Switchable deep network for
pedestrian detection.In CVPR, 2014. 2, 5
[14] [14] 61. Yin, X.C., Yin, X., Huang, K.: Robust text detection in natural scene
images. CoRR abs/1301.2628 (2013)
[15] [15] Min Lin, Qiang Chen, and Shuicheng Yan Network In Network arXiv
1312.4400v3,4 Mar 2014
[16] [16] Dan Ciresan, Ueli Meier and JurgenSchmidhuber, Multi-column Deep
Neural Networks for Image Classification CVPR 2012
[17] [17] JulienMairal, PiotrKoniusz, ZaidHarchaoui, and CordeliaSchmid,
Convolutional Kernel Networks arXiv 14 Nov 2014 .
[18] [ 18] Ross Girshick Jeff Donahue Trevor Darrell Jitendra Malik Rich feature
hierarchies for accurate object detection and semantic segmentation
arXiv:1311.2524v5 [cs.CV] 22 Oct 2014
[19] [19] Kaiming, He and Xiangyu, Zhang and Shaoqing, Ren and Jian Sun Spatial
pyramid pooling in deep convolutional networks for visual recognition European
Conference on Computer Vision, 2014
[20] [20] Jan Hosang Mohamed Omran Rodrigo BenensonBerntSchiele, Taking a
Deeper Look at Pedestrians arXiv:1501.05790v1 [cs.CV] 23 Jan 2015
[21] [ 21] S. Paisitkriangkrai, C. Shen, and A. van den Hengel. Strengthening the
effectiveness of pedestrian detection with spatially pooled features.In ECCV,
2014. 1, 2, 4, 7
[22] [ 22] Rodrigo Benenson Mohamed Omran Jan HosangBerntSchiele, Ten Years
of Pedestrian Detection,What Have We Learned?
315
(ICoIT'17)
10thof April, 2017.
[23] [23] W. Ouyang and X. Wang.Joint deep learning for pedestrian detection.In
ICCV, 2013. 2, 4, 5
[24] [24] Hiroshi Fukui Takayoshi Yamashita, Yuji Yamauchi, Hironobu Fujiyoshi,
Hiroshi Murase2 Pedestrian Detection Based on Deep Convolutional Neural
Network with Ensemble Inference Network
[25] [25] Xiaogang Chen, Pengxu Wei, Wei Ke, Qixiang Ye, Jianbin Jiao Pedestrian
Detection with Deep Convolutional Neural Network
[26] [26] X. Zeng, W. Ouyang, and X. Wang.Multi-stage contextual deep learning for
pedestrian detection.In ICCV, 2013. 2
[27] [27] Ross Girshick, Fast R-CNN arXiv preprint arXiv:1504.08083, 2015
[28] Y. Jia, Caffe: An open source convolutional architecture for fast feature
embedding, http://caffe.berkeleyvision.org/, 2013.
316
(ICoIT'17)
10thof April, 2017.
FINDING SHORTEST PATH BASING ON DIJKSTRA'S

ALGORITHM FOR BAGHDAD UNIVERSITIES BY USING
GEOGRAPHIC INFORMATION SYSTEM APPLICATIONS
Roaa Ebada Saeed
Amer Saleem Elameer
Rafah Shahab Alhamdani

As the population increases, the transport network becomes
Article History: difficult and enormous. Finding demanded location is becomes a
Received: 20 March 2017 troublesome task. After finding, a location individual gets confused
Accepted: 1 April 2017 to access that location because of the routes contains several modes.
Published: 10 April 2017 This drawback is even more substantial for those who may have to
Keywords: Route go to strange parts of the city. Just in case of the universities, it is
analysis, GIS troublesome to find the specified university and it's the shortest
applications, Service path to access therefore takes longer to access it. In this paper
area analysis, and solved this problem by find the shortest path from the user location
shortest path analysis. to the university or between two universities. The area for study is
Baghdad city and the coordinate for universities is obtained from
Google earth program. The ArcGIS software system used and also
the Dijkstras algorithm to produce the shortest path from one
location to a different and ArcGIS cloud publisher for publish map
from ArcMp to ArcGIS cloud and then can be edited and updated
1. INTRODUCTION
U
NESCO ask the Iraqi ministries of Education and Higher Education to build
educational atlases for the Iraqi Education Systems. The ministry of education
have its own Atlas system but in the Iraqi higher education, there is no such atlas
system. This research is a part from the Atlas system, which we start to build it for the Iraqi
higher education and the first step of the atlas is to complete the system for Baghdad
universities only. UNESCO requirements for this atlas system are big and need huge quantity
of data, but we have fix only some of the required data to start with. The Atlas system need
for a special geographical system to be completed.
A geographic information system (GIS) may be a computer system designed to
capture, save, treatment, analyze, manage, and display all kinds of spatial or geographical
data [1]. Geographic information system (GIS) is a robust tool and has the capability to
handle and treatment spatial data in a massive volume. GIS becomes a common technique
that may produce maps, integrate info, visualize and solve issues, and develop valuable
solutions [2].
Network analysis remains one in all the foremost important and protracted analysis and
application areas in the geographic data system. Networks can perform an alternative datum
317
(ICoIT'17)
10thof April, 2017.
for locations in the context of linear referencing and support a set of tools for the graphical
present. Many network location problems are the most troublesome to solve in terms of their
combinatorial complication. The shortest path analysis is not only one of the major network
operations in GIS analysis, transportation manufacture and many other fields but also the
foundation of other network analysis, such as the most credible path problem, maximum
capacity path problem etc. The shortest path problem is a problem of finding the shortest
path or route from one point to another [3]. In this study describes how network analysis based on
Dijkstra algorithm in ArcGIS can be found
Shortest path for Baghdad universities.
2. STUDY AREA BAGHDAD CITY

Baghdad is that the capital metropolis and a key cultural, instructional and economic
center of Iraq. Geographically the metropolis is found right within the center of the country,
and within the finish of the last century, the metropolis has been one in all the foremost
developed big cities within the Middle East, with terribly advanced infrastructure, schools
and universities, stadiums and different facilities for sports. Geographic coordinates of
Baghdad, Iraq (Latitude: 332026 N, Longitude: 442403 E) [4].
3. LITERATURE REVIEWS
In route optimization of the school Busses advancement of the school transport and
decrease of the transportation cost is potential by using the system dissection based mostly
requisitions. numerous inquiries concerning have as of currently been Performed known with
course improvement of school transports. various school overseers have perceived a
substantial backlog once the route and calendars are overseen utilizing GIS. GIS
programmings improve the abilities of school sheets to transfer a productive and effective
transportation framework [5][6].
4.METHODOLOGY
Figure 1: Shows the steps of the methodology that proposed for this work
318
(ICoIT'17)
10thof April, 2017.
Figure 1: Methodology for proposed work
A. Data Collection
1. Base Map
It is the primary and necessary step towards the completion of the project. For this work,
Bagdad city (332026 N and 442403 E) is taken into account because the study space
that is located within the Iraq. Top sheet of Bagdad city is obtained from the Google earth
application that is taken into account because of the base map for future work(Fig2).
Figure 2: Base map of Baghdad city
319
(ICoIT'17)
10thof April, 2017.
2. Universities Data
The university information will be obtained by visiting the required university and taking
the necessary data concerning with the staff academic, relations, projects, researches and
university ranking. The attribute data like the name of universities, email of universities,
number of colleges, website address etc., can be obtained by visiting the website for each
university. The data needed is in the form of spatial and attribute data. This spatial data is
obtained through the method of digitizing base map of the Bagdad city. The coordinates of
the universities were taken by using Google earth program. For universities spatial data, it
should be connected to the information services provided at every university. Attribute data
that require being saved within the database are universities name, Longitude, and Latitude
as shown in (Table 1).
Table1: University information
B. Geo-referencing
Geo-referencing of base map is an essential stage, as a result of geo-referencing permits
to register base map with respect to the earths surface. Geo-referencing is often done by
choosing the four management points on the corners of the scanned map and inserting that
point by giving acceptable latitude and longitude [1].
C .Mapping of Universities
In this step, the coordinates of the university are taken by collected from Google earth
software.
5. IMPLEMENTATION
The software used to implement this project:
ArcGIS
Dijkstras algorithm
320
(ICoIT'17)
10thof April, 2017.
ArcGIS cloud publisher

ArcGIS cloud
After the information gathering, it is necessary to locate it on the map. The university
location is executed using the ArcGIS tool. Coordinates the universities are put on the map
as a point feature and this point can be represented as a symbol. When university location is
completed, it is necessary to locate a road map. It has as well done using the ArcGIS. When
making a network routing system, all the roads at the network are connected, is vital as a
result of it permits connecting through the system.
ArcGIS Network Analyst could be a powerful tool of ArcGIS that supply network-based
special analysis together with routing, travel directions, closest facility, and service area
analysis.
The Dijkstras algorithm use for the network analysis and finding the shortest path. ArcGIS
Network Analyst permits to unravel common network issues, like finding the most effective
route across a metropolis, distinguishing a place around a location, service a collection of
orders with a fleet of vehicles, or selecting the most effective facilities to open or close[7].
The commercial GIS software that was used to carry out the route analysis for this research
is Esri ArcGIS Network Analyst. ArcGIS is appropriate for this type of research as a result
of its commercially obtainable, and therefore the Network Analyst extension is contained
within the student edition of ArcGIS. The route problem solver in Network Analyst to locate
the shortest path is depended on Dijkstra's algorithm [8].
Dijkstra's algorithm is a graph search algorithm that solves the only supply shortest path
drawback for a graph with non-negative edge path value, generates the shortest path tree.
This algorithm is commonly utilized in routing and as a subroutine in different graph
algorithms. For a given supply vertex (node) within the graph, the algorithm finds the route
with the lowest value (i.e. the shortest path) between that vertex and each different vertex[9].
5.1. Dijkstra' algorithm
321
(ICoIT'17)
10thof April, 2017.
6. EXPERIMENTAL RESULT
6.1. Creation of Map for Universities
geo-referencing for the base map can be created by choosing the four management points
on the corners of the base map and inserted that point by giving acceptable latitude and
longitude, and then inserted the coordinates of the university are taken by collected from
Google earth software as shown in Figure( 3). Finally located road on the map as shown in
figure (4).
Figure 3: Geo-referencing and coordinate base map
Figure 4: Located road on the base map
322
(ICoIT'17)
10thof April, 2017.
6.2. Shortest path analysis

The shortest path created by network analysis for finding the minimum distance path
from the user location to the university or between two universities this will help the user to
minimize traveling time to access a specified university. Figure(5) shows the shortest route
from Al-Nahrain University to University of Baghdad and from University of Baghdad to
Baghdad University for Women.
Figure 5: shortest path analysis

6.3. Publish map on the ArcGIS cloud
GIS Cloud Publisher for ArcMap extension used to publish map from ArcMap to GIS
Cloud mechanically uploads data, symbology, layer structure and spatial references As
Shown in figure (6).The idea is that what see in desktop GIS is immediately replicated on
GIS Cloud account.GIS Cloud Publisher enable to publish a map to the general public or
included into website/blog while not a desire for having own servers. After published map
on ArcGIS cloud, map it's edited and updated the maps. as shown in figure(7).
323
(ICoIT'17)
10thof April, 2017.
Figure 6: Using GIS cloud publisher
7. CONCLUSIONS AND FUTURE WORK

This study for find shortest path from the user location to the University for Help the
user to minimize traveling time to access a specified university. This can be done by using
ArcGIS and Dijkstra algorithm that calculate distant between two points depended on route
length. And use GIS cloud publisher for publish the map on the ArcGIS cloud and then can
be edit and update and modify.
As a future work, it can be performed measure distance between two universities or
between user location and university; this can be accomplished by ArcGIS cloud.
8.REFERENCES
[1] Dabhade, Amrapali C., and K. V. Kale, GIS Based Health Care Information System for
Aurangabad City. International Journal of Engineering and Innovative Technology, 2014.
4(1).
[2] Bhanumurthy, V., et al. , Route analysis for decision support system in emergency
management through GIS technologies. Int. J. Adv. Eng. Glob. Technol, 2015. 3(2):P. 345-
350.
324
(ICoIT'17)
10thof April, 2017.
[3] Zhang, Jiyi, et al. , Shortest path algorithm in GIS network analysis based on Clifford
algebra. Proceedings of The 2nd International Conference on Future Computer and
Communication. Wuhan: Wuhan University. 2010.
[4] Al Obaidy, Abdul Hameed MJ, and Athmar AM Al Mashhadi. Heavy metal
contaminations in urban soil within Baghdad City. JORNAL OF ENVIRONMENTAL
PROTECTION, 2013. 4:72-82
[5] Belcher, Johnny, et al. Bus Routing Algorithms: Application to a Rural School District.
Working Paper No. 27." Appalachian Collaborative Center for Learning, Assessment, and
Instruction in Mathematics (ACCLAIM), Ohio University (2005).
[6] Yomralioglu, T. "Cografi Bilgi Sistemleri Temel Kavramlar ve Uygulamalar
(Geographic Information Systems Basic Concepts and Applications)." Karadeniz Technical
University, Trabzon (2002).
[7] Kai, Ni, Zhang Yao-ting, and Ma Yue-peng. Shortest Path Analysis Based on Dijkstra's
Algorithm in Emergency Response System. Indonesian Journal of Electrical Engineering and
Computer Science, 2014. 12(5):P. 3476-3482.
[8] Winn, Michael T. A Road Network Shortest Path Analysis: Applying Time-varying
Travel-time Costs for Emergency Response Vehicle Routing, Davis County, Utah: a Thesis
Presented to the Department of Humanities and Social Sciences in Candidacy for the Degree
of Master of Science. 2014.
[9] Gupta, Nitin, et al. , Applying Dijkstras Algorithm in Routing Process. International
Journal of New Technology and Research (IJNTR), 2016. 2(5):P.122-124.
325
(ICoIT'17)
10thof April, 2017.
Image Encryption Based on Partitioning

and Rearranged Pixels Position
Dr. Jamila H. Al-Ameri

Computer Science Dept., College of Science, AL-Mustansiriyah University - Iraq
Dr.jameelahharbi@uomustansiriyah.edu.iq
Salah T. Allawi
Computer Science Dept., College of Science, AL-Mustansiriyah University Iraq
Salahta1970@gmail.com
May M. Abbas
Ministry of Higher Education and Scientific Research, Legal and Administrative Directorate Iraq
May.M.Altaie@gmail.com
ARTICLE INFO
ABSTRACT
Article History: Protect important information transmitted via the channels of

Received: 9 March 2017 communication it's one of the important area. Today,
Accepted: 1 April 2017 especially with the increasing use of the Internet in the
Published: 10 April 2017 process of transferring this information you had to find ways
Keywords: to protect this information for fear of being captured by
Data protection, unauthorized people to look at it. The image and what it
Image encryption, contains is one of the fore mentioned knowledge that should
Image Decryption, be protected. In our paper a new method is suggested for
Random number , protecting information of colored image by rearranging the
LFSR. distribution of image pixels. This way depends on the image
partition into four equal parts where each data stored in a 1-D
array, then the data of these parts are re-arranged through
random numbers generated by using LFSR to produce a new
(2-D) array combines the data of the four-part. Finally, the
new array is converted to another color space YCbCR and the
result will be representing the stego image.
1. INTRODUCTION
I
nternet is used to get quicker transmission of large amount of essential and important
data, because internet contains numeral points of attack, it faces a large number of
attacking types, to avoid unauthorized access this information should be protected,
thus many data protection methods like (Encryption, Watermarking, Masking Data,
etc. ) are done[1]. Protecting digital images has become very important because of the
Internet fastness in the digital world at this time. Protection of digital images drawn a lot of
interest nowadays, and several image encryption techniques suggested to make better safety
of images [2]. Where nobody can read encrypted text without decrypting. The changing
method of the modified data into its native plaintext is defined Decryption, for it can be read
easily [3]. Algorithm of encryption image is divided into three groups [4]:
326
(ICoIT'17)
10thof April, 2017.
1- Visual transformation based algorithm,

2- Position permutation based algorithm,
3- Value transformation based algorithm.
This paper consists of the following parts:
Part 1- Provide general information about the subject of image encryption.
Part 2- View previous research that has been done within this topic.
Part 3- Review the way it has been used in this research.
Part 4- Explain some of the points that have been got out from work.
2. LITERATURE SURVEY
M. A. B. Younes and A. Jantan [5]. Proposed Blowfish forward algorithm depending on
the Selection and choosing of image transformation and a well-defined encryption and
decryption. In their algorithm is divided main image into groups that reconstructed into a
changed image. Blowfish algorithm is used to encrypt transformed image. The correlation
was decreased between image elements by using the suggested method.
Y. Rajput and A, K. Gulve [6]. are suggested method to make better a secure image can
be obtain from the scheme encrypt. The suggested method is classified into 3 steps. Firstly,
the single digit number from the original image calculated if it can be partitioned. Secondly,
bit rotation, reversal & for each block of the image is done. Thirdly, the extended hill cipher
method is used upon an output of second step. The receiver can be generating the original
image if he has suitable decryption key.
A. Gupta, N. Tiwari, M. Chawla and M. Shandilya [7], are introduce image encryption
methods which gather the idea of block based on pixel treatment and transformation. The
suggested technique contains two phases:
1- Block based matrix transformation for pixel place treatment was applied.
2- Applying a bit moving method which alters the data of every pixel.
S. P. Nichat and S. S. Sikchi [8], are introduced a hybrid model for image encryption
gathers chaotic function and genetic algorithm. Their technique is contains two phases:
1- Secure images is built using chaotic function and secret key.
2- Using the result images as seed value for genetic algorithm.
In their suggested way used genetic algorithm to have a better result. The best cipher
image is chosen on finding the correlation entropy and coefficient. Also the best cipher image
in case of having highest entropy and lowest correlation coefficient.
3. Proposed Method
The proposed Image Encryption Based on Partitioning and Rearranged Pixels Position
(IEBPRPP) method is suggested to protect information of colored image by rearranging the
distribution of image pixels. The framework of IEBPRPP method is classified into two parts;
image encryption and image decryption as it will be shown below:
A- Encryption Part
The IEBPRPP method is illustrated in Fig.1. First step includes entering cover image
(RGB) used in the IEBPRPP method. Second step: partitioning the cover image into four
equal parts, and then converting the pixels value of each part into new 1-D array.
327
(ICoIT'17)
10thof April, 2017.
Enter RGB image
spilt image into four equal parts

part_size = (high /2)*(width /2)
Generate
Random Seed value
Register1= high
Number Register2= width
Register3= image size
Count= 0
Generate random number
Count = count +1
No
Count = part_size
Yes
Reorder Using random number to
Pixels reorder pixel position
Position
Count = count -1
No
Count = 0
Yes
Convert result image into

YCbCr color space
Save image
FIGURE1. Proposed method outline (Encryption Stage)
Third step, is generate a set of random numbers by using LFSR. The length of
sequence bits that will be generated by using LFSR and then convert it to random number
depending on the size of any part of cover image (part image =(high/2)*(width/2)) therefore
this length will be changed with each cover image that has a new size. LFSR consists of three
registers that have different primary lengths (29, 31, 37) respectively with different
connecting function. Initial values of registers based on the original image (height, width and
size of image). These numbers are stored in a new 1-D array where the size is equal to any
part of sub square partition see Fig.2.
328
(ICoIT'17)
10thof April, 2017.
Random
No.
4
1
3
:
:
FIGURE 2. Generate random number
Each number indicates the pixel location that will be called from the four respectively
to be rearranged in a new array, whose size is equal to the size of the main image see fig. 3.
Fourth step, after the rearrangement of all the pixels complete the result image will be
converting to another color space (YCbCr) and then the result represent encrypted image.
Random
No. P1 P2 P3 P4 Image Encryption
No.
4 1 120 33 231 210 144 167 150 188 120 33
1 2 200 123 220 213 231 210 136 52
3 3 136 52 158 240
5 4 144 167 150 188
2 5 80 189 158 160
FIGURE 3. Process of rearrangement image pixels
The algorithm is illustrating the main steps of encryption image:

1. Enter color Image (RGB).
2. Partitioning the image into four equal parts.
3. Constructing four 1-D arrays to store pixels value of the parts.
4. Generating random numbers by using LFSRs.
5. Depending on the random number, reordering pixels position for all parts.
6. Constructing a new 2-D array to store pixels value after reordering (encrypted image).
7. Convert the encrypted image into another color space YCbCr.
8. Save the Image encrypted.
B- Decryption Part
Decryption part of IEBPRPP method is consists of four steps: First step: input the
encrypted image (YCbCr). Second step: conversion the encrypted image from YCbCR to
RGB color space. Third step: constructing 2-D array to store the pixels value of encrypted
image. Fourth step: using the same initial values in encryption side to produce random
numbers. Then reconstructing four equal arrays their sizes are equal to quarter part of the
whole image. Relying upon the random numbers, the pixels are rearranged into their original
positions in the four arrays. Finally the image has been reconstructed from the four parts that
were obtained, see fig.4.
329
(ICoIT'17)
10thof April, 2017.
Enter YCbCr image
Convert YCbCr into RGB
Convert image pixels value

into 2-D array
Generate the same random

number
Reorder image pixels position

in Four 1-D array
Reconstruct original image

from the four 1-D array
Save image
FIGURE4. Proposed method outline (Decryption Stage)
The algorithm is illustrating the main steps of decryption image:

1. Entering encrypted image (YCbCr).
2. Convert image from (YCbCr) into RGB
3. Constructing new 2-D arrays to store all the pixel values of the encrypted image.
4. Generating the same random numbers by using LFSRs
5. Using the random number to getting back the data of the four 1-D arrays.
6. Reconstructing the original image from the four 1-D arrays.
7. Save the result image.
4. Experimental Results
The proposed method is implemented on color images. Figs. (5 and 6) are illustrated the
result of IEBPRPP method (encrypted and decrypted). The result of the images are analyzed
the performance by using some types of visual observation.
a) Original image b) Spilt image c) Encrypted image d) Stego image

Figure 5. Encrypted image
330
(ICoIT'17)
10thof April, 2017.
a) stego image b) Encrypted image c) reconstruct image d) original image

Figure 6. Decrypted image
4.1. Number of Pixel Change Rate (NPCR):

NPCR is known as the different value of pixels of original and encrypted images. By
regarding two images P(i, j) and C(i, j), an array X(i, j) means X(i, j) = 0, if P(i, j) = C(i, j),
else X(i, j) = 1. If the output is equal to zero both images are same otherwise equal to one
[9]. The value of NPCR is calculated by using the mathematical methods found in the eq. 1.
Where: W and H are the image width and height.

Encrypted and original images gave NPCR value as shown in table 1. Through applying the
proposed method it is clear that the obtained NPCR values are different from one image to
another.
TABLE1. Showing the difference of NPCR value
Encryption method NPCR Value %

Suggest method 100
R. Venkatesan [9] 99.5468
K.Bhoopathy Bagan [10] 98.4754
Jean-Yves Chouinard [ 11] 99.5850
C.W. Liao [12] 99.5400
4.2. Unified Average Changing Intensity (UACI)

The (UACI) shows the rate of intensity variation of original and encrypted images.
NPCR focuses on the unlimited value of pixels that alters rate in different attacks, as the
UACI optimal value UACI is around 33% [9]. The UACI is computed by using the eq. 2.
Where: p(i,j) and c(i,j) are two images.

H and W the height and width of the images.
Table2 shows The UACI value of the encrypted and main images and a compares with the
ways which are already found.
331
(ICoIT'17)
10thof April, 2017.
TABLE2. The comparison value of UACI
Encryption method UACI Value %

Suggest method 24.3333
R. Venkatesan [9] 28.8065
K.Bhoopathy Bagan [10] 32.2128
Jean-Yves Chouinard [ 11] 28.6210
C.W. Liao [12] 28.2700
4.3. Histogram Analysis

Graphical show of the many times of pixels in an image likes a function of their intensity
values which is known histogram image. The degree for the variation of inspecting between
encrypted and main images visually at pixel level is done by Image histogram is a. Fig 7
shows The histogram of the result of process image encrypted fig.5.
A- Red color before encrypted B- Green color before encrypted C- Blue color before encrypted
D- Red color after encrypted

FIGURE7. E- Green
Difference histogram of 3color after
bands encrypted
between original andF-encryption
Blue color image
after encrypted
4.4. Noise Attack Analysis

Information transmitted over the internet is subjected to attempts by the attackers or
unauthorized persons to destroy them, so that the receiver cannot understand the information
sent to him after decrypted it. There is several ways to destroy this information such as:
1- Add noise
2- Cut of part of the information
In IEBPRPP method, the way of adding noise or cutting parts of the image, do not affect
the clarity of the image because this noise will be distributed randomly on the image after
the decryption process. This noise can be removed by using medium filter. Fig.8 illustrates
adding noise and cropping parts of encryption image and the result after its decryption.
332
(ICoIT'17)
10thof April, 2017.
A- Add noise B- Decryption Image C-Medium filter Image
D- Cutting parts E- Decryption Image F- Medium filter Image
FIGURE8. Add noise , decrypted image and applying medium filter
5. Conclusions
A new suggested IEBPRPP method is introduced to protect the information in color
images in this paper. This method depends on fragmentation the image into four parts and
rearranging pixels position through generated random numbers by using LFSRs. The initial
values are taken from image data which cannot be changed if exposed for compression
process. The numbers of random numbers that are generated represent 1/4 of the size of the
image. One of the benefits of this method the result image can be sent to the receiver with
JPEG form and the receiver can recover the original image. Another benefit is not sending
any additional information with the encrypted image.
REFRENCES
[1] J. Shah, and Dr. V. Saxena Performance Study on Image Encryption Schemes" , IJCSI
International Journal of Computer Science Issues, Vol. 8, Issue 4, No 1, July 2011, p 349-
355
[2] V. V. Divya, S. K. Sudha, and V. R. Resmy "Simple and Secure Image Encryption",
IJCSI International Journal of Computer Science Issues, Vol. 9, Issue 6, No 3, November
2012, p 286- 289
[3] R. Pakshwar, V. K. Trivedi, and V. Richhariya " A Survey On Different Image
Encryption and Decryption Techniques", IJCSIT International Journal of Computer
Science and Information Technologies, Vol. 4 (1) , 2013, p 113 116
[4] K. D. Patel, and S. Belani " Image Encryption Using Different Techniques: A Review ",
International Journal of Emerging Technology and Advanced Engineering , Volume 1,
Issue 1, November 2011, p 30-34
[5] M. A. B. Younes, and A. Jantan "Image Encryption Using Block-Based Transformation
Algorithm", IAENG International Journal of Computer Science, 35:1, IJCS_35_1_03,
(Advance online publication: 19 February 2008
333
(ICoIT'17)
10thof April, 2017.
[6] Y. Rajput and A K. Gulve " An Improved Cryptographic Technique to Encrypt Images
using Extended Hill Cipher", International Journal of Computer Applications, Volume 83
No 13, December 2013, p 4-8.
[7] A. Gupta, N. Tiwari, M. Chawla and M. Shandilya " An Image Encryption using Block
based Transformation and Bit Rotation Technique", International Journal of Computer
Applications, Volume 98 No.6, July 2014, p 30-32.
[8] S. P. Nichat, and Prof. Mrs. S. S. Sikchi " Image Encryption using Hybrid Genetic
Algorithm", International Journal of Advanced Research in Computer Science and
Software Engineering, Volume 3, Issue 1, January 2013, p 427- 431.
[9] T. Sivakumar, and R. Venkatesan " Image Encryption Based on Pixel Shuffling and
Random Key Stream", International Journal of Computer and Information Technology,
Volume 03 Issue 06, November 2014, p 1468- 1476.
[10] G. A. Sathishkumar and K. B. Bagan, A novel image encryption algorithm using pixel
shuffling and BASE 64 encoding based chaotic block cipher, WSEAS Transactions on
Computers, Vol. 10, No. 6, pp169-178, 2011.
[11] K. Loukhaoukha, J.Y. Chouinard, and A. Berdai, A secure image encryption algorithm
based on Rubik's cube principle, Journal of Electrical and Computer Engineering, Vol.
20, No. 12, pp113, 2011.
[12] C. K. Huang, C. W. Liao, S. L. Hsu, and Y. C. Jeng, Implementation of gray image
encryption with pixel shuffling and gray-level encryption by single chaotic system,
Telecommunication Systems, Vol. 52, pp563 571, 2013.
334
(ICoIT'17)
10thof April, 2017.
Image Enhancement
By Using Homomorphic Filtering Model
Assist.Prof. Amel H.Abbas
Dr_amal2006@yahoo.com
Prof.Dr.Jamila Harbi S.
Dr.jameelahharbi@uomustansiriyah.edu.iq

A number of techniques have been proposed in the literature
Article History: to deal with the illumination induced appearance variations
Received: ranging from simple image enhancement techniques, such as
histogram equalization or gamma intensity correction. In this
paper we proposed a homomorphic filtering model or the
Published: 10 April 2017 logarithmic total variation model to enhance non uniform
illumination in real experimental image. This model will be
Keywords: uniform enhancing the brightness in high frequency and in low
Illumination, Gamma frequency. We improved the image by using the output
intensity, Image resulted from our proposed model as input again to this
enhancement, model, the final enhanced image is de-noised from all
Homomorphic filtering. unwanted details and the edge will be enhanced in all
direction. Some fidelity parameters is applied and we obtain
an acceptable results.
1. INTRODUCTION
I Image enhancement is used when the image has a degradation of natural look
and the contrast is higher. The contrast enhancement is a widely using image
enhancement technique. The difference between contrast of color image and the
brightness makes an object in the scene distinguishable. The brightness is the feature of
vision that defines how much light the source reflects makes the source appears to be
reflecting the light. But the brightness and the lightness are two different aspects [1]. The
contrast can also be defined as the amount of color or the grey scale differentiation value that
exists between various images. Many algorithms, such as the Retinex based algorithms [2],
the unsharp masking algorithms [3], and the histogram equalization (HE) algorithms [4], etc.,
are proposed to enhance the image brightness. These algorithms are focused on detail
enhancement, but the result of these algorithms is given unnatural looks, caused by light
source confusion and artifacts. Therefore others attempt to reduce over-enhancement at the
cost of the details [5]. A strong correlation with reflectance and the amount of visible light
335
(ICoIT'17)
10thof April, 2017.
reaching observers depends on the product of reflectance and illumination assumed in

Retinex theory [6]. In fact, the reflectance value range is [0, 1], which means the surface
cannot reflect more light than that it receives. Moreover, it is unreasonable to simply remove
the illumination which is essential to represent the ambience [7]. The bright-pass filter which
is used to preserve the natural look in the non-illumination images also restricts the
reflectance to 0 and 1. This will give the output as grey scale image. But the complexity of
implementing the bright-pass filter is more compared with the other enhancement algorithms
[6].
The intensity is needed to preserve in an enhancement applications. Therefore, for non-
uniform illumination images, brightness preservation is disadvantageous to detail
enhancement in the areas of inappropriate intensity, such as the dark areas [8].
2. PROPOSED IMAGE ENHANCEMENT MODEL
In this paper an enhancement model is proposed to enhance an image that collected

from real experimental of LASER generation. This model will be applied on these images in
frequency domain by using homomorphic filter. This model is depend on the image has
produced high energy that received from object. The energy received is determined by
illumination radiation source and the features of reflectance of the object itself. The
relationship is viewed as product between two components of energy (i.e., illumination and
reflectance components) to produce an optical image [9] [10]:
(, ) = (, )(, ) . (1)
Where: f(x,y) is an optical image, and i(x,y)&r(x,y) are the illumination and reflectance
components.
As previously said the reflectance component is means the features of the object
itself. Therefore we more interest with reflectance than illumination component. Eq.1can is
converted into linear model by using the log operation such as:
(, ) = (, ) + (, ) . (2)
Form Eq,2 can be easy to process in frequency domain, because the illumination component
lies in low frequency band and reflectance in high frequency band. Discrete Fourier
transform (DFT) is applied on eq.2, and then high pass filter is used to enhance reflectance
part and degrade the other. After enhancing done the inverse DFT (IDFT) is used to
reconstruct the original image after exponential operation is used to convert the illumination-
reflectance original space. All these process called Homomorphic filter is shown in Fig.1.
336
(ICoIT'17)
10thof April, 2017.
.
FIGURE 1: Homomorphic filter steps
Pearson Correlation Coefficient factor can be calculated from eq.3

1 2
= . (3)
(1) (2)
3. RESULTS AND DISCUSSION
In the real world the corrupted images are expected to be the way the image that is corrupted
by this type of non-uniform illumination. The corrupted image under investigation is as
shown in Fig.2. Also Fig.2 showed the original color image and the converted gray scale
image. The aim of the paper is to use homomorphic filter out the illumination component to
enhance the original image.
Image1Size=198x198 Image3Size= 270x270

Image2Size 383x383
(a)
337
(ICoIT'17)
10thof April, 2017.
(b)
FIGURE 2: original non-uniform illumination image, a) color image and b) gray scale image
Fig.3 illustrated the brightness distribution by drawing histogram in 3-D of the original
images.
(a) (b) (c )
FIGURE 3: 3-D histogram of original images: a) image1, b) image2, and c) image3.
Before enhancement model applied we modify the brightness by using equalization

histogram, Fig.4 is illustrated the gray scale image modification with its histogram.
Brightness modification insufficient which is shown in Fig.4, therefore we used
another modification to enhance the non-uniform illumination. Homomorphic filtering is
used in frequency domain to enhance the brightness more than histogram equalization that
applied in spatial domain. Also, homomorphic filtering applied twice time on the resultant
image to get more enhance on the brightness without loss any information.
(a)
(b)
PSNR=10.72dB and SNR = 2.51dB
FIGURE.4: Image2 enhancement by using equalization histogram: a) image2 and its histogram before
enhancing the brightness, and b) image2 and its histogram after enhancing the brightness
338
(ICoIT'17)
10thof April, 2017.
The experimental result is demonstrated in Fig.5. RMSE value is 65.57, MSE= 4299.6,
PSNR = 11.83dB, Pearson Correlation Coefficient between original image and enhance
image (PCC1) value is 38320.85 and Pearson Correlation Coefficient between original image
and original image (PCC2) value is 39203.
FIGURE 5: Homomorphic filtering applied on Image1and the reconstructed image.
Fig.6 is shown the modification of the brightness distribution by using homomorphic filtering
model.
FIGURE 6: Histogram modification by using homomorphic filtering model.
If we use the output image of Fig.5 as input to the enhancement model, therefore we obtain
the more enhance image as shown in Fig.7. RMSE value is 45.998, PCC1 value is 37897.38
and PCC2 value is 39203.
(a) (b)
FIGURE 7: Modify homomorphic filtering: a) first applied enhancement, and b) second enhancement.
Figs.8, 9, &10 are shown the real enhancement by using homomorphic filtering model. In
Fig.8 the results values are: RMSE = 59.302, MSE = 6154.83, PSNR = 9.77dB,
339
(ICoIT'17)
10thof April, 2017.
PCC1=127335.17 and PCC2=146688. Fig.9 demonstrated the histogram of each steps in

homomorphic filtering model. Fig.10 is illustrating the output of Fig.8 used as input of Fig.10
and the values of image fidelity are: RMSE= 69.509, PCC1= 139365.96 and PCC2= 146688.
FIGURE 8: Homomorphic filtering applied on Image2, and the reconstructed image.
FIGURE 9: Histogram modification of each steps of homomorphic filtering model.
(a) (b)
FIGURE 10: Modify homomorphic filtering: a) first applied enhancement, and b) second enhancement
Figs.11, 12, &13 are shown the real image3 enhancement by using homomorphic filtering
model. In Fig.11 the results values are: RMSE = 42.60, MSE = 1717.61, PSNR = 15.35dB,
PCC1= 66184.02, PCC2=72899.
Fig.12 showing the histogram graph of each steps of homomorphic filtering model. Fig.13 is
illustrating the output of Fig.8 used as input of Fig.9 and the values of image fidelity are:
RMSE = 132.59, PCC1= 70035.66, PCC2= 72899.
340
(ICoIT'17)
10thof April, 2017.
FIGURE 11: Homomorphic filtering applied on Image3, and the reconstructed image .
FIGURE 12: Histogram modification of each steps of homomorphic filtering model.
(a) (b)
FIGURE 13: Modify homomorphic filtering: a) first applied enhancement, and b) second enhancement
4. CONCLUSIONS
The proposed modification model is enhanced the illuminate component of the non-
uniform brightness images that collected from the output LASER experiment. The
brightness is modified the brightness in high frequency and in low frequency therefore, we
improved the results by using homomorphic filtering two times and we get acceptable
results of some fidelity parameters.
5. REFERENCES
[1] T. Arici, S. Dikbas, and Y. Altunbasak, A Histogram Modification Framework

and Its Application for Image Contrast Enhancement, IEEE Trans. Image
process., Vol. 18, No. 9, pp. 19211935, Sep. 2009.
341
(ICoIT'17)
10thof April, 2017.
[2] D. J. Jobson, Z. Rahman, and G. A. Woodell, Properties and Performance Of A

Center/Surround Retinex, IEEE Trans. Image Process., Vol. 6, No. 3, pp. 451
462, Mar. 1996.
[3] G. Deng, A Generalized Unsharp Masking Algorithm, IEEE Trans. Image
Process., Vol. 20, No. 5, pp. 12491261, May 2011.
[4] C. Wang and Z. Ye, Brightness Preserving Histogram Equalization With
Maximum Entropy: A Variation Perspective, IEEE Trans. Consum. Electron.,
Vol. 51, No. 4, pp. 13261334, Nov. 2005.
[5] Shuhang Wang, Jin Zheng, Hai-Miao Hu, and Bo Li, Naturalness Preserved
Enhancement Algorithm for Non-Uniform Illumination images, IEEE Trans.
Image Process., Vol. 22, No. 9, pp.3538-3548, Sep. 2013.
[6] M. Bertalmio, V. Caselles, and E. Provenzi, Issues about Retinex Theory and
Contrast Enhancement, Int. J. Comput. Vis., Vol.83, No. 1, pp. 101119, 2009.
[7] B. Li, S. Wang, and Y. Geng, Image enhancement based on Retinex and
lightness decomposition, in Proc. IEEE Int. Conf.Image Process., Sep. 2011, pp.
34173420.
[8] H. Ibrahim and N. Kong, Brightness Preserving Dynamic Histogram Qualization
for Image Contrast Enhancement, IEEE Trans. Consum. Electron. Vol. 53, No.
4, pp. 17521758, Nov. 2007.
[9] Delac, K.; Grgic, M. & Kos, T.,Sub-Image Homomorphic Filtering Techniques
for Improving Facial Identification under Difficult Illumination Conditions,
International Conference on System, Signals and Image Processing, Budapest,
2006.
[10] Megha Rajan, Radhakrishnan B, and Raji P G, Combined Contrast Stretching
and Homomorphic Normalized Filtering for Color Image Enhancement,
International Journal of Engineering Research and General Science Volume 3,
Issue 4, Part-2, July-August, 2015.
342
(ICoIT'17)
10thof April, 2017.
Implementation of an Economic Light Duty Three-Axis

Computer Numerical Control Machine
Ali A. Abed (SMIEEE)
Computer Engineering, College of Engineering, University of Basra- Iraq
ali.ahmed.abed@gmail.com
Murtadha A. Al-Ibraheem, Auday A. Mohammed
Computer Engineering, College of Engineering, University of Basra- Iraq
malibraheem21@gmail.com, uday.adil.h@gmail.com

The Computer Numerical Control (CNC) is an advanced
Article History: technology that has automated many manufacturing
Received: 17 March 2017 applications concerning parts machining as well as other
applications. This work presents the implementation of a low-
cost desktop size 3-axis CNC machine. It has been built from
Published: 10 April 2017 free software and cheap components. It is able for light duty
Keywords: : CNC, Grbl, cutting, painting, carving, and drawing. The control system of
microcontroller, G-code, the machine consists of a PC and a microcontroller running
CAD/CAM, laser diode Grbl software, which is an open source free program. The PC
has the task of loading files written in a special programming
language for CNCs, which is called G-code. It sends the lines
of this file to the microcontroller, which parses the code,
interpolates the motion commands, and drives the stepper
motors of the machine. The system is built and tested
experimentally to prove its operation, robustness, and
accuracy.
1. INTRODUCTION
umerical Control (NC) is a programmable automation technology that controls a
N process by letters, symbols, and numbers. NC machines are mechatronic products

composed of mechanical and electrical components in which physical movements
in one axis or more are controlled by instructions [1]. NC machines come in a
variety of kinds that serve different automated industrial needs such as drawing, cutting,
engraving, plotting, 3D printing and others. NCs also provide advantages in digital
fabrication of customized parts due to their accuracy and reduced time-to-market. NC
machines controllers can be built using transistors, circuits, and logic elements. Modern NC
controllers are optimizing the advanced technology of microprocessors and computers to
implement NC control functions. NC functions implemented with microprocessors are called
Computer Numerical Control (CNC). The CNC workflow starts with an engineered design
using Computer Aided Design (CAD) software. The CAD file is processed with a Computer
Aided Process Plan (CAPP) that generates the machining necessary information such as
343
(ICoIT'17)
10thof April, 2017.
machine tools selection, jig and fixture, cutting conditions, and machining sequences. Based
on the information provided by CAD and CAPP, the Computer Aided Machining (CAM) is
carried out to generate the tool paths. The generated tool paths are coded with a special
programming language standard (ISO 6983-1) which is known as G-code [2]. G-code
contains information that guides the motors of the system and sets some parameters of the
CNC control system. CAM software can generate a wide range of G-code extensions based
on the type and manufacturer of the CNC machine. Generally, the CNC system composed of
Man Machine Interface (MMI) unit, Numerical Control Kernel (NCK) motors and drivers
unit.
MMI: composed of a Graphical User Interface (GUI) application that offers the interface
between the motion controller and the user. It provides the facilities for the user to load the
part program, display the machine coordinates and tool paths, and a command line to send
single instructions to the motion controller.
NCK: the part that is responsible for reading the G-code lines, extracts the motion
commands, determine the tool paths, and position control the motors by delivering the motors
driver units with appropriate signals.
Motors and drivers unit: Includes servomotors or stepper motors and their respective
drivers.
Figure 1 shows a typical CNC structure with labels for the key components. It also
indicates the 3 axes of movement (X, Y and Z). This is an example of a simple category of
the available CNCs. There are tremendous numbers of configurations that differs in the
number of supported axes of movements, sizes, and other machining features.
2. SYSTRM DESCRIPTION
The constructed machine composed of a GUI running on a Windows operated PC. It

transmits G-code lines to a microcontroller, which is the core of the machine. It runs software
that receives G-code lines, processes them, and generates the appropriate signals for the
motor drivers. The motor drivers operate the machine stepper motors.
2.1. STRUCTURE
Figure 2 (a) shows the hardware of the designed CNC. In this work, the three primary
linear axes X, Y, and Z are considered.
A ball screw shown in Figure 2 (b), which is a lead screw operated by a ball bearing, takes
the role of converting the rotation of the motor to linear movement. It also determines the
Basic Length Unit (BLU) of the machine tool. Here, three ball screws are coupled with each
of the three stepper motors shafts using a plastic tube to reduce the sensitivity to misaligned
rotation centers.
344
(ICoIT'17)
10thof April, 2017.
FIGURE 1. 3-axis CNC structure
FIGURE 2. a) The CNC System (left), b) Ball Screws (right)
In order to increase the linear movement accuracy and smoothness, aluminum drawer
slides shown in Figure 3 (a), have been proposed as the linear movement guide of the table
(Y), the gantry (X) and the spindle (Z).
The work area of the machine is 30x20 cm, and various options of light weight drills or
low watts laser diodes can be used as cutting tools.
FIGURE 3. (a) Drawer Slide (left), (b) Stepper motor coupling with slide (right)
345
(ICoIT'17)
10thof April, 2017.
2.2. MOTORS
Generally, CNC machines movements can be accomplished through the usage of stepper
motors or servomotors. The control system is of open loop type when stepper motors are
used. While in the case of servomotors, the control system would be a closed loop system.
Servomotors work in a closed loop setting where they continuously receive feedback
signals forms the encoders. These signals dictate the current position and check whether the
control signals executed properly. This setting gives the servomotors advantage over stepper
motors in that there is no possible way of losing steps. In addition, servomotors offer high
torque and speeds.
Stepper motors are DC motors move in discrete steps. Very precious speed and
positioning control can be obtained using stepper motors. They are good for low speed/high
precision applications, but their torque is less at high speeds. Stepper motors do not have
position feedback. They operate in open loop control system. Figure 3 (b) illustrates the
coupling of the stepper motor shaft with the ball bearing screw.
In this work, the used stepper motors are of size NEMA-17 and operated by 12 volt DC
power supply.
2.3. STEPPER MOTORS DRIVERS

According to the way of energizing stepper phases, there are two distinct types of stepper
drivers, namely, bipolar and unipolar. Unipolar drivers could be implemented with simple
transistors circuitry. Their disadvantage is less available torque. Bipolar drivers use more
advanced circuitry (H-bridge) that actually energizes all the coils of the steppers which
deliver higher torque.
In this work, the A4988 micro-stepping motor driver is adopted. This driver operates
bipolar stepper motors in full, 1/2, 1/4, 1/8, and 1/16 step modes [3]. The minimal wiring
needed to connect the driver to the microcontroller is shown in Figure 4.
2.4. MOTION CONTROLLER

Grbl motion controller is a C program optimized to utilize all the tremendous features
of the Atmega328p chips used in Arduino UNO microcontrollers boards. It is a high
performance, free, open source software, and available for three-axis machines (X, Y, and
Z). It interprets the G-code lines received serially from the GUI running on a PC, processes
them, and delivers the control pulses to the motor drivers. G-code file is of ASCII format,
and is composed of blocks. Each block contains a number of words [4]. A word consists of
two parts: an address, which is an alphabetical character, and a numerical value. The
incorporated G-code interpreter follows a subset of the Linux CNC standard. Grbl (last
version is v1.1) supports arcs, circular and helical moves [5] and also capable of processing
346
(ICoIT'17)
10thof April, 2017.
commands that are related to the coolant system and spindle control.
FIGURE 4. A4988 connection diagram FIGURE 5. Grbl main processing steps
Table 1 lists some of the G and M commands that Grbl

v1.1 can handle [5]. The processing steps of Grbl are illustrated in the flow chart shown in
Figure 5.
347
(ICoIT'17)
10thof April, 2017.
FIGURE 4. A4988 connection diagram FIGURE 5. Grbl main processing steps
TABLE 1: Supported G-codes in Grbl v1.1
G0, G1 Linear Motions

G2, G3 Arc and Helical Motions
G4 Dwell
G10 L2, G10 L20 Set Work Coordinate Offsets
G17, G18, G19 Plane Selection
G20, G21 Units
G28, G30 Go to Pre-Defined Position
G28.1, G30.1 Set Pre-Defined Position
G38.2, G38.3, Probing
G38.4, G38.5
G40 Cutter Radius Compensation Modes
G43.1, G49 Dynamic Tool Length Offsets
2.5. MACHINING TOOL

Various types of CNCs are classified based on the tools that are used for machining. Laser
diodes can be used as a tool for engraving and cutting materials. Drills can be attached to the
tool holder and do various cutting and carving operations. Grbl v1.1 offers modes for using
drills and lasers. It supports a set of commands that controls the laser power or the spindle
speed. The spindle tool of the proposed machine is a 745 mW laser diode that is suitable for
cutting and engraving thing materials.
348
(ICoIT'17)
10thof April, 2017.
2.6. SYSTEM INTERFACE

Figure 6 gives an abstract view of system connections while Figure 7 shows the Grbl
Controller application. Grbl Controller provides a simplified graphical user interface with all
basic functions needed to communicate with it. It opens a USB port with the Arduino UNO
at a specific baud rate e.g. 9600 is taken in this work. It offers a single line command that
gives users the ability to manually give commands to the machine besides the ability to load
a stored G-code file. It also offers a visualizer that shows the tool path of a loaded file. Figure
8 illustrates the G-code file that is loaded to the Grbl Controller GUI. G-code file is obtained
from Aspire software which is CAD/CAM software.
FIGURE 6. System Interface FIGURE 7. Grbl Controller
A CNC shield has been utilized to simplify the wiring of the system. CNC shield as shown
in Figure 9 is a breakout board that has been designed specifically to work with Grbl CNCs.
It has four sockets that are compatible with A4988 driver, which provides the ability to drive
4 stepper motors, namely, X, Y, and Z, plus one auxiliary motor. It also has additional
connectors that facilitate the connection of limit sensors and control buttons.
FIGURE 8. Typical G-code file FIGURE 9. CNC Shield
3. PRACTICAL RESULTS
A designed part using 3DMAX, which is CAD software, is shown in Figure 10. This
design is saved as a 2D vector file. This file is opened with Aspire, which is CAD/CAM
349
(ICoIT'17)
10thof April, 2017.
software, to create accurate and efficient toolpaths to cut the designed shape, as indicated in
Figure 11.
FIGURE 10. A 2D design using Aspire. FIGURE 11. Tool paths generation with Aspire
The final product after cutting using a laser diode attached to the machine spindle is shown
in Figure 12.
FIGURE 12. Designed shape engraved by laser on acrylic material
4. CONCLUSIONS AND FUTURE WORK
In this paper, a light duty 3-axis CNC machine structure and control system have been
discussed. The machine has been built using wood, drawer sliders and ball screws. Three
stepper motors have been used to move the axes of the machine in the X, Y, and Z directions.
The core control of the CNC is achieved by utilizing the considerable features of Grbl v1.1.
The system has been tested to draw different shapes designed with CAD software (3D
MAX). CAD and CAM software are embedded in current available programs such as Aspire.
This program offers the facilities to design the parts and generate the tool paths based on the
available profiling options. The generated tool path file, which is called G-code, is loaded to
Grbl through serial communication with a Grbl Controller application running on a personal
350
(ICoIT'17)
10thof April, 2017.
computer. The constructed machine can be utilized in drawing, laser engraving and carving
on several types of material such as wood and plastic. This gives the ability to invest this
machine in producing numbers of products that could be marketed for small business
projects.
Through utilizing the advanced technology of System on Chips (SoC), the CNC control
system can be integrated in a single chip. Integrated control system of CNC omits the need
for external PC and provides more portability and speeds processing of the system.
5. REFERENCES
[1] S.-H. Suh and S.-K. Kang, Theory and Design of CNC Systems, Springer, 2008.
[2] T. R. Kramer, F. M. Proctor and E. Messina, "The NIST RS274NGC Interpreter -
Version 3," NISTIR 6556, 2000.
[3] Allegro Microsystems, "DMOS Microstepping Driver with Translator", A4988
datasheet, Rev. 5.
[4] Roland DG Corporation, NC Code Reference Manual, Roland DG Corporation,
2009.
[5] "Grbl Wiki," [Online]. Available: https://github.com/gnea/grbl.
[6] D. V. Can, N. Q. Dich and N. P. Quang, "Proposed Structure Hardware of the CNC
for Machine Tools Based on System on," in The 2nd International Conference on
Control, Automation and Robotics, 2016.
[7] A. Albert, Understanding CNC Routers, vol. I, FP Innovations.
[8] J. Bi, T. Yu and Q. Li, "I2NC: A New Type of Computer Numerical Control," in
Proceedings of the 2006 IEEE International Conference on Mechatronics and
Automation, June 25 - 28, 2006, Luoyang, China, 2006.
351
(ICoIT'17)
10thof April, 2017.
Intelligent Fault Recovery Controller for Power Generator at

Perdawd CCGS in Kurdistan-Iraq
Wassan Adnan Hashim
Computer Science, Information Technology, Bayan University - Iraq
wassanhashim@gmail.com
Faiz Al Alawy
College of Technology, Kent State University U.S.A.
Email: falalaw@kent.edu.

Different Artificial Intelligent (AI) tools are becoming
Article History: commonly used to design different intelligent controllers. The
Received: 9 March 2017 objective of this work is to investigate the performance of a
suggested intelligent controller of a specific power generation
system during and after sadden faults. Two types of AI tools,
Published: 10 April 2017 genetic algorithms and adaptive neuro-fuzzy inference
system, have been used in this work to design a control unit
Keywords: ANFIS, for the power station mentioned above as a case study.
CCGS, KRPS, Genetic Simulated models were designed to optimize the control
algorithm, Power System parameters to enhance the performance. A wide range of
Stabilizer, objective training data pairs was used to train the controllers and then
function. to test the system performance. A comparison between the
performance of the intelligent and conventional controllers
has been introduced. The simulation results show a clear
prove that the intelligent controller is outperform the classical
one especially during the fault recovery time. Damping
performance, oscillation, the system stability and the dynamic
performance of electrical power system have been studied.
1. INTRODUCTION
M
ost of the complex control problems cannot be solved by using conventional
techniques. Modern intelligent tools are adopted to simplify these complex
problems and design more robust controllers using what is known as a
Computational Intelligence (CI) [1].
During the last few decades, many researches have studied and applied different Artificial
Intelligence (AI) techniques as the most popular methods in Heuristic optimization
techniques. There is a growing interest in using AI tools in various fields electrical power
and power generation plants such as using the AI techniques to design a hybrid control
systems which integrate the conventional (mathematical approaches) together the intelligent
tools in controlling the performance of different electrical power systems. [2, 3]
352
(ICoIT'17)
10thof April, 2017.
There are many AI techniques that aimed to compute the optimal or near optimal solutions
for different optimization problems such as: Artificial Neural Network (ANN), expert
system (ES), Evolutionary Computation (EC), Genetic Algorithms (GA), Particle Swarm
Optimization (PSO), Ant colony Optimization, Fuzzy logic and others. Many of these
techniques are used to find the optimal working parameters of power control systems,
excitation systems and power system stabilizer (PSS) [4-7].
2. ARTIFICIAL INTELLIGENT TECHNIQUES

2.1. GENETIC ALGORITHM
GA is based on the Darwinian principles of inheriting and evolving the best offsprings
features which can be used as a search method in finding the optimization and best fit global
or parallel solution.
Genetic algorithms operate on a population of potentially huge search spaces that applying
the principle of survival of the ttest to look for optimal aggregation of better and better
approximations to a solution. [8]
The standard GAs can be implemented through a coding that represents the parameters or
variables and select the best criteria for reproduction, recombination (crossover) and
mutation. The initial set of chromosomes is representing the initial and each chromosomes
is composed of many genes which represent given measures of each single variable. A
fitness function (objective function) is normally used to transform the search problem into a
minimization problem with which we should find nearest fit (or exact fit in some cases) to
the optimal solution or performance [9, 10]:
F (x ) g (f (x )) (1)
Where:
f, is the objective function;
g, transforms the value of the objective function to a non-negative number;
F, is the resulting relative tness.
A simple GA can be represented as follows:
Create a population (string of chromosomes)
Evaluate a population through fitness factor
Check the solution if optimum:
o Reproduction (Selection);
o Crossover;
o Mutation;
Generate a new offspring
The basic elements of a genetic algorithm consist of three types of operators:

Selection: randomly selecting chromosomes from the available the population in order to
carry out reproduction operation. Chromosomes are allocated specific priorities proportional
to their fitness and thus the fittest individuals are selected as the best chromosomes (solution).
Crossover: the specific or randomly chosen point within the chromosome to exchange
between two chromosomes in order to create two offspring (next generation). After that the
new string is subjected to a specific probability of mutation operation.
353
(ICoIT'17)
10thof April, 2017.
Mutation: is used to avoid a deadlock or falling in a local minimum point which might
lead a non-accurate solution. The importance of this operation is extracted to maintain variety
in population through performing a gene or bit-wise process.
The below diagram illustrates the whole cycle of the GA operation:
T=0
Initialize Evaluate
start
population Solutions
Is optimum Yes
solution Stop
(Fitness)?
NO
G
T=T+1 A Selection
o
p
e Crossover
r
a
t
o Mutation
r
FIGURE 1. A simple cycle of Genetic Algorithm

In this work, an initial population of randomly selected chromosomes have been
represented to simulate the variables of the power plant control unit. The chromosomes
representation is used to describe the structures and value of each of the addressed variables.
Each individual or chromosome that mapped the variables of control unit is made up from a
binary strings (0's, 1's). A suitable fitness function was designed to measure the outcome
fitness after each iteration or new generation. A good objective function is a core issue in
determining the best solution for the addressed control problem. The optimization solution
here can be defined as following:
F(x)Max : x min x x max; (2)
Where x is a parameter of genetic algorithm which must be design.
2.2. ADAPTIVE NEURO FUZZY INFERENCE SYSTEM

Adaptive Neural network based fuzzy inference system is a type of artificial neural
network which is based on Takagi-Sugeno fuzzy inference system. ANFIS was presented in
the early 1990 from Wang, since, it's combined both neural networks and fuzzy logic
principle and convened the benefits of both in a one framework [11].
ANFIS is a hybrid learning techniques which uses back propagation training algorithm
during the initial training section to optimize the parameters of fuzzy system. Three types
of membership functions are usually used to implement ANFIS, these three types are:
membership function, Gaussian, generalized bell and sigmoid. Each input variable of fuzzy
inference system can have an arbitrary number and mix of these membership functions.
354
(ICoIT'17)
10thof April, 2017.
In this work, the first order Sugeno system with two fuzzy IF-THEN rule has been
implemented as follows: [12, 13].
Rule 1: if x is A1 and y is B1 then f1= p1x+q1y+r1
Rule 2: if x is A2 and y is B2 then f2= p2x+q2y+r2 (3)
Where: A1, A2, B1, B2, are membership functions for inputs (x, y), and (p1,q1,r1) ,
(p2,q2,r2)are the parameters of output functions. The architecture of ANFIS can be
represented as shown in figure (2).
FIGURE 1. A simple cycle of Genetic Algorithm
A set of numerical data consisting of an inputoutput space is usually used to construct a

neuro-fuzzy system. The construction of the system involves two essential phases:
(I) Optimization of the structure or structure learning phase and
(II) The parameter learning phase.
The partitioning of inputoutput space is a technique used to form and visualize the
mapping relation of fuzzy rules and explain how the fuzzy rules are related to the input space.
The structure and the number of fuzzy rules are inuenced by the partitioning techniques.
The most common partitioning method is grid-type partitioning and clustering. Figures 3 and
4 illustrate the input-output space partitioning[14].
355
(ICoIT'17)
10thof April, 2017.
FIGURE 3. Grid type partition.
FIGURE 4. Clustering based partitioning.
3. APPLY AI FOR ENHANCING THE DYNAMIC STABILITY OF ELECTRICAL

POWER SYSTEM
Case study of this work is the Perdawd Combined Cycle Gas Station (PCCGS). The
Kurdistan Regional Power System of Iraq (KRPS) comprises two hydro power stations,
Dokan (400 MVA) and Darbandikhan (249 MVA) in addition to two gas stations cited at
Perdawood (500 MVA) and Chamchamal (750 MVA). Another two more gas stations are
under construction in Duhok (200 MVA) and TaqTaq(200 MVA). The four gas stations are
private sector investments. Currently, KRPS uses 35 load nodes at 132 kV substation ends
totaling 2202 MW. The system serves five million population in an area of 80,000 square
km[15].
356
(ICoIT'17)
10thof April, 2017.
3.1. THE MODEL OF PCCGS

A two axis nonlinear machine model is representing the generator of the suggested case
study (Perdawd CCGS). The control unit of power station is implemented through excitation
system, type IEEE ST1A, and power system stabilizer, type conventional Lead/Lag (CPSS).
The electrical part of Perdawd CCGS is modeled by Simulink tools of using MatLab software
program as shown in figure (5).
Figure.5. model of Perdawd CCGS at Simulink\Matlab
3.2. TUNING PARAMETERS OF THE CONTROL UNIT USING GA

The objective is to optimize the key parameters of control unit (exciter and CPSS) using
the genetic algorithms in order to reduce the angular speed oscillation () and the load angle
() at minimum values with best settling time for transient state for the suggested case study
under any possible sudden disturbance.
3.2.1. Fitness function
It's used to provide an index of how the best solution is achieved in the problem search
space, can be seen as a minimization problem. The multi objective function of our case
study is characterized such as:
Minimum J for CPSS and exciter parameters = ( W J);

Where J is defined as a sum of square error which is represented the disparity between
desired and real value of variables (terminal voltage of exciter (vt), rotor speed (m) and
settling time (ts)).
J1 vt desired vt real
2
(4)
J 2 mdesired mreal
2
(5)
J 3 tsdesired ts real
2
(6)
The weight coefficients (W) are employed to include the effective weight or the
importance of each individual parameter in the objective functions.
Based on the evaluation of fitness function at nominal operating of power station and
randomly initialization generation of individuals, the genetic algorithm will produce a new
version of population in a process which should continue until convergence is accomplished.
357
(ICoIT'17)
10thof April, 2017.
TABLE 1.
Genetic algorithm variables for case study
Mutation Rate, % 0.8
Population Size, Npop 20
Number Of Variables 13
Maxi. Gen. 250
Gen. Gap 0.75
Precision 20
3.2.2. Simulation Analysis

High value of crossover and mutation rates were used in the proposed simulation method
in this simulation to help find the best choice of PSS and Exciter parameters which can
guarantee speedy converges of GA. Matlab software program is used to simulate the power
station model under disturbance (three phase fault) at nominal operation of station. The
performance of the control unit (under different parameter values) has been evaluated for
both situation (Conventional and GA based controllers) and it is clearly proving how the GA
based control system is outperforming the conventional one especially when recovering a
faulty operation. The CG based controller is also better than the classical control unit in
reducing oscillation and increasing the cleaning of critical time.
Figure 6 illustrates the dynamic response of station studied at classical control, and with
genetic control.
358
(ICoIT'17)
10thof April, 2017.
c
FIGURE. 6. Simulation of power station with GA for three situations: (a) exciter voltage
vf, (b) speed of rotor, m, (c) the load angle delta ( )
3.3. ENHANCING THE CONTROLLER DYNAMIC STABILITY USING ANFIS-

GRID
The core of the proposed ANFIS based controller is designed using fuzzy linear model
type Sugeno which converts a fuzzy inference engine into an adaptive network which is used
to train the pairs of inputs, dedicated by both factors, the operating conditions and parameters
of excitation system (Tlead, Tlag, K=f(k1, k2), Ti, KF, TF), consequently with outputs,
implemented by the parameters of PSS (Kpss,Tw,Tw1,TLead1,2,3,TLag1,2,3 ). 1600
samples of training inputoutput data pairs are used to train the ANFIS based power station
control unit. It's worth noted that, these pairs of data are partitioned in two sets, one for
training set which is used as a training data and the other is to system testing to evaluate
performance of ANFIS during a validation state [16].
In this study, three membership functions (MFs) are assigned to each variable, and the
desired network has 751 fuzzy rules as shown in figure 7. The training algorithm is
implemented on the PSS parameters of studied station. The raining process of ANFISGRID
network has consumed (50200 sec) of CPU time for each parameter of PSS
FIGURE.7. The ANFISGRID structure for station studied
359
(ICoIT'17)
10thof April, 2017.
3.3.1. SIMULATION ANALYSIS

Time domain simulations with ANFIS-PSS has been simulated for Perdawd CCGS under
three-phase fault (5-cycle of simulation time) at nominal loading conditions.
MATLAB\Simulink has been used to run the analysis and evaluate the performance. Figure
8 illustrates the dynamic response of station studied with CPSS, and with adaptive PSS
(ANFIS-PSS).
(a)
(b)
(c)
FIGURE. 8. Simulation of power station with ANFIS for three situations :( (a) Exciter
voltage vf, (b) speed of rotor, m, (c) the load angle delta ( )
4. DISCUSSION AND CONCLUSION

A case study analysis has been carried out on Perdawd CCGS power generator controller
when it was at transient state. The simulation result shows that a high oscillation and
360
(ICoIT'17)
10thof April, 2017.
instability of voltage is detected in response to a faulty situation as shown in the above

figures. Conversely, the efficiency of the GA based controller is verified and the evolved
values of the optimal control unit parameters have led to reduce the oscillation and increase
the ability to produce better damping over a wide operating range and enhances the
performance of the power station. It has been noticed that the response of the system studied
V
with CPSS is unstable at external voltage ( f =1 pu) during the disturbance, could cause to
melting. While, the dynamic stability of system studied enhances and stability by using
adaptive power system ANFIS-PSS. Also, the critical clearing time (CCT) of the control
system at CPSS is very high compared with other situation (ANFIS-PSS) that is equal (35
sec).
As a general conclusion, it is confirmed that the outcome of the intelligent controller is
much better than the classical (conventional) controller from different point of views. Adding
intelligent component to the controller has contributed to reduce the oscillation of power
signal when fault occurs, increase capability to recover on critical time of instability and
return the situation of the power signal to stability extremely when exposing fault tolerance
to the power station
5. REFERENCES
[1] Michael S. Bittermann," Artificial Intelligence (AI) versus Computational Intelligence
(CI) for treatment of complexity in design," Design Computing and Cognition
DCC10. J.S Gero (ed), Springer, 2010.
[2] A. Konar, Computational Intelligence: Principles, Techniques and Applications,
Springer, 2005.
[2] A. Kusiak, Computational Intelligence in Design and Manufacturing. Wiley-

Interscience, 2000.
[3] Xiaofei Wang, Xiuaua LI and Victor C. M. Leung, " Artificial Intelligence-Based
Techniques for Emerging Heterogeneous Network: State of the Arts, Opportunities,
and Challenges," in IEEE Access SPECIAL SECTION ON ARTIFICIAL
INTELLIGENCE ENABLED NETWORKING, Volume 3, 2015, pp. 1379-1391.
[4] Kunjin Chen, Caowei Huang, Jinliang He, " Fault detection, classification and location
for transmission lines and distribution systems: a review on the methods," High
Voltage, Review Article, Vol. 1, Iss. 1, 2016,pp. 2533.
[5] Graciliano Nicols Marichal, Mara Lourdes Del Castillo, Jess Lpez, Isidro Padrn
and Mariano Arts, " An Articial Intelligence Approach for Gears Diagnostics in
AUVs," Sensors 2016, 16 (www.mdpi.com/journal/sensors).
[6] Riad Al-Mustfa, Anas Fattouh, and Wassan A. Hashim, Robust Coordinated Design
of PSS and Excitation System using Genetic Algorithm to Enhace the Dynamic
Stability of Al-Zara Thermal Power Station Generation in Syria, in JKAU: Eng. Sci.,
Vol. 23, No. 1, 2012
[7] Alkhatib H. and Duveau J., "Robust Design of Power System Stabilizers Using
Adaptive Genetic Algorithms", World Academy of Science, Engineering and
Technology, pp:267 272, 2010
361
(ICoIT'17)
10thof April, 2017.
[8] Rahul Malhotra, Narinder Singh, and Yaduvir Singh, Genetic Algorithms: Concepts,
Design for Optimization of Process Controllers, Computer and Information Science,
Vol. 4, No. 2; March 2011.
[9] David A Coley, An Introduction to Genetic Algorithms for Scientists and Engineers,
World Scientific Publishing Co. Pte. Ltd, 1999.
[10] Ulrich Bodenhofer, Genetic Algorithms: Theory and Applications, 3d. ed, Lecture
Notes, Winter 2004 (www.flll.jku.at).
[11] Michael Negnevitsky, Articial Intelligence A Guide to Intelligent Systems, 2d. ed.,
Addison-Wesley, Pearson Education, 2006.
[12] Jyh-Shing Roger Jang , Chuen-Tsai Sun, & Eiji Mizutani, Neuro-Fuzzy and Soft
Computing A Computational Approach to Learning and Machine Intelligence,
Prentice-Hall, Inc, 1997.
[13] Fakhreddine O. K., Clarence de Silva, Soft Computing and Intelligent Systems design:
Theory, Tools and Application, Addison Wesley, Prentice-Hall, 2004.
[14] Werner Dubitzky, Data Mining Techniques in Grid Computing Environments, John
Wiley & Sons, Ltd.,2008.
[15] Ibrahim Hamarash, "Small Signal Stability Analysis of Kurdistan Regional Power
System", [on line], http:// www.researchgate.net/publication/268745758.
362
(ICoIT'17)
10thof April, 2017.
Internet of Things (IoT): Readme

Istabraq M. Al-Joboury
Al-Nahrain University, College of Information Engineering, Department of Networks Engineering,
Baghdad, Iraq
estabriq_94@coie-nahrain.edu.iq
Emad H. Al-Hemiary
Al-Nahrain University, College of Information Engineering, Department of Networks Engineering,
Baghdad, Iraq
emad@coie-nahrain.edu.iq

Internet of Things (IoT) integrates billions of the
Article History: heterogeneous IoT things with the Internet in which the
Received: 19 March 2017 embedded systems such as sensors and actuators linked
together to improve quality of life, and becomes the future of
technologies in any field of human daily life. These IoT
Published: 10 April 2017 devices cooperate with each other and generate useful
information to provide better services and applications to the
Keywords: Internet of governments and the society. Also, there is a need to store
Things (IoT), Internet these data on Cloud for monitoring. This paper, surveys IoT
of Everything (IoE), applications, new challenges and issues arise in different
Machine to Machine fields and provides IoT architecture, focuses on explanation
(M2M), IoT of IoT protocols and their operations and functionalities,
architecture, IoT presents different microcontroller types used by researchers.
protocols, Cloud With the huge amount of data generated from IoT devices, the
Computing, integrating Cloud and IoT may helpful, Therefore, a survey
Microcontrollers on open issues faced when these two concepts integrating
together is discussed. The objective of this paper is to provide
a survey for everything related to IoT and direct it to all
beginners in this filed or academic researchers.
1. INTRODUCTION
T
he IoT is a field that consists of the newest technologies make a revolution of
Internet. It includes a wide range of applications such as smart cities, e-health,
smart home, smart grid and where it monitors smart road, smart waste and patients
in real time [1]. The term "things" consists of interconnected smart devices like
sensors and actuators. These devices are communicating with each other, referred to as
Machine to Machine communication (M2M) [2]. With the rapid growth of the Internet, Cisco
estimates that the world will have 50 billion of IoT devices within the next 20 years. Also,
these devices have sensing feature such as temperature sensor in order to measure Room
temperature [3]. Thus, a large number of data needs to be stored on servers. This will lead to
363
(ICoIT'17)
10thof April, 2017.
integration of IoT and Cloud Computing, which Cloud offers services, tools and software to
end users [4].
The phrase of " Internet of Things" was created by Kevien Ashton on supply-chain
management where he was cofounder and executive director of the MIT Auto-ID Center. He
said regarding to IoT: "I could be wrong, but I'm fairly sure the phrase "Internet of Things"
started life as the title of a presentation I made at Procter & Gamble (P&G) in 1999. Linking
the new idea of RFID in P&G's supply chain to the then-red-hot topic of the Internet was
more than just a good way to get executive attention. It summed up an important insight
which is still often misunderstood."[5]. After years, it was introduced by David L. Brock in
2001. The "thing" is a part of people`s life to interact within the physical world that surrounds
us. The phrase " Auto-ID" includes any identification technology such as bar codes and smart
cards that are used for reducing errors, automating, and increasing efficiency. In 2003, Radio-
Frequency Identification (RFID) application expanded into wider fields at the MIT`s Auto-
ID center, which built the basis for the current IoT sight [6].
People have an addiction to purchase multiple devices, (I Pads, mobile handsets, Tablets,
Laptops, etc.) therefore, the number of devices have been increased and is expected to exceed
the number of people [7]. Data that are extracted from these devices have an impact on
society and economy. Thus, IoT is easy to monitor operations and manage them far away
from organization, track data as it moves from one place to another and even if it crosses the
ocean and detect the changes in the human body like temperature and heartbeat that could
save lives [8]. IoT includes essential types of networks like distributed, grid, ubiquitous, and
vehicular; these have turned towards the world of IT with the five coming years [9].
The original Internet was formed its architecture before the explosion of huge and smaller
devices such as sensors/actuators. With the emergence of IoT, it destroys all previous ideas
regarding network architecture. Thus, create challenges for the current networking paradigm
[10]. IoT develops the enormous applications which makes use of data generated from
different objects that interact and cooperate with each other to provide better services to
governments, organizations and humans.
The rest of this paper is organized as follows: Section 2, discusses the architecture of IoT.
Section 3, presents protocols and standards involved in IoT and their challenges. Section 4,
classifies the most important microcontrollers. Section 5, introduces the integration between
cloud and IoT paradigm. Section 7, focuses anda derives the open issues in Cloud of
Everthing.
2. IOT ARCHITECTURE
The IoT handles billions of different objects that integrate and corporate with each other
to provide the scalability, configuration and modularity of IoT deployments in different
fields, therefore, the layered architecture has taken into consider. It depends on various kinds
of technologies supporting the IoT and describes how these technologies related to each
other. The basic IoT layers consists of three model [11] are: Application, Network, and
Perception Layers. new layers are added [12]: Business and Middleware Layers are recently
proposed.
364
(ICoIT'17)
10thof April, 2017.
2.1. PERCEPTION LAYER

The perception or so-called object/ sensor connectivity layer it's called that, since it
represents the smart objects that are integrated with IoT sensors. Real-time information can
be collected and processed by enabling the interconnection of the digital/physical world by
sensors. The shrinking of hardware size was to produce a smaller form and more powerful
sensors. The sensors can be different types for different purposes. The aim of this layer is
collecting the information from different devices like sensors/actuators, tags (RFID and
Barcode, GPS, camera), creating big data and then performing different functionalities like
taking the measurements such as temperature, humidity, movement and electricity. Sensors
have a number of features and characteristics, for instant: they have a memory-sized for
recording a certain number of measurements. Furthermore, they have an ability to convert
signals from digital to analog that can be understood by devices [13]. Most sensors need a
gateway for aggregating data, which require a connectivity such as LAN (Local Area
Network): Wi-Fi, Ethernet and PAN (Personal Area Network): Bluetooth and ZigBee. Also,
sensors sent the data directly to servers or applications using WAN (Wide Area Network):
GPRS, GSM and LTE. WSNs are part of IoT, these sensors use low-power and low-data-
rate. The WSNs are better than the traditional sensors because they keep sufficient battery
life and covering large areas.
2.2. NETWORK LAYER

The Network or so-called the Object Abstraction Layer. Massive volume of data is
produced from the first layer, that require a high-performance transport medium (gateway)
to collect the information and send to Middleware. Gateways refer to WAN, Mobile
Communication Network, Wireless/ Wired Network as well as provide routing protocols and
reliable delivery. Networks may include several of technologies and protocols to support IoT
and M2M applications to provide new capacities such as context-aware and to support the
communication requirements such as delay, bandwidth, security and privacy. Hence,
Network layer of IoT model is similar to both Network and transport layers of OSI (Open
System Interception) model. The aim of this layer is to transmit the data to Internet through
different technologies like Wi-Fi, 3G, Bluetooth, XBee, etc. [13].
2.3. MIDDLEWARE LAYER

Middleware Layer or so-called Service management layer, it makes the processing of
information possible. It provides the important functions like device modeling, management,
analytics and configuration, etc. The aim of this layer is to receive and store the date form
the pervious layer, processes the information, make decisions and then deliver it to the
application layer. It provides information in the form of events for example good temperature
for the patient`s health condition. These data may require an immediate response to
emergency`s station. In addition, hiding technological detail from the user to enable the
developers to develop applications more easily, quickly and without the complexity. At this
layer, some middleware protocols may be included in order to make applications compatible
with different objects without consideration according to hardware platforms. Cloud
computing is an important platform of Middleware Layer that it may analyze the data [14].
365
(ICoIT'17)
10thof April, 2017.
2.4. APPLICATION LAYER

Application layer provides different services to the end user after the processing and
analyzing the data from Middleware Layer. For instant, after a patient measures his blood
glucose and temperature the application layer can provide these measurements to who asked
for them. The application Layer presents several of fields like smart home, smart city, smart
building, smart healthcare, education, industry, etc. [15].
2.5. BUSINESS LAYER

As the name suggests, its purpose is to build a business model. The data from the
application Layer will be processed more in order to provide significant services to meet
customer needs. In addition, Business layer can provide different functionalities such as
analyzing, monitoring and evaluating data. It also enhances the service by comparing the
output of each layer with the expected output, which can make money from the service
provided. [15].
There are some problems with IoT architecture which still get the attention from
researchers:
Vertical vs. Horizontal [1]:
In vertical, operation and control of various services of different layers (from the
perception layer to the application layer) should studied.
In horizontal, different devices and technologies are used, so that further
enhancements of user-centric communication capabilities should considered (one object
to another object).
Security:
Security must be included across IoT layers from object layer to application layer.
Security is an important issue as the integrity of data, since the integrated physical sensors
are able to access applications organizations and domains.
3. PROTOCOLS
In this section, an overview is provided of the most common protocols specified for IoT,
by giving their functionalities and their challenges.
3.1. APPLICATION PROTOCOLS

Two of the most common protocols are motioned, that are used for constrained devices:
3.1.1 MESSAGE QUEUE TELEMETRY TRANSPORT (MQTT):
The history of MQTT was introduced by Andy Stanford-Clark and Arlen Nipper in 1999
and then MQTT v3.1.1 becomes an open standard protocol by OASIS (Advancing Open
Standards for the Information Society) in 2014. It is a lightweight application protocol
optimized for devices with limited constrained such as power and bandwidth. Also, it is
designed for IoT applications and M2M [16].
MQTT runs over TCP/IP and based on publish/subscribe model topic-based. In this mode,
it consists of two types of clients and server. Each client can be publisher or subscriber to
specific topic, while server or called broker mediates publishers and subscribers and pass
messages form publishers to subscriber. The subscribers receive messages related to
subscribed topics [17].
366
(ICoIT'17)
10thof April, 2017.
MQTT is better than the HTTP protocol because it has a lower overhead, a synchronous
and reliable with different levels of quality of services used for reliable networks. QoS level
0, the client receives the message at most once, an acknowledgement not used in this level,
and the server does not retry sending the message at all. QoS level 1; the server delivers a
message to the client at least once, here an acknowledgement used in this level. QoS level 2,
this level is used when lost or duplicated of the message that is accepted by the client, this
requires 4-way handshake to deliver the message at exactly once, hence this coz increase in
the overhead [16].
3.1.2 CONSTRAINED APPLICATION PROCOLCOL (CoAP):

CoAP is web transfer protocol based on the Representational State Transfer [REST]
architecture for Resource-Oriented such as 6LowPAN networks and nodes. In 2010, CoAP
was created by CoRE (The Constrained RESTful Environments) working group. The aim of
this protocol was to optimize application for IoT and M2M by reducing message overhead
[18]. CoAP uses the same functionality as the HTTP protocol (supports request/response
model), but CoAP adds some additional features such as low processing power and energy
consumption; so that the protocol can be suitable for constrained environments such as
sensors/actuators. Thereby, it will offer a new feature like multicast,
asynchronous/synchronous message exchange, low-bandwidth and reliable transmission (the
bandwidth and data can be exchanged via the client-server model). CoAP runs overTCP so
that it will support unicast only, while running over UDP, it will support both unicast and
multicast [19]. All above features make the use of CoAP more favorable than using
traditional HTTP [20].
CoAP supports publish/subscribe model by using messaging queue (MQ-CoAP). Here,
MQ-CoAP is using URI (Universal Resource Identifier) to access resource on the destination
instead of using topics like MQTT. each specific resource has a unique URI. CoAP has two
layers: request/response layer and transaction layer. In the Request/Response layer, it is
responsible for influencing the resource by using these commands (GET, PUT, POST and
DELETE) to provide resource oriented between client and server. Publishers send the data
to subscribers which request the data through URI [21]. While in the transaction layer, it is a
reliable because it has multiple different levels of QoS. This layer is used to exchange a
single message between client and server. There are four types of messages that require QoS.
The CON (Confirmable), is required to return an acknowledgment to the client that could be
sent as synchronous or asynchronous. The NON (Non- Confirmable), does not require to
return an acknowledgment. The ACK (Acknowledgment), is used to confirm the reception
of a CON message type. The RST (Rest), is used if the message cannot be processed.
HTTP-CoAP mapping offers transformation between Client HTTP and CoAP server by
allowing a HTTP client to access resource at the CoAP server using a proxy, but if the process
of mapping is reversed, a reverse proxy will be used [20].
CoAP is designed with no security feature, therefore the IETF proposed two protocols to
secure CoAP: DTLS and IPsec. Recently, IPSec stops securing CoAP. CoAP with DTLS
(Datagram Transport Layer Security) is similar to HTTPs. DTLS functionality is similar to
TLS (Transport Layer Security) functionality. However, the difference between TLS and
DTLS is that DTLS runs on top of UDP. DTLS provides authentication, data integrity,
confidentiality and automatic key management. But the problem with DLTS is that it's
367
(ICoIT'17)
10thof April, 2017.
unsuitable for constrained devices because it does not support multicast, which is the
important role in CoAP [22]. Some researchers have proposed some solutions to that problem
[23].
There are many challenges that CoAP that may face, some of them are mentioned:
Congestion control: CoAP runs on the top of UDP protocol, UDP does not
provide reliable or congestion control mechanism, therefore, CoAP must deal with the
congestion control mechanism by itself. It has a basic mechanism and it is achieved by
"Confirmable" and "NON Confirmable" messages. However, this mechanism could
not be handled with constrained resources. Some researchers have proposed some
solution to that problem [24].
Size of MTU: constrained environments have very small size of MTU. The
frame size of IEEE 802.15.4 is 127 bytes as a maximum, while the size of MTU of
IPv6 is 1,280 bytes as a minimum, but the IP network supports MTU 1500 byte or
higher. 6LowPAN provides an adoption layer between the link layer and the network
layer to provide fragmentation method and header compression to recover the above
problem. Some researchers have proposed some solution to that problem [25].
TABLE 1: MQTT vs. CoAP
MQTT CoAP
Messaging Multipoint-to- Point-to-Point
transformation Multipoint communication
commination
Topics URI
Transport TCP-Based UDP-Based
Architecture Publish/Subscribe Request/Response
model model
Layers Single layered Two sublayer:
Request/Response
and Transaction
Security SSL/TLS SSL/DTLS
Reliability 3 QoS levels 4 QoS levels
Performance Lower delays Lower overhead and
analysis lower packet loss
3.2. INFRASTRUCRE PROTOCOLS

3.1.3 ROUTING PROTOCOL FOR LOW POWER AND LOSSY NETWORKS
(RPL)
RPL is a distance vector routing protocol for LLNs (Low power and Lossy Networks).
LLNs consist of thousands of embedded devices which are constrained like limited power,
memory and processing resource. Routers are interconnected by lossy links; however, these
links have a problem on a low data rate. In addition, LLNs support three types of traffic:
point-to-point (transfer traffic between nodes), point-to-multipoint (transfer traffic from root
368
(ICoIT'17)
10thof April, 2017.
to other nodes) and multipoint-to-point (transfer traffic from nodes to root) [26]. For the
above reasons, LLNs must have specific routing requirements. The previous routing
protocols were like OSPF, IS-IS, AVOD and OSLR but all these protocols did not meet LLN
routing requirements. IETF forms a new Working Group to meet LLN solutions. The
Routing Over Low power and Lossy (ROLL) proposed an RPL routing protocol for various
application scenarios like healthcare, industrial, urban, smart home, smart cities and for
WSNs and IoT. RPL was standardized in 2011 and defined as RFC6650. RPL is committed
to IPv6-based to prevent fragmentation and intends to support a different kinds of link layer.
RPL supports various technologies such as Wi Fi and IEEE 802.15.4 [27]. RPL routing has
two attributes that are used for path calculation: the first attribute is constraints, the node and
link clipped from the path that did not commit the constraint; the second attribute is metrics
that determine the least-cost path [28].
Objective Function uses metrics and constraints to determine the best path [29]. RPL
mechanism is based on Directed Acyclic Graphs (DAGs) which consists of Destination
Oriented Directed Acyclic Graphs (DODAGs). DODAGs have one root that is called LBR
(LowPAN Border Router) or it can have a multiple root in the network. In addition, they
have a preferred parent node in which each router can selects its parent. Furthermore, they
have a several sink routers that are called a leaf node. The DODAG root advertises the routing
metrics and filters the link to look as if it fulfills the properties [30].
RPL supports new messages of Internet Control Message Protocol (ICMPv6) for the IPv6
specification to exchange related information [31]. These messages are DIO (DODAG
Information Object), DIS (DODAG Information Solicitation) and DAO (DODAG
Information Object). RPL uses Trickle algorithms to exchange messages dynamically
between nodes. [50]. The DIO message is used to advertise the table of neighboring nodes
of the graph, determine distance from each node to the root based on certain rules (objective
function, DAG characteristics, advertised path cost and local policy), and to make a decision
whatever the node is to join the network or not. The DAO message is used to advertise the
upward traffic and downward traffic. The DIO message is used by a node to collect DIO
messages from neighboring nodes.
3.3. IPv6
IPv6 was created by the IETF and defined as RFC-791. It is considered as an interactive
of IPV4. It is similar to IPv4, but adds some additional features such as the length of IP
address is longer than IPv4 and it is extended from 32bit into 64bit. IPv6 becomes the future
of IoT.
The following aspects of IPv6 make it suitable to be used by IoT:
The IP is a requirement to any connection that is used to transfer data between devices
over the internet.
It offers scalability by providing 2^128 unique address space which covers all old and
future devices.
It supports the Network Address Translation (NAT) which enables multiple users to
share the same public IP address that solves the limitation on IPv4 problem.
It provides solutions to support mobility of end-to-end devices and routing mobility of
the network.
369
(ICoIT'17)
10thof April, 2017.
It provides the stateless mechanism, in another word the nodes can define their address
and hence reduce the cost and configuration effort.
If the number of devices is increased, then a security issue is considered, such as network
attack. For example; in the case of smart house, there will be a security risk on information,
therefore IPv6 provides security such as IPSec and VPN (Virtual Private Network) where
the last is supported by all compatible devices.
ISP (Internet Service Provider) may suffer from a problem, when the network is not
supporting IPv6. Some researchers have proposed some solutions to that problem [32].
3.4. IEEE802.15.4
The IEEE802.15.4 is an IETF standardized protocol for constrained devices such as
sensors/actuators, designed for power consumption, low data rate, low cost, low complexity
and a few meters of radio frequency; specified for low-rate wireless personal area networks
(LR-WPANs), IoT and M2M. It designates a physical layer and Medium Access Control
(MAC) sublayer [33]. IEEE802.15.4 supports two types of devices: a full-function device
(FFD) and a reduced-function device (RFD). The FFD device serves as a personal area
network (PAN) coordinator or just a coordinator. On the other hand, the RFD cannot serve
as PAN coordinator or just a coordinator. The RFD is designed for simple node like a switch
or a passive infrared sensor with limited resources and memory. The network has at least one
FFD and multiple RFD which can communicate with each other. IEEE802.15.4 consists of
two types of topologies and operates one of them: the star topology or the peer to peer
topology (or called mesh). The star topology contains of one central FFD and some RFD
devices. The FFDs control, initiate, terminate and manage the other RFD devices. The star
topology offers services in the fields such as smart home, healthcare and computers. While
the mesh topology contains one FFD coordinator and multiple RFD devices which can
communicate with each other through intermediate nodes. The mesh topology is more
complex than star topology. It supports applications like WSN and industrial. Cluster tree
topology is extended from mesh; the most devices are FFD. The other devices from outside
the network cannot associate with cluster tree due to RFD devices at the end of the cluster
[34].
There are other protocols developed to support IEEE802.15.4 such as ZigBee, WiFi and
WirelessHART [35].
4. MICROCONTROLLER
In this section, we consider the microcontroller as one of the most important and
fundamental elements of IoT, by giving the main types used by researchers, where the
selection of microcontrollers depends on many factors to choose the suitable one for a project
for example, processing speed, energy computation and the supported programming
languages, also we mention some important characteristics for micro-controllers, with
comparison between some important versions which related to micro-controllers and to make
the selection and learning easier. Also, some simulators programs which are available are
mentioned with books.
370
(ICoIT'17)
10thof April, 2017.
4.1. Arduino
Arduino is an open-source microcontroller used by teachers and students to prototype a
powerful, low cost and flexible electronic project. Arduino consists of hardware and software
that is called IDE (Integrated Development Environments). It takes input from different
sensors like light, temperature, degree of flex, pressure, proximity, acceleration, carbon
monoxide, radioactivity, humidity and barometric pressure, and then sends them to the
Arduino board to convert it to digital output. Arduino`s software used to control the board to
take different actions. It supports variety of products like: IoT application, wearable, entry
level and enhanced features. In addition, it is compatibles with shields that are used to provide
additional capabilities such as connecting to the Internet (Wi-Fi, Bluetooth, 6LowPAN) [36].
TABLE 2: Arduino features
Board Features Communication I/O

Arduino ARM, 7- ETH, BLE, Wi- 6 ADC
PRIMO 15V, Fi, CAN
64MHZ
Arduino ARM, 3.3, ETH, BLE, Wi- 2 ADC
STAR - 180MHZ Fi, 2CAN
OTTO
Arduino UNO AVR, ETH, BLE, Wi- ADC
WIFI ESP8266, Fi, CAN
5V, 16-
80MHZ
Arduino ARM, 5V, ETH, Wi-Fi, -
INDUSTRIAL 16MHZ CAN
101
4.2. Raspberry Pi
Raspberry Pi is a small-sized computer and low-cost device intended for the education
field. It is used by people to improve programming skills. The Raspberry Pi advances the
Arduino microcontroller. The physical hardware is an open source and run on Linux,
Windows and Risc OS. The board consists of important things: processor, RAM and ports.
There are different types of models. The first is the Raspberry Pi A, this model is low price,
has only one USB port, uses less power computation and only 256 MB of RAM. the
Raspberry Pi A is upgraded to the Raspberry Pi A+, suitable for robotics and embedded
system. The second is the Raspberry Pi B, this model has two ports one for the Ethernet and
the other one for 512 MB of RAM. The Raspberry Pi B is also upgraded to the Raspberry Pi
B+, this model has a double number of USB ports and improved power consumption. The
third is the Raspberry Pi 2, is the replacement of B+, the important features including 1GB
of RAM and 4 USB ports. The Raspberry Pi 3 is the last model, supports important features
like it has built such as Wi-Fi IEEE 802.11n and Bluetooth [37]. There are several book`s
guides for the Raspberry Pi in [38]. Also, Raspberry Pi simulators in [39].
371
(ICoIT'17)
10thof April, 2017.
4.3. TelosB
TelosB sensor is an open-source platform used in WSNs. The platform includes: USB
programming and data collecting capability, microcontroller, transceiver, 10KB RAM,
250Kbps and IEEE802.15.4 radio with antenna built-in and low power consumption (which
offers a long-life battery and fast wake ups from deep sleep). It uses an open-source, simple
and energy efficient TinyOs image, which was developed by US Berkeley. It is integrated
with five sensors: Temperature, Sound, Solar radiation, Light and Humidity sensors [40].
4.4. NodeMCU-ESP8266
The Node MCU is an open-source microcontroller that is used in IoT applications and
was designed by Espressif system. It is based on the ESP8266 Wi-Fi SoC (System on Chip)
and ESP-12 module. Node MCU was created on 13 OCT 2014, Arduino developed their IDE
(Integrated Development Environments) software to be compatible with Node MCU. It uses
two programming languages: The Lua scripting language and C/C++ Arduino language.
TABLE 3: NodeMCU types
Generation Version Features

1st 0.9V1 ESP-12,
4MBflash,
47mm*31mm
2nd 1.0V2 ESP-12E,
4MBflash
3nd 1.0V3 The as V2 but it
has two reserve
pins for USB
power out and the
other for an
additional GND.
4.5. Onino-OMEGA 2
Omega 2 is a tiny microcontroller, cloud integrated and designed for building IoT projects.
It is a Linux computer server with Wi-Fi built-in and designed for connectivity and on-board
flash memory. It is a small size and has power efficiency as Arduino and flexible as
Raspberry Pi. It has new expandable docks like Expansion Dock, Mini Dock, Power Dock,
Relay Expansion Dock, Bluetooth, Cellular, GPS and Arduino Dock R2 (make use of
Arduino shield). It is a simple and an affordable to start with it. It offers users an application
store to get solutions for IoT. It can be programmed with many languages like C, C++,
Node.js, Python and PHP. It has nine sensors: humidity, temperature, gyroscope, light, LED
driver, potentiometer, pressure, gas and buzzer [41].
TABLE 4: Onino types
372
(ICoIT'17)
10thof April, 2017.
Omega Omega2 Omega2

Plus
Memory 64MB 64MB 128MB
CPU 400MHZ 580MHZ 580MHZ
Storage 16MB 16MB 32MB
Price 19$ 5$ 9$
USB 2.0 2.0 2.0
Wi-Fi b/g/n b/g/n b/g/n
GIPO 18 15 15
PWM - 2 2
I2C - 1 1
SPI - 1 1
I2S - 1 1
4.6. Intel Galileo

The Intel Galileo is the first microcontroller based on the Intel architecture. It is designed
to be compatible with Arduino software and hardware. It uses software Arduino development
IDE, which makes easier getting started. The hardware is operating at 3.3v or 5v for Arduino
shield. It is the best choice for students and professional developers who seek for simple,
efficient and low-cost board due to its features like: a full sized mini-PCI Express slot, Micro-
SD slot, 100MB Ethernet port, USB host port, USB TTL UART header and 8Mbyte NOR
flash. Galileo supports many programming languages such as Python, Videio4linux, Secure
Shell (SSL), Advanced Linux Sound, OpenC, Node.js and Advanced Linux Sound
Architecture (ALSA). Galileo uses the default Linux image [42].
5. INTERGRATING CLOUD AND IOT

The new revolution within the field of devices and objects of the real world that these will
integrate into the virtual world. In 2011, the number of connected devices was about 12.5
Billion and they were exceeded the number of people on the world [43]. Since the number
of devices has increased, there will be Big Data. Storing that Big Data temporally will not be
the solution for IoT. There must be another solution with low-cost and light-weight to store
that big data as permanently in the cloud [44]. Integrating the cloud and internet of things,
it becomes the new wave of computing. We call this new paradigm Cloud of Everything
(CoE). The two terms cloud and IoT have grown exponentially. Furthermore, their
architectures are different from each other, IoT uses Cloud to improve its processing,
computing, energy and power. Many researchers are interested in this integration and
proposed a new solution of architecture [15].
373
(ICoIT'17)
10thof April, 2017.
6. COE ISSUES
In this section, some problems and difficulties are explained which still appear when
integrating between IoT and cloud.
6.1. PROTOCOLS
The protocols represent a major role in complete fulfillment of the sensing requirements.
They consider as a data tunnel between IoT devices and the outside world [45]. The sensing
devices use IoT protocols like IEEE802.15.4 or IEEE802.11 to transmit data to the person
concerned. However, some gateways do not support some of the protocols that the sensor
being used. People at the most time choose a cheap and easy sensor to work with. If a sensor
is added to the network, it cannot be guaranteed that sensor will be compatible with the
gateway. It must consider whether the gateway supports the standardized protocols [15].
Using MAC (Medium Access Control) protocols make the system work efficiently. MAC
protocols (collision free) have been proposed for several areas like: DMA, TDMA, CSMA
and FDMA. However, these schemes are impossible with billions of available IoT devices.
This becomes an issue, which requires further study [46].
6.2. ENERGY EFFICIENCY

In the near future, WSNs will be a part of IoT. When sensor nodes join the network and
sense data this will lead to the appearance of power consumption problems. A WSN has four
basic elements: a sensing unit, a processing unit, a transmitting/receiving unit, and a power
unit. The challenge is to design energy-efficiency sensors instead of the need of battery
replacement in a traditional way. For example, design new video sensing, encoding and
decoding to consume low-power. In general, the encoding is more complex than decoding
because the encoders have to compress the video efficiently by analyzing the redundancy in
the video. This process is acceptable because the encoding process is done only once while
the decoding process takes multiple times [47].
The inappropriate way is to have a sensor with batteries, especially, if there are billions
of sensors all over the world, it`s difficult to change the huge number of batteries. The most
suitable way is to have sensors with permanent power supply, as well as taking advantage of
the environment that generate electricity like vibrations, light, and airflow. Furthermore,
sciences publicized a new chip called Nan-generator that used body movement to generate
electricity. Moreover, enable sleep-mode [15].
6.3. IDENTITY MANAGEMENT

There are many "things" communicate over the Internet; these are becoming part of IoT
in order to understand, collaborate and to do their job of sensing. There are two types of
"things", the first type, dies over a short time, which doesnt need an identity. The other type
requires identification, for example, mobile nodes in smart city with motivation. The
identifying object in IoT is one of the main problems in the resource constraint like energy,
lifetime, point-to-point delay, memory, routing overhead latency and bandwidth [47].
Identity management (IDM) is defined as the management of things involves identifying
devices in the network to make things distinguishable and assigning identifiers. It provides
authentication, authorization and polices and controls access to resources. Normally, things
have only one identity, but sometimes they have multi identifiers. Various components of
374
(ICoIT'17)
10thof April, 2017.
the IDM system include; directory services, access management and Password
administration [48].
6.4. IPv6
In spite of the IPv6 gets widely used to identify devices, but some existing companies,
technologies and customers still use the IPv4. The problem comes with development of IPv6
to be compatible with IPv4 devices [49]. A solution is to enable seamless communication by
using a transition technology that is called dual stack routers. Nodes may have dual stack
support both IPv4 and IPv6 packets. In addition, use tunneling techniques in cloud
virtualization: 6to4, Teredo, and ISATAP.
6.5. QUALITY OF SERVICE

By default, heterogeneous devices support multi traffic types and multi services; a single
network can handle with all services and traffic without QoS. Communication networks
offer to the end user three types of multimedia: high quality and smooth video, no drop calls
and real time voice, and data. As IoT expanded, there are sensors provide different services
at the same time; this leads to a bottleneck, hence, QoS becomes an issue. Achieving QoS
are guaranteed by maximizing the bandwidth, control delay (the limited time that a packet
takes to reach to the destination), jitter (the variation in the end-to-end delay of received
packet) and packet loss (the number of packet that failures to reach the destination) to meet
customer`s need [50].
6.6. DATA LOCATION

Since cloud and IoT support data mobility, location of data becomes an important issue
for several reasons; for user`s perspective, it does not matter to know the location of their
data. For example, people share their photos on social networks and they do not concern
where their data are stored. However, some systems have a sensitive data that preferred to
know the location; and in some cases, they specify their own location. The service provider
side should also take care of systems security. In addition, they may take contract with
another cloud, so that when the data moved from one place to another, it will not become a
problem. The cloud services provider may depend on several factors including the cost of
data center (researchers proposed a system reducing the cost [51]), the temperature of servers
and dynamic routing may help to optimize resource and reduce costs [52]. Another
consideration, the time that takes to access the data should be minimalized; the data should
be stored in closest location to the user [15].
7. CONCLUSION
IoT is a new concept and becoming the future of Internet, which a huge number of data
generated from things and shared across Internet around the world. A billion of devices,
applications and new technologies will connect with each other to improve the quality of life
more than anything did before. This paper shows an overview of IoT and about its history
and application domains. Some remarks are concluded related to this topic; IoT architecture
has been presented to understand the role of each IoT layers to know which the IoT
components and technologies are suitable with each layer. IoT protocols functionality existed
375
(ICoIT'17)
10thof April, 2017.
nowadays have been explained their functionality with open research challenges. Finally, for
providing new services arising from this integration to users, Integration cloud with IoT has
been emerged. This integration termed as Cloud of Everything, including some issues which
have been explained.
In further, we are looking into surveying the next-generation technologies and networks.
The 5G networks have been expected to handle trillions of objects by 2020 more. However,
it will be a part of IoT; connecting all mobile nodes and fixed devices with each other. New
challenges will arise from this integration that should be studied study.
8. REFERENCES
[28] A. H. H. Ngu, M. Gutierrez, V. Metsis, S. Nepal, and M. Z. Sheng, IoT Middleware:

A Survey on Issues and Enabling technologies, IEEE Internet of Things Journal, pp.
11, 2016.
[29] L. Durkop, B. Czybik, and J. Jasperneite, Performance evaluation of M2M protocols
over cellular networks in a lab environment, 2015 18th International Conference on
Intelligence in Next Generation Networks, 2015.
[30] S. N. Srirama, "Web services and cloud enabling mobile Internet of Things" CSI
Transactions on Information and Communication Technology, April 2017.
[31] D. Kelaidonis, P. Vlacheas, V. Stavroulaki, S. Georgoulas, K. Moessner, Y. Hashi, K.
Hashimoto, Y. Miyake, K. Yamada, and P. Demestichas, Cloud Internet of Things
Framework for Enabling Services in Smart Cities, Designing, Developing, and
Facilitating Smart Cities, pp. 163191, Jun. 2016.
[32] O. Vermesan and P. Friess, Eds., Internet of things: Converging technologies for smart
environments and integrated ecosystems. Aalborg: River Publishers., 2013.
[33] K. Rose, S. Eldridge, and L. Chapin, "The internet of things: An overview", The
Internet Society (ISOC), pp. 150, Oct. 2015.
[34] J. Hller, V. Tsiatsis, C. Mulligan, S. Karnouskos, S. Avesand, and D. Boyle, From
Machine-to-Machine to the Internet of Things: Introduction to a New Age of
Intelligence. Amsterdam, The Netherlands: Elsevier, 2014.
[35] J. Manyika, M. Chui, P. Bisson, J. Woetzel, R. Dobbs, J. Bughin and D. Aharon. The
Internet of Things: mapping the value beyond the hype. June, 2015.
[36] F. Hu, Security and privacy in Internet of things (IoTs): Models, Algorithms, and
Implementations. United Kingdom: CRC Press, 2016.
[37] F. DaCosta, Rethinking the Internet of things a scalable approach to connecting
everything. New York, NY: ApressOpen, 2013.
[38] M. Yun and B. Yuxin, Research on the architecture and key technology of Internet of
Things (IoT) applied on smart grid, 2010 International Conference on Advances in
Energy Engineering, 2010.
[39] M. A. and S. T., "Internet of Things: Architecture, Security Issues and
Countermeasures", International Journal of Computer Applications, vol. 125, no. 14,
pp. 1-4, Sep. 2015.
[40] R. Khan, S. U. Khan, R. Zaheer, and S. Khan, Future Internet: The Internet of Things
Architecture, Possible Applications and Key Challenges, 2012 10th International
Conference on Frontiers of Information Technology, 2012.
376
(ICoIT'17)
10thof April, 2017.
[41] M. Aazam, I. Khan, A. A. Alsaffar, and E.-N. Huh, Cloud of Things: Integrating
Internet of Things and cloud computing and the issues involved, Proceedings of 2014
11th International Bhurban Conference on Applied Sciences & Technology (IBCAST)
Islamabad, Pakistan, 14th - 18th January, 2014, 2014.
[42] "FAQ - Frequently Asked Questions | MQTT", Mqtt.org. [Online]. Available:
http://mqtt.org/faq. [Accessed: 9- March- 2017].
[43] S. Lee, H. Kim, D.-K. Hong, and H. Ju, Correlation analysis of MQTT loss and delay
according to QoS level, The International Conference on Information Networking
2013 (ICOIN), 2013.
[44] Z. Shelby, K. Hartke, C. Bormann, and B. Frank, draft-ietf-core-coap-18 - The
Constrained Application Protocol (CoAP), Internet Eng. Task Force (IETF), Fremont,
CA, USA, 2013. [Online]. Available: https://tools.ietf.org/html/draft-ietf-core-coap-
18. [Accessed: 9- March- 2017].
[45] W. Colitti, K. Steenhaut, and N. D. Caro, "Integrating wireless sensor networks with
the web," Extending the Internet to Low power and Lossy Networks (IP+ SN 2011)
,2011.
[46] M. R. Palattella, N. Accettura, X. Vilajosana, T. Watteyne, L. A. Grieco, G. Boggia,
and M. Dohler, Standardized Protocol Stack for the Internet of (Important) Things,
IEEE Communications Surveys & Tutorials, vol. 15, no. 3, pp. 13891406, 2013.
[47] M. Koster, J. Jimenez, and rfcmarkup version 1, "Publish-subscribe broker for the
constrained application protocol (CoAP)", Tools.ietf.org. [Online]. Available:
https://tools.ietf.org/html/draft-koster-core-coap-pubsub-05. [Accessed: 9- March-
2017].
[48] S. Raza, H. Shafagh, K. Hewage, R. Hummen and T. Voigt, "Lithe: Lightweight Secure
CoAP for the Internet of Things", IEEE Sensors J., vol. 13, no. 10, pp. 3711-3720, Oct.
2013.
[49] H. J. Ban, J. Choi, and N. Kang, Fine-Grained Support of Security Services for
Resource Constrained Internet of Things, International Journal of Distributed Sensor
Networks, vol. 2016, pp. 18, 2016.
[50] A. Betzler, C. Gomez, I. Demirkol and J. Paradells, "CoAP congestion control for the
internet of things", IEEE Communications Magazine, vol. 54, no. 7, pp. 154-160, Jul.
2016.
[51] A. Betzler, C. Gomez, I. Demirkol, and J. Paradells, CoAP congestion control for the
internet of things, IEEE Communications Magazine, vol. 54, no. 7, pp. 154160,
2016.
[52] J. Hui, D. Culler, and V. Manral, An IPv6 Routing Header for Source Routes with the
Routing Protocol for Low-Power and Lossy Networks (RPL), 2012.
[53] "Routing Protocol for LLN (RPL) Configuration Guide, Cisco IOS Release 15M&T",
Cisco. [Online]. Available: http://www.cisco.com/c/en/us/td/docs/ios-
xml/ios/rpl/configuration/15-mt/rpl-15-mt-book.html. [Accessed9- March- 2017].
[54] P. Thubert, "RPL Objective Function Zero", 2012. [Online]. Available:
https://tools.ietf.org/html/draft-ietf-roll-of0-19. [Accessed: 9- March- 2017].
[55] JP. Vasseur, M. Kim, K. Pister, N. Dejean, and D. Barthel, draft-ietf-roll-routing-
metrics-19 - Routing Metrics Used for Path Calculation in Low-Power and Lossy
Networks, Internet Engineering Task Force (IETF) Request for Comments, No. RFC
377
(ICoIT'17)
10thof April, 2017.
6551, 2011. [Online]. Available: https://tools.ietf.org/html/draft-ietf-roll-routing-

metrics-19. [Accessed: 9- March- 2017].
[56] A. Conta and S. Deering, Internet Control Message Protocol (ICMPv6) for the
Internet Protocol Version 6 (IPv6) Specification, 2006.
[57] P. Levis, Clausen, T. Heide, J. Hui, O. Gnawali, J. Ko and rfcmarkup version 1, "The
trickle algorithm", 2011. [Online]. Available:
https://tools.ietf.org/html/rfc6206#section-1. [Accessed: 9- March- 2017].
[58] P. Kinney, IEEE 802.15 WPAN Task Group 4 (TG4). [Online]. Available:
http://www.ieee802.org/15/pub/TG4.html. [Accessed: 9- March- 2017].
[59] S. Ziegler, P. Kirstein, L. Ladid, Antonio S. and Antonio Jara, "The Case for IPv6 as
an Enabler of the Internet of Things - IEEE Internet of Things", Iot.ieee.org, 2015.
[Online]. Available: http://iot.ieee.org/newsletter/july-2015/the-case-for-ipv6-as-an-
enabler-of-the-internet-of-things.html. [Accessed: 9- March- 2017].
[60] IEEE 802 Working Group, IEEE Standard for Local and Metropolitan Area
NetworksPart 15.4: Low-Rate Wireless Personal Area Networks (LR-WPANs),
IEEE Std, pp. 1314, Sep, 2011.
[61] J. Song, S. Han, A. Mok, D. Chen, M. Lucas, M. Nixon, and W. Pratt, WirelessHART:
Applying Wireless Technology in Real-Time Industrial Process Control, 2008 IEEE
Real-Time and Embedded Technology and Applications Symposium, 2008.
[62] Arduino, "Arduino - Introduction", Arduino.cc. [Online]. Available:
https://www.arduino.cc/en/Guide/Introduction. [Accessed: 9- March- 2017].
[63] "Introduction to Processing | Raspberry Pi Learning Resources", Raspberrypi.org.
[Online]. Available: https://www.raspberrypi.org/learning/introduction-to-
processing/. [Accessed: 9- March- 2017].
[64] E. Upton and G. Halfacree, Raspberry pi user guide. Chichester, United Kingdom:
Wiley-Blackwell (an imprint of John Wiley & Sons Ltd), 2014.
[65] A. Lofquist, "EmulationStation", Emulationstation.org. [Online]. Available:
http://emulationstation.org/gettingstarted.html. [Accessed: 9- March- 2017].
[66] "Telosb Sensors | Telosb Motes |Telosb", Telosb Sensors | Telosb Motes |Telosb, 2014.
[Online]. Available: https://telosbsensors.wordpress.com/. [Accessed: 9- March-
2017].
[67] O. Corporation, "Onion invention platform for IoT". [Online]. Available:
https://onion.io/. [Accessed: 9- March- 2017].
[68] I. Corporation, "Intel Galileo gen 2 development BoardEmpower your prototype",
Intel. [Online]. Available: http://www.intel.com/content/www/us/en/do-it-
yourself/galileo-maker-quark-board.html. [Accessed: 9- March- 2017].
[69] P. N. Howard, "How big is the Internet of Things and how big will it get?" The
Brookings Institution, Jun, 2015.
[70] D. Evans, The Internet of Things How the Next Evolution of the Internet is
Changing Everything, White Paper-Cisco Internet Business Solutions Group, April
2011.
[71] K. Chen, "Challenges and opportunities of internet of things." 17th Asia and South
Pacific Design Automation Conference. IEEE, 2012.
[72] P. N. Mahalle and P. N. Railkar, Identity Management for Internet of Things. Vol. 39.
River Publishers, 2015.
378
(ICoIT'17)
10thof April, 2017.
[73] R. Shaikh and M. Sasikumar, Identity Management in Cloud Computing,

International Journal of Computer Applications, vol. 63, no. 11, pp. 1719, 2013.
[74] J. Gubbi, R. Buyya, S. Marusic and M. Palaniswami, "Internet of Things (IoT): A
vision, architectural elements, and future directions", Future Generation Computer
Systems, vol. 29, no. 7, pp. 1645-1660, Sep. 2013.
[75] Y. Cui, P. Wu, M. Xu, J. Wu, Y. Lee, A. Durand and C. Metz, "4over6: network layer
virtualization for IPv4-IPv6 coexistence", IEEE Network, vol. 26, no. 5, pp. 44-48,
Sep. 2012.
[76] M. Aazam and E.-N. Huh, "Impact of ipv4-ipv6 coexistence in cloud virtualization
environment," annals of telecommunications - annales des tlcommunications, vol.
69, no. 9-10, pp. 485496, Aug. 2013.
[77] "Quality of Service (QoS)", Cisco. [Online]. Available:
http://www.cisco.com/c/en/us/products/ios-nx-os-software/quality-of-service-
qos/index.html. [Accessed: 9- March- 2017].
[78] P. T. Jaeger, J. Lin and J. Grimes, "Cloud Computing and Information Policy:
Computing in a Policy Cloud?", Journal of Information Technology & Politics, vol. 5,
no. 3, pp. 269-283, Oct. 2008.
[79] Z. Mahmood, Data Location and Security Issues in Cloud Computing, 2011
International Conference on Emerging Intelligent Data and Web Technologies, 2011.
379
(ICoIT'17)
10thof April, 2017.
Internet of Things Based Cloud Smart Monitoring for

Asthma Patient
Tabarek G. AL-Jaf
Networks Engineering Department, College of Information Engineering, Al-Nahrain University
Baghdad - Iraq
tabarekaljaaf@yahoo.com
Emad H. Al-Hemiary
Baghdad - Iraq

The Internet of Things (IoT) is becoming the sinew of today
Article History: and future technology. It can be exploited to make life much
Received: 20 March 2017 easier for human beings by taking the advantage that they
have their own smart devices connected to the Internet. From
this perspective, many society related issues like chronic
Published: 10 April 2017 diseases can be solved. In this paper, an IoT based remote
monitoring for Asthma patients is designed and implemented.
Keywords: Internet of The work is initiated with a survey on Asthma chronic disease
Things, IoT, Asthma by asking specialists in this field then discussing how to
monitoring system, IoT- monitor and control it using IoT technology. The prototype
Cloud-HTTP IoT Asthma monitoring system consists of a heart pulse
sensor controlled by ESP8266. The data collected from the
sensor is passed wirelessly to the ESP8266 in real time which
in turn is send to the cloud using HTTP protocol. In the cloud,
a MySQL database is used to store the received readings. The
monitoring facility; hospital, doctor or person can browse the
webpage written in PHP to monitor live update of patient
condition. The work includes simulation of the network using
packet tracer to verify its operation.
1. INTRODUCTION
I oT connects everything together via Internet using sensors, network devices,

communication, and software. Internet of things includes many applications cover whole
life fields offer a cloud storage which is used to expand the availability of data
transferring. It aims to create smart vision of living. With smart network among devices in
lower cost, power with highest security and scalability, the core of this technology is to
connect the virtual reality with physical world this technology represents the revolution of
information technology to have infinite opportunities of impacts and all these benefits help
380
(ICoIT'17)
10thof April, 2017.
to build smart planet [1]. IoT can be embedded in all aspects of life like 'Smart Home', it
includes the ways that controls the environment of the place of living like the lighting,
temperature, home privacy and security this technology can also go wider to the large
environment like cities including sensors for water and gas management, smart garages,
roads, traffics and street lighting as well as mobility these controlled sensors all embedded
in 'Smart City' that manage the quality of life (QoL) [2] and for healthcare field there's also
sensors works to evaluate the health of bodies like pulse sensor, blood sensor which is
represented by wearable devices or in other way as tablets. IoT uses real time transmitting
protocols which is the main challenge to enable real-time transferring information to create
smart system connect a huge network of devices through internet along with the security
feature [3]. There are many protocols that work with the IoT such as MQTT (Message
Queuing Telemetry Transport), CoAP (Constrained Application Protocol), XMPP
(Extensible Messaging and Presence Protocol), AMQP (Advanced Message Queuing
Protocol) and HTTP (Hyper Text Transmission Protocol) and knowing the right type of
protocol is the first design requirements. Today the cloud also become in the top of
requirements to build IoT systems for hosting the data and remotely managing the device
everywhere, it is working as database that contains millions of members to deal with daily.
IoT applications gain data from enormous resources sensors connecting between different
materials and upload them to the cloud. The cloud computing deals with the application layer
as an important layer because it's working to process and analyze the data. IoT devices are
rapidly increasing and this will lead to the 'Big Data' concept where huge amount of data
need to be processed and analyzed [4,5,6]. In this work, IoT is used to monitor Asmtha
patients by sensing their heart beat behaviour and send the recorded pulse counts to the
Cloud. The hospital, doctor, or patient relatives can continuously monitor his state and a
feedback can be send to the sensor to increase recording rate. The rest of the paper is
organized as follows: section 2 gives related researches, section 3 presents the system model,
section 4 and 5 discuss problems and challenges accompany system implementation while
section 6 state the related references.
In [7], the authors present a Module based on ArduinoUNO with GSM shield to connect
to the Internet and send an SMS remotely over the GSM network to the user Mobile phone
by using GSM library to have a connection with the shield without using cloud database to
save the reports of the patient in order to review and diagnosis the patient case. The work in
[8] proposed healthcare system based on IoT and using Arduino UNO platform to read sensor
data and send it to PC platform then processed with an application to display the graph of
heartbeats without mention the Wi-Fi shield or other shields that it used with the Arduino to
transfer collected data to the cloud also the authors did not explain how the remote user could
access and monitor the information from the cloud but display the heartbeats senses on the
local machine. In this paper [9] the authors use architecture of local server to detect
arrhythmia via ECG signal without describing how will be notified the medical experts in
emergent case, so in this paper the data will be send through the Internet to remote server
and there is a website to allow access by authentication person or medical experts to monitor
the patient case as well as the notification send via SMS in abnormal case. Finally, this paper
381
(ICoIT'17)
10thof April, 2017.
[10] describes machine-to-machine M2M connection on local cloud platform to transfer data
to and from local environment while M2M connection can be accrued through WAN
environment.
3. SYSTEM MODEL
According to physicians and references related to Asthma patients, heart rate beats
increases rapidly when the patient suffers from symptoms of breathing difficulties.
Therefore, the decision to use heart beat sensor to monitor heart beat rate. The proposed
network design is devided into three layers: Embedded, gateway, and cloud layers and they
are explained in the following sections with the aid of figure 1:
Cloud
Cloud Layer
Routers
Gateway + Access Point

Type IEEE802.11n WiFi Gateway Layer
Asthma
Embedded Layer
Patient Manager
NodeMCU
FIGURE 1. IoT based Cloud layering model

3.1 EMBEDDED LAYER
This layer includes the end-users (patient and doctor) the first one has wearable device
including pulse sensor used to read the heartbeat rate and send them through the other layers
to the destination (Doctor) for monitering the patient status to make decisions. As stated
before, the pulse reads the patient heart beat rate continuously and periodically each t seconds
configured in the microcontroller software of NodeMCU 'esp8266'. The NodeMCU
'esp8266' microcontroller board equipped with a low cost Wi-Fi module to enable wireless
connection to a nearby access point type IEEE802.11n WiFi. The interface and programming
languages used are the Arduino language and Lua language [11]. As shown in figure 1, the
382
(ICoIT'17)
10thof April, 2017.
manager (PC based software) is used to display the collected data from the sensor and send
them via HTTP protocol to the Cloud to be permanently save for any future analysis. The
core language used in this phase is C++.
3.2 GATEWAY LAYER:
The gateway layer is a router based network which routes the traffic coming from the
microcontroller to the required Cloud. It is evident that when public Cloud is used then this
layer might contain single or multiple hops, and in the proposed model it is fixed to a single
router.
3.3 CLOUD LAYER:

The Cloud layer is used for storing the sensor data using PHP-MySQL and allow users
'doctor or patient' to access remotely or even to mentoring the patient state from any available
device. The basic step is to insert the sensor data into the database after making a successful
connection using HTTP. Since the readings become old with time, an averaging process is
performed to optimize the storage and prevent large amount of data in the database especially
when many sensors use the same database. The monitoring person or even an auto
programmed device receive heart rate values when HTTP connecting is established with the
cloud and this is follow an authentication process. A live graph is displayed on the monitoring
device to highlight the Asthma patient condition in real time.
The verification of IoT-Cloud network is performed using Cisco Packet Tracer as shown
in figure 2. In this simulation, a sensor IoE (Internet of Everything) is used to collect sensed
data and send them to the Cloud denoted by Server0 through Gateway-Router-Router1. The
monitoring is done using a smartphone connected wirelessly to the cloud using wireless
router.
4. PROBLEMS, CHALLENGES, AND RESULTS

Throughout this work, some problems and challenges rises and described as follows. The
first challenge was the type of the sensor used to detect abnormal breathing for asthma
patient. The first idea was to use a breathing sensor named MQ-9 for but then the problem
was that the sensor is designed to measure the rate of gases in the air beside the position of
the sensor which should be placed near the patient mouth. Therefore, the decision was to use
the heartbeat pulse sensor instead which can be easily installed around the arm of the patient,
as an example. The second challenge was the type of protocol to connect to the Cloud.
Different protocols exists and each one has its own features. Therefore, to verify the proposed
architecture, the HTTP protocol is used although other protocols gives better performance.
As shown in figure 3(a), the connection to the cloud is prepared by HTTP protocol and the
data is sent using HTTP GET method. The readings is continuous as long as the
microcontroller is operational this is shown in figure 3(b) and (c).
383
(ICoIT'17)
10thof April, 2017.
FIGURE 2. Network verification using Cisco Packet Tracer.
(a) (b) (c)
FIGURE 3. Sample result of sensor data store in cloud. (a) HTTP connection with GET
method. (b) Sample readings of heart beat rate and (c) MySQL data insertion
5. CONCLUSION
In this paper, a proposal for healthcare system based IoT-Cloud is explained. The system
monitors the Asthma patients through sensing their heartbeat rate and store them for future
analysis. The use of heartbeat sensor is not optimistic but closely reflects the patient
condition since the one of the main asthma symptoms is fast breathing and rapid increase in
heartbeat rate. Therefore, using the heartbeat sensor is a contribution to this field as far as
test results show. For future work, the system could be extended to include multiple type
sensors and include another types of protocols.
384
(ICoIT'17)
10thof April, 2017.
6. REFERENCES
[80] A. Ngu, M. Gutierrez, V. Metsis, S. Nepal, and M. Sheng, "IoT Middleware: A Survey
on Issues and Enabling technologies," IEEE Internet of Things Journal, Vol. 4, Issue 4,
Feb. 2017, pp. 1-20.
[81] D. Minoli, K. Sohraby, and B. Occhiogrosso, "IoT Considerations, Requirements, and
Architectures for Smart Buildings Energy Optimization and Next Generation Building
Management Systems," IEEE Internet of Things Journal, Vol.4, Issue 1, Feb. 2017, pp.
269-283.
[82] D. Chen, N. Zhang, Z. Qin, X. Mao, Z. Qin, X. Shen, and X. Li. "S2M: A Lightweight
Acoustic Fingerprints based Wireless Device Authentication Protocol." IEEE Access,
Vol.4, Feb. 2016, pp. 766-773.
[83] Y. Sun, H. Song, A. Jara, and R. Bie, Internet of Things and Big Data Analytics for
Smart and Connected Communities IEEE Internet of Things Journal, Vol.4, Issue 1,
Feb. 2017, pp. 88-100.
[84] H. Cai, B. Xu, L. Jiang, and A. Vasilakos, IoT-Based Big Data Storage Systems in
Cloud Computing: Perspectives and Challenges, IEEE Internet of Things Journal,
Vol.4, Issue 1, Feb. 2017, pp. 75-87.
[85] "Oracle: Big Data for the Enterprise - White Paper." Jun. 2013.
[86] U. S.U, O C. O and U. M. E, "Heartbeat Monitoring and Alert System Using GSM
Technology," International Journal of Engineering Research and General Science,
Volume 3, Issue 4, 2015, pp. 26-34.
[87] D. Singh and R. Gour, An IoT Framework for Healthcare Monitoring Systems, Int.
Journal of Computer Science and Information Security (IJCSIS), Vol. 14, No. 5, May
2016.
[88] M. Hassanalieragh, A. Page, T. Soyata, G. Sharma, M. Aktas, G. Mateos, B. Kantarci,
and S. Andreescu." Health Monitoring and Management Using Internet-of-Things (IoT)
Sensing with Cloud-Based Processing: Opportunities and Challenges.", IEEE Int. Conf.
on Service Computing, July 2015.
[89] B. Anggorojati, N. Prasad, and R. Prasad, "Evaluation of secure capability-based access
control in the M2M local cloud platform," 10th Int. Conf. on Telecommunication Systems
Services and Applications (TSSA), Oct. 2016.
[90] "ESP8266 WiFi Module Quick Start Guide, [online]
http://rancidbacon.com/files/kiwicon8/ESP8266_WiFi_Module_Quick_Start_Guide_v_
1.0.4.pdf, (Accessed: 19 Mar. 2017).
385
(ICoIT'17)
10thof April, 2017.
MARKOV-BASED DEPLOYMENT APPROACH TO IMPROVE

WSN COVERAGE
Abd alnasir R. Finjan
Software Dep., Information Technology College, University of Babylon - Iraq
naser_reyadh@yahoo.com
Dr. Saad T. Hasson
Software Dep., Information Technology College, University of Babylon Iraq
Saad.aljebori@gmail.com

A "wireless sensor network (WSN)" represents a gathering of
Article History: limited number of sensors that are closely deployed in a
Received: 19 March 2017 recognizing area. The efficiency of any WSNs is heavily
depending on the coverage delivered by the deployed sensors.
Published: 10 April 2017 In this paper, a developed "deployment approach" is

suggested to improve the WSN coverage, connectivity and
Keywords: Sensors reliability. This approach is based on the "Markov process".
deployment, Coverage, The distances between sensor node and its neighboring
WSN, Markov process,
GSO.
sensors are calculated, and then converted to the probabilities
that create the transition matrix. Depending on this transition
matrix the distance and toward movement for each sensor are
estimated in each iteration.
The Simulation results were compared with the GSO results.
Our results show that this deployment approach can provide
high coverage and good reliability.
1. INTRODUCTION
I
n the latest years, many researchers focus on the "wireless sensor networks
(WSNs)"constructions and applications. WSNs have potential applications in
different vital areas. Wireless Sensor networks represent systems that encompass
of certain numbers of wirelessly linked sensors that are spatially deployed in a
certain region. Sensors can be deployed in large numbers due to its accepted cost
and small size [1].
One of the key challenges facing the WSNs applications is how to determine the physical
locations of the deployed sensor nodes. This problem represents an active research topic in
the recent years. The supreme solution for such problem is to provide the sensor nodes by
"Global Positioning System (GPS)". GPS is costly, power consuming and restricted for
outdoor applications. To find the optimum sensors location, a set of nonlinear equations must
be solved to reach the best prospect purpose in some cases. The effectiveness in this approach
386
(ICoIT'17)
10thof April, 2017.
is how to investigate the sensors deployment by searching for the optimal or the near optimal
solution [2].
2. RELATED WORK
Wen-Hwa Liao et al., presented in 2011 sensors deployment system based on "Glowworm
Swarm Optimization (GSO)" to improve the nodes coverage after a process of random
deployment. Each node was considered as "individual glowworms emitting". A "luminant
substance" called "luciferin" and the strength of the "luciferin" is relies on the distance
between the sensor node and its neighboring sensors. A sensor node is appealed towards the
lower intensity of "luciferin". The sensing field coverage was maximized when the sensor
nodes must move towards the lower sensor density area [3].
Guo, et al., proposed in 2012 a technique with goal coverage based on lattice scan. They
divided the area into lattices. Then the best lattice was selected as a position to the next
sensor. This method used smallest number of nodes to attain the coverage goal and get
enhanced positions for the deployed sensor nodes [1].
Yourim Yoon, et al., proposed in 2013 an effective "genetic algorithm" using a "novel
normalization method". A "Monte Carlo method" was assumed in designing an effective
assessment function. They showed that their computation time can be reduced without losing
the quality of the solution. They selected a small number from random samples and steadily
increase the number for the succeeding generations [4].
Alduraibi et al., They proposed in 2016 three optimization models for determining the node
density that varies in the objective. The first opts to achieve a desired level of detection
fidelity while minimizing the number of deployed sensors. The second model considers the
scenario with a constrained node count and determines the position of the available nodes
such that the coverage is maximized. In the third model, they strive to minimize the number
of deployed nodes when the desired fidelity is not uniform and some locations require higher
coverage than others. The proposed optimization formulations are generic in nature and can
be applied to any sensor coverage model [5].
3. WSN APPLICATIONS
WSN can be used to sense the object movement, sound, and temperature. It can be used
to collect information about the system and transmit it to the Base station (BS). In military
applications, WSNs can be used to observe and control the front line. Figure (1) presents a
view of some WSN architecture [6].
387
(ICoIT'17)
10thof April, 2017.
Figure (1): WSN architecture sample [6].
4. SENSORS DEPLOYMENT
The deployment approaches in WSN was divided into two groups as "sparse" and "dense
deployments". In the given field of interest the "sparse deployment" can be used with little
number of nodes deployed while the "dense deployment" can be used in high number of
deployed sensors [7]. The "dense deployment" technique was used in the case of every event
wanted to be noticed or there are multiple sensors are deployed to cover the same area. The
sparse deployment is suitable to achieve maximum sensing area with minimum number of
sensors. The sensor nodes are either static or mobile so they can change their positions with
time [2].
Sensors are deployed in certain field either by locating them in "predetermined places" or
randomly deployed. "Random sensors deployments" approach is one type of the "dense
deployments". WSNs with mobile sensors are normally initiated with a "random
deployment" and by their mobility change their locations to the best suitable location [1] [2].
4.1 STATIC DEPLOYMENT

In "static deployment", sensors are static and their locations must be selected due to certain
"optimization strategy". Sensors locations will stay fixed and not changed during the
operation of the network. "Static deployment" strategy can be achieved by deterministic
steps and random deployments. This approach (the "deterministic deployment") starts by
surveying the area of interests and continuing with the deployment process [2].
4.2 DYNAMIC DEPLOYMENT

One important application of a "dynamic deployment" was used in robots. Mobility helps
in attending the sensing goal by letting the sensor move towards a maximum sensing
performance. In "random deployment" most of the nodes are tossed in the first step, then a
special utilization and reformation estimations to select the next movement [8]. Many
developed algorithms were built such as "virtual force oriented particles
388
(ICoIT'17)
10thof April, 2017.
algorithm"[9],"simulated annealing algorithm"[10], "particle swarm optimization algorithm"

[11] and "simulated annealing genetic algorithm" [3].
5. SENSING AREA
The essential aim of any WSN is to have maximum sensing area (maximum coverage area).
Coverage or sensing in WSN represents the backbone issue due to its relation with the energy
consumption, network reconfiguration and connectivity. A basic problem is how to achieve
maximum effective sensing area by deploying the sensors in an optimal manner. A good
sensing area is that area in which every point in the area will covered and monitored by at
least one sensor [12].
6. GSO APPROACH
There are many approaches were suggested and implemented to achieve certain coverage
connectivity and reliability. Wen-Hwa, et al., for example suggested in 2011 a sensor
deployment
System based on "glowworm swarm optimization" to improve the coverage after an "initial
random deployment" of the sensors. They considered each sensor node as "individual
glowworms emitting a luminant" body called "luciferin" and the strength of the "luciferin"
is depending on the space isolating the sensor node with its neighboring. A sensor node can
be appealed to its adjacent with lower strength of "luciferin". When the sensor tends to move
towards the area having lesser sensor concentration then the sensing coverage area will be
maximized. Their simulation results showed that "GSO-based sensor deployment approach"
can deliver great coverage with restricted number of the sensor movements [3]. They
concluded the following GSO results:
When 50 sensors were deployed in a center of 100 X 100m area with sensing range of 5m &
communication range of 10m, they found a coverage percentage of about 35.4 %. When 100
sensors were deployed at the same area and same sensing and communication ranges, they
found coverage of 64.6 %. While when they used 200 sensors with the same area, same
sensing and communication ranges satisfied coverage of 92.2 %. This method was failed to
achieve connectivity and /or reliability with less numbers of sensors deployed randomly in
such area.
7. MARKOV MODELS
A "Markov chain" represents one aspect of the "mathematical models". It has certain
random occurrence depending on time in a specific style that the future state depends only
on the current state. The applied time can be described as discrete or continuous. In some
mathematical applications, such phenomenon which depends on time and only the present
activity will affects the next activity was called a "dynamical system". For any time t, the
upcoming path (A(x), x t) can be entirely stated by the current state A(t), without any need
to the past states. In most applications they dealt with random variables rather than the
deterministic values.
Markov chains represent a very important modelling tool that can be applied for the situations
in which the trials do not happen according to an "independent and identically distributed"
389
(ICoIT'17)
10thof April, 2017.
mechanism. Markovian property means that any stochastic process {xt} must have the
following property:
P{xt1= j\ x0= k0, x1 =k1, . . . , xt-1=kt-1, xt=i} = p{xt+1=j\xt=i}, for t = 0, 1, . . . and every
sequence i, j, k0, k1, . . . , kt-1. Markovian property means that "the conditional probability of
any upcoming event /(any preceding events and the current state Xt= i), is independent of the
historical events and depends only on the current state" [14].
Markov chain can be generally defined as follows: for a "set of states, S = {s1, s2,...,sn}". The
movement will be initiated in any one of these states and goes sequentially from state to state.
Each change is titled as a step. If the chain is found in certain state si, then it can go to state
sj at its next step with certain probability equal to pij. This probability value will depends on
states i and j only.
The transition probability represented by (pij) means the probability of going from state i to
state j on the upcoming step. The transition matrix collects all the transition probabilities
from state to state [2].
8. TRANSTION MATRIX
The transition probabilities are commonly expressed as an M X M matrix called the transition
probability matrix (or transition matrix), P. The transition probability matrix P and its
characteristics are given as follows [13]:

=1 = 1 for I = 1 , 2 ,.,m (1)
Based on the Chapman-Kolmogorov equation, the probability of the system moving from
State i to State j after n periods (n transitions), that is, the n-step transition probability matrix,
P (n) , can be obtained by multiplying the matrix P by itself n time. Thus:
() = (2)
0
Let the initial state vector, Q , be the probability that the Markov chain is in State i at time
0. Then, the state vector, Qn , which is the probability that the chain is in State j after n
transitions, can be expressed as:
= 0 () (3)
0
Where Q =[q1 ,q2 ,,qm]; and qi = probability of being in State i at Time 0.
9. SUGGESTION DEPLOYMENT APPROACH

Markov approach is utilized in this study as a new approach to find the best possible
sensors deployment. The following network conditions are suggested:
390
(ICoIT'17)
10thof April, 2017.
The area of interest is of (100 x 100m), numbers of the available sensors are (50, 100 and
200), and the sensors are initially deployed in a random manner. Net Logo is suggested to be
used as a simulation tool to create, implement and evaluate each model in this study.
Case 1:
50 sensors are deployed randomly in a 100x 100m area as shown in figure 2-a. A suitable
developed markov algorithm is used to change these sensors positions in sequential steps to
reach the best possible state shown in figure 2-a.The initial transition matrix of (10) sensors
is shown in figure 3-a. The transition probability is calculated with respect to certain
developed criterion, and after 9 iterations the transition matrix becomes as figure 3-b.
a: initial random deployment

b: final deployment
FIGURE 2 : initial and final deployment of 50 sensors
0 0.159 0.076 0.18 0.129 0.094 0.044 0.105 0.057 0.156

0.163 0 0.162 0.099 0.034 0.112 0.188 0.093 0.115 0.035
0.074 0.152 0 0.13 0.121 0.141 0.051 0.143 0.055 0.134
0.149 0.079 0.111 0 0.071 0.148 0.15 0.137 0.102 0.052
0.159 0.04 0.154 0.106 0 0.118 0.186 0.099 0.098 0.038
0.094 0.109 0.146 0.18 0.096 0 0.136 0.019 0.091 0.127
0.038 0.161 0.047 0.16 0.133 0.119 0 0.127 0.063 0.152
0.11 0.095 0.155 0.173 0.084 0.02 0.151 0 0.096 0.116
0.072 0.142 0.072 0.158 0.101 0.116 0.092 0.116 0 0.13
0.167 0.036 0.149 0.067 0.033 0.135 0.185 0.118 0.109 0
a: Initial Transition Matrix
391
(ICoIT'17)
10thof April, 2017.
0 0.137 0.093 0.288 0.047 0.097 0.13 0.066 0.047 0.094

0.099 0 0.154 0.156 0.065 0.101 0.187 0.074 0.129 0.034
0.07 0.158 0 0.223 0.097 0.139 0.035 0.119 0.035 0.123
0.113 0.084 0.117 0 0.1 0.134 0.134 0.116 0.117 0.086
0.049 0.092 0.134 0.263 0 0.096 0.177 0.049 0.092 0.049
0.07 0.101 0.135 0.247 0.068 0 0.155 0.034 0.1 0.089
0.079 0.156 0.029 0.207 0.104 0.129 0 0.118 0.05 0.127
0.06 0.092 0.145 0.268 0.043 0.043 0.177 0 0.103 0.07
0.042 0.157 0.042 0.265 0.08 0.123 0.074 0.101 0 0.116
0.084 0.043 0.148 0.196 0.043 0.111 0.189 0.069 0.118 0
b: Final transition Matrix
Figure 3: sample of the transition matrix for the 10 sensors
Case 2:
100 sensors are deployed randomly in a 100x 100m area as shown in figure (4).
a: initial random deployment b: final deployment
Figure 4: initial and final deployments of 100 sensors
392
(ICoIT'17)
10thof April, 2017.
Case3:
200 sensors are deployed randomly in a 100x 100m area as shown in figure (5).
a: initial random deployment b: final deployment
Figure 4: initial and final deployments of 200 sensors
Table (1) summarizes the results of the above three cases. Figure (6) shows a clear
comparison between this paper results and the GSO (developed by Wen-Hwa, et al.,) results.
From the results it's clear that the developed approach gave a good results comparing with
the GSO.
Table (1) the results of the suggested cases.
Cases No. of Field area Sensing Communication Total coverage
Sensors Range Range area
Case 1 50 100 X 100 5 10 37.8786
Case 2 100 100 X 100 5 10 74.326
Case 3 200 100 X 100 5 10 95.5595
393
(ICoIT'17)
10thof April, 2017.
Number of deployed sensors
Coverage Rate
200 100 50
Figure 6: comparison between this paper results and the GSO approach
10. CONCLUSION
In this study, we have suggested and developed a markovian algorithm to develop the
process of sensors deployment. The proposed approach is to model the sensing field, nodes
positioning and the sensors communication range. The results indicate a best approach in
achieving good coverage of the sensing field and connectivity of the network as compared
to other methods such as GSO. This method ensures full connectivity and good system
reliability. Different numbers of sensor nodes are deployed randomly and then re-change
their positions in a sequential manner till reaching the final optimal location for each node.
The end position for each node is ensuring good and reliable connectivity with all neighbors
in addition to reduce the overlapping. A connected graph representation is also used to show
the network connectivity.
REFERENCES
[1] X. M. Guo, C. J. Zhao, X. T. Yang et al, A Deterministic Sensor Node Deployment
Method with Target Coverage Based on Grid Scan, Chinese Journal of Sensors and
Actuators, Vol.25, No.1, 2012, pp.104-109.
[2] Hasson, Saad Talib, and Abdul Nasir Reyadh. "A Modified Virtual Approach to Deploy
Border Line Sensors." Asian Journal of Information Technology 15.16 (2016): 2750-
2755.
[3] Liao, Wen-Hwa, Yucheng Kao, and Ying-Shan Li. "A sensor deployment approach using
glowworm swarm optimization algorithm in wireless sensor networks." Expert
Systems with Applications 38.10 (2011): 12180-12188.
[4] Y. Yoon and Y. H. Kim, An efficient genetic algorithm for maximum coverage
deployment in wireless sensor networks, IEEE Transactions on Cybernetics, (2013).
[5] Alduraibi Fahad, Noureddine Lasla, and Mohamed Younis. "Coverage-based node
placement optimization in wireless sensor network with linear
topology." Communications (ICC), 2016 IEEE International Conference on. IEEE,
2016.
394
(ICoIT'17)
10thof April, 2017.
[6] Abbasi, Abu Zafar, Noman Islam, and Zubair Ahmed Shaikh. "A review of wireless
sensors and networks' applications in agriculture." Computer Standards &
Interfaces 36.2 (2014): 263-270.
[7] Noureddine Boudriga; "AWSN-Based system for country boarder surveillance and target
tracking", advances in remote sensing, 5, 2016. (http://www.scrip.org/gurnal/ars).
[8] Luo, Qiang, and Zhongming PAN. "An algorithm of deployment in small-scale
underwater wireless sensor networks." Chinese Journal of Sensors and Actuators 24.7
(2011): 1043-1047.
[9] Wang, Xue, Sheng Wang, and Jun-jie Ma. "Dynamic sensor deployment strategy based
on virtual force-directed particle swarm optimization in wireless sensor
networks." Acta Electronica Sinica 35.11 (2007): 2038.
[10] Lin, Frank YS, and Pei-Ling Chiu. "A near-optimal sensor placement algorithm to
achieve complete coverage-discrimination in sensor networks." IEEE
Communications Letters 9.1 (2005): 43-45.
[11] Yi,L. "Wireless sensor network deployment based on genetic algorithm and simulated
annealing algorithm." Computer simulation (2011): 171-174
[12] Boukerche, Azzedine, Xin Fei, and Regina B. Araujo. "WSN04-4: A Coverage-
Preserving and Hole Tolerant Based Scheme for the Irregular Sensing Range in
Wireless Sensor Networks." IEEE Globecom 2006. IEEE, 2006.
[13] David A. Levin, Yuval Peres, and Elizabeth L. Wilmer. Markov chains and mixing
times. American Mathematical Society, Providence, RI, 2009. With a chapter by James
G. Propp and David B. Wilson.
[14] Ding, Xin, Ziyi Qiu, and Xiaohui Chen. "Sparse transition matrix estimation for high-
dimensional and locally stationary vector autoregressive models." arXiv preprint
arXiv:1604.04002 (2016).
395
(ICoIT'17)
10thof April, 2017.
NEURAL FUZZY PETRI NET BASED ARABIC

PHONEME CLASSIFIER WITH MFCC FEATURE
EXTRACTION
Abduladhem Abdulkareem Ali
Department of Computer Engineering , College of Engineering , University of Basrah, IRAQ.
abduladhem@ieee.org
Ghassaq S. Mosa
Department of Computer Engineering , College of Engineering , University of Basrah, IRAQ

In this paper Arabic phoneme classification is employed
Article History: using Mel Frequency Cepstral Coefficient (MFCC) as the
Received: 8 March 2017 basic recognition features. These features are first calculated,
then used as an input to fuzzy neural Petri net. One network
are used for each phoneme. The network was first trained
Published: 10 April 2017 based on a set of training recoded data, then the network are
validated based on another set of data. Classification accuracy
Keywords: neural fuzzy were then calculated and it have been found that the resulting
Petri net, Phoneme total accuracy reached 74.94%.
recognition, speech
recognition, Mel
Frequency Cepstral
Coefficient, pattern
recognition.
1. INTRODUCTION
A
Rabic language are the 5th spoken language. Its importance also comes from that
many non-Arabic origin people speaks this language. It is the language of Quran.
All Muslims who reads Quran have to read it in this Language. Arabic phoneme
recognition attracts many researchers in the world. Many techniques have been
used for this purpose. Ahmad etal [1] employed neural network with Mile Frequency
Cepstral Coefficient (MFCC), and Linear Predictive Coding (LPC) for the recognition of
isolated Arabic words. Ali etal [2] used neural network with Fast Fourier Transform(FFT),
Wavelet Transform (WT) and a combination of both for the recognition of Arabic phonemes.
MFCC have also been adopted in [3] using a decision tree based on hierarchical neural
networks. Fuzzy Systems have also been investigated for such applications. Debyeche etal
[4] used fuzzy vector quantization for the Arabic speech recognition. Taleb and Benyettou
[5] employed fuzzy neural systems and frequency parameters to identify Vowels. In [6] and
[7] Hidden Markov Model (HMM) are employed for the Arabic phoneme and speech
recognition. Fuzzy Neural Petri net have also been used for this purpose but with LPC
396
(ICoIT'17)
10thof April, 2017.
features only[8]. This paper investigates farther the use of Fuzzy Neural Network as a
decision making logic but with MFCC as a feature extraction technique. In MFCC the
features mimics the Human ear by adding a mile scale filter.
A/D
Segmentation Preprocessing Feature Decision making
Converter extracting
Figure (1) General Speech Recognition System Module
Figure (1) shows the block diagram of a phoneme classification system. At first the
speech signal is recorded via a microphone. The recorded signal is then passed into a band
pass filter to limit its frequency and avoids aliasing during sampling. The signal is then
sampled and digitized using analogue to digital converter. After digitization the signal is
segmented to extract and isolate the phonemes that have to be classified. Segmentation into
phonemes avoids the recognition of huge amount of vocabularies exist in the language if
isolated word recognition is to be used. The isolated phonemes are then feed into filtering to
reduce noise and then the signal is scaled and unified in order to remove the effect of
recording gain and speaker voice level variations. The features are then extracted from the
phonemes signal and then feed into the decision making in order to classify the phoneme.
2.MEL FREQUENCY CEPSTRAL COEFFICIENT
The MFCC is the mostly used feature extraction technique in speech recognition system.
This feature extraction technique was developed in 1940 by Stevens and Volkman and
employed by Mermelstein and Davis in 1980 to extract the features of the speech signal [9].
This technique is based on the modeling of human auditory system. The study of human ear
shows that its sensitivity to low frequency components of the speech signal are linear. If the
frequency increases the ear sensitivity change to vary in logarithmic scale to the variation of
audio signal frequency to capture only important features of the speech signal. This
phenomena is implemented in MFCC scale through linear scale below 1kHz and logarithmic
scale above 1kHz. Figure (2) shows the relation between the mile scale and frequency
physical scale. This relation can be expressed mathematically as:
Mel (f) = 2595 log10 (1 +f/700) (1)
In which (f ) represents the signal frequency
and the "Mel" represent the unit that measures the frequency or the perceived pitch or of a
tone. Figure(3) shows the steps required for the calculation of MFCC. This calculation
requires seven steps they are:
397
(ICoIT'17)
10thof April, 2017.
Fig. (2) Relation between Mel scale & physical scale
Fig.(3) The block diagram of MFCC calculation
A. Frame Blocking: Frame the speech signal into blocks of approximately (10 to 30 msec)
in length (at sampling rate 16 kHz) over which the characteristics of the speech Signal are
fairly stationary. Results of the framed signals (blocking into blocks of N sample) [10].
B. Windowing: The samples of speech within each frame are multiplied by a finite-
duration window. The shape of duration of the window determines the time and frequency
resolution that can be represented in the data. One of the most popular window function in
ASR systems is the hamming window defined in Equation (2). The effect of widowing is
shown in figure (4). [10].
398
(ICoIT'17)
10thof April, 2017.
2n
w(n) = 0.54 0.45cos(N1)
0nN1 (2)
. Fig.(4) Effect of window on signal
C. Fast Fourier Transform: After windowing, the power spectrum of each data frame is
calculated. The most commonly used spectral estimation technique is the Fast Fourier
Transform FFT. The vocal tract shape is revealed by the spectral analysis.
For the finite duration sequence {y(n)}, the Discrete Fourier Transform is defined as:
Y(k) = N1
n=0 y(n)e
jkn/N
= N1
n=0 y(n)W
nk
In which (0 n, k N 1) (3)
D. Mel Filter Bank Processing: This Mel filter function is to smooth the frequency
spectrum, such that it closely model the human ear sensitivity. The Mel frequency scale
calculation is performed using a set of band-pass filters varies between 13 and 24, depending
on the bandwidth of the signals that are being processed. The filter output Z(m) can be given
as[9]:
k
1
2
Z(m) = |Y(k)|Sm (k)

k=0
m = 1M (4)
In which M: is the number of the filter banks.
Sm(k): the triangular filter banks as shown in figure (5).
Fig. (5) Mel scale filter bank
399
(ICoIT'17)
10thof April, 2017.
E. Log Magnitude: After Mel scale filtering the logarithm of the output of each filter bank
is calculated. The logarithm can be regarded as a crude approximation of the non-linear
intensity-loudness relationship associated with the human auditory system. An important
advantage of this processing step is that it makes the shape of speech spectra insensitive to
different loudness levels amplification factors in the linear energy domain become constant
offsets in the log-energy domain and if these are properly compensated for, the resulting
spectra are independent of differences in absolute loudness [11].
F. Discrete Cosine Transform or DCT: The discrete cosine transform (DCT) is applied to
Mel-scaled log energy vectors in order to obtain features that are statically independent.
The resulting Mel-scaled cepstral coefficients (MFCCs) are ordered such that higher order
coefficients explain less variance [10,11].Because the higher order coefficient contain less
information after the application of the DCT, the cepstral vector can be truncated to
Contain fewer components than the number of filters in the Mel-scaled filter bank.
M
(2m 1)j
Cx (j) = (log(Z(m)) cos [ ]
2M
m=1
j = 1,2, . . J 1 (5)
Where Cx (j): is the jth Mel Frequency Cepstrum Coefficients.
J: is the desired length of the cepstrum.
M: is the number of filter banks.
G. Energy Features: In addition to the features describing the spectral envelop, a caustic
feature vectors usually also contain features that are related to the energy in the signal. The
cepstral coefficient, c is often used as a measure of the mean log-energy in each frame. It is
sometimes replaced by (or used in addition to) the log-energy that is directly computed from
the signal at frame level (Log E) [10].
E n =
[x(m). w(n m)]
2
(6)
where, x(n) represent the signal, w(n) represent a window function (such as Hamming
window) and En is the short time energy at discrete time n.
Another time dependant sequence, which can serve as representation of speech signal is the
short-time average zero-crossing rate. A simple technique to measure the frequency contents
of a signal can be found by the rate at which zero crossing occurs. The short-time average
zero-crossing can be given by the following Equation [10].
400
(ICoIT'17)
10thof April, 2017.
1
Zn = 2N
|sgn[x(m)] sgn[x(m 1)]w(n m)| (7)
1 () 0
sgn[x(n)] = { (8)
0 () < 0
w (n): is a window function.
N: is the window length.
According to the MFCC algorithm, the performance of MFCC might be affected with the
following factors: The first factor is number of the filters used in the filter bank. The second
is the shape of the filters used (triangular, rectangular or Schroeder). The third factor is the
spacing of these filters, whether they are overlapped or not. And the fourth factor is the way
that power spectrum is warped.
3.FUZZY NEURAL PETRI NET
The fuzzy neural Petri net uses the Petri net stricture, embeds the fuzzyfication in the input
place and employs the back propagation algorithm as a learning role. Figure (6) shows the
basic structure of the fuzzy Petri net. In general Petri net consists of a number of places
which are input, hidden or output places. These places are connected by arc and separated by
transitions. In the fuzzy neural Petri net the input places are the input layer, the transitions
are the hidden layer and the output places are the output layer.
Fig. (6) The structure of the Neural Fuzzy Petri Net.
401
(ICoIT'17)
10thof April, 2017.
The features are the MFCCs are fuzzyfied and marked in the input places. The transition are
fired based on the value calculated based on the weights connecting input places and
transitions and also based on threshold limit of each transition. The output places are marked
based on the weights connecting transitions with the output place. Figure (7) shows a typical
of the net showing the notation used. The equations relating different parts of the network
are as follows:
x f (input( j )) (9)
j
Where f is a triangular mapping function defined by

Fig.(7) One section of the Fuzzy Neural Petri Net
x min(x)
average( x) min(x) , if x average( x)
(10)
max(x) x
f ( x) , if x average( x)
max( x) average( x)
1 , if x average( x)

n
Zi T [Wij S (rij X j )] (11)
j 1
Wij is the weight between the i-th transition and the j-th input place;
- rij is a threshold level associated with the level of marking of the j-th input place and the
i-th transition;
- Zi is the activation level of i-th transition
T is a t-norm, S denotes an s-norm, while stands for an
Implication operation expressed in the form
a b sup{c [0,1], aTc b} (12)
Where a, b are the arguments of the implication operator confined to the unit interval.
If t-norm is defined as a multiplication operator () then
402
(ICoIT'17)
10thof April, 2017.
xj
, if rij x j
rij x j rij (13)
1, otherwise

xj
n
, if rij x j
Zi Wij rij (14)
j 1 1, otherwise

- Yk is the marking level of the k-th output place produced by the transition layer and
performs a nonlinear mapping of the weighted sum of the activation levels of these
transitions (Zi) and the associated connections Vki
No.ofTransitions
Yk f ( V Z ),
i1
ki i k 1,2,...,m (15)
Where f is a nonlinear monotonically increasing function from R to [0,1].

The parameter update is derived in a similar way to back propagation algorithm. The sum
squared error is used as an optimization criteria. The update is based on minimizing this
measure.
1 m
E (tk yk ) 2 (16)
2 k 1
tk is the k-th target;
yk is the k-th output.
The updates of the parameters are performed according to the gradient method
param(iter 1) param(iter) paramE

(17)
Where Eparam E is a gradient of the performance index E with respect to the network
parameters, is the learning rate coefficient, and iter is the iteration counter.
403
(ICoIT'17)
10thof April, 2017.
The nonlinear function associated with the output place is a standard sigmoid described as
1
yk
1 exp( ZiVki ) (18)
Hence, the parameter update is derived as follows:
E
1- Vki (19)
Vki
E E yk

Vki yk Vki (20)
E
(tk yk )
yk (21)
yk
yk (1 yk )Zi
Vki
exp(ZiVki )Zi

1 exp(ZiVki ) (22)
exp(ZiVki )Zi (tk yk )

Vki (23)
1 exp(ZiVki )
Vki (iter 1) Vki (iter) Vki (24)
for k=1,2,,m, i=1,2,, no. of transitions
E
2- rij (25)
rij
404
(ICoIT'17)
10thof April, 2017.
E E yk Zi

rij yk Zi rij (26)
E
(tk yk )
yk (27)
yk
yk (1 yk )Vki
Zi (28)
n
Zi T [Wij S (rij x j )] (29)
j 1
Zi
A (Wij (rij x j ) Wij (rij x j ))
rij rij

A(1 Wij ) (rij x j )
rij (30)
Where
n
A T [WiS (ri x )] (31)
1
j
xj x j
, if rij x j 2 , if rij x j
(rij x j ) r rij
rij rij ij 0, otherwise
1, otherwise
(32)
405
(ICoIT'17)
10thof April, 2017.
x j
exp(ZiVki )Vki n , if rij x j
rij (tk yk ) T [WiS (ri x )](1 Wij ) rij2
1 exp(ZiVki ) 1 0, otherwise
j
xj x j
n
, if ri x , if rij x j
Vki * Wi ri *(1 Wij ) rij2
1 1, otherwise
0, otherwise
j
(33)
rij (iter 1) rij (iter) rij (34)
E
3- Wij (35)
Wij
E E yk Zi

Wij yk Zi Wij , k=1,2,,m, i=1,2,, hidden, j=1,2,, n.
(36)
E
(tk yk )
yk (37)
yk
yk (1 yk )Vki
Z i
exp(Z iVki )Vki

1 exp(Z iVki ) (38)
406
(ICoIT'17)
10thof April, 2017.
Zi
A (Wij (rij xi ) Wij (rij x j ))
Wij Wij
A(1 (rij x j )) (39)
n
A T [WiS (ri x )]
1
j
xj
, if rij x j
(rij x j ) rij
1, otherwise
(40)
exp(Z iVki )Vki n

Wij (t k yk ) [WiS (ri x )](1 (rij x j ))
1 exp(Z iVki ) T
1
j
xj x j
n
, if ri x , if rij x j
Vki * Wi * ri *(1 rij2 )
1 1, otherwise
0, otherwise

j
(41)
Wij (iter 1) Wij (iter) Wij (42)
Hence, the update of parameter learning is performed using equations (23 and 24) for Vki
equations (33 and 34) for the update of rij , and equations (41and 42) for the update of Wij
4. EXPERMENTAL RESULTS
In the recognition system proposed her, 33 Neural Fuzzy Petri Nets NFPN are used.
One network for each phoneme. 23 input places are used for each network, corresponding to
407
(ICoIT'17)
10thof April, 2017.
the input feature for MFCC. The feature vector is first fuzzified, then used as input to each
of the NFPN the number of transitions was selected between 22 to 33 based on the phoneme.
Each NFPN has one output place. Twelve cases are token for each phoneme and applied to the
network. The network is trained such that its output place is one corresponding to the phoneme
that it represents as zero otherwise. (1*10-5) error goal was selected as stopping criteria for
the training phase of the network. The network is then tested using the twelve set for each
phoneme (testing phonemes) as input to the network and total recognition accuracy is
calculated to be as:
Number of correctlyclassified phonemes

phoneme classification accuracy *100%
Total number of test phonemes
Table (1) shows the resulting total classification accuracy obtained in the testing phase. As can
be seen from the results, some of the phonemes classification reaches 100%. However, others
have poor classification. The total classification accuracy for all phonemes are 74.93%.
5.CONCLUSIONS
This work proposes Arabic phoneme classification technique uses MFCC and FNPN. The
proposed system was tested and it have been found that the total classification accuracy is
74.93% comparing the results with that obtained in [8] that uses linear predictive coding
which have accuracy of 71% and most of the phonemes, have improved accuracy.
408
(ICoIT'17)
10thof April, 2017.
Table (1) Classification accuracy
409
(ICoIT'17)
10thof April, 2017.
References:
[1] A. M. Ahmad, S. Ismail and D. F. Samaon, "Recurrent neural network with back propagation
through time for speech recognition," IEEE International Symposium on Communications and
Information Technology, 2004. ISCIT 2004., , pp. 98-102 vol.1. 2004.
[2] A. A. Ali, M. A. Alwan, and A. A. Jasim. "Hybrid Wavelet-Network Neural/FFT Neural
Phoneme Recognition." Proceedings of The 2nd International Conference on Information
Technology. pp. 39-47, Amman Jordan. 2005.
[3] Abduladhem A. Ali and I. T. Hwaidy "Hierarchical Arabic Phoneme Recognition Using
MFCC Analysis". Iraq J. Electrical and Electronic Engineering, vol.3, pp.97-106, 2007.
[4] M. Debyeche, A. Amrouche and J. P. Haton, "Distributed TDNN-Fuzzy Vector Quantization
for HMM speech recognition," 2009 International Conference on Multimedia Computing and
Systems, Ouarzazate, pp. 72-76, 2009.
[5] A. Taleb and A. Benyettou, "Arabic Vowels Fuzzy Neural Network Recognition". Journal of
Applied Sciences, vol.10, pp.848-851, 2010.
[6] A. A. Abushariah et al. "Arabic speaker-independent continuous automatic speech recognition
based on a phonetically rich and balanced speech corpus." International Arab Journal of Information
Technology (IAJIT) vol.9, no.1, pp.84-93, 2012.
[7] Gadeed, Ashwag, and Talaat Wahbi. "The Recognition of Holy Quran Reading types
Rewaih." International Journal of Advanced Research in Computer Science vol.5. no.3,pp.37-40,
2017.
[8] Ghassaq S. Mosa, and Abduladhem Abdulkareem Ali. " Arabic Phoneme Recognition Using
Neural Fuzzy Petri Net and Lpc Feature Extracting." The International Arab Conference on
Information Technology (ICIT-2009).
[9] R. M. Hegde, "Fourier Transform Phased-Based Features for Speech Recognition," Ph.D. Thesis,
Department of Computer Science Engineering, Institute of Technology, India, 2005.
[10] L. Rabinar and R.W. Schafar "Fundamental of Speech Recognition ",Prentice Hall, 1993.
[11] H. Xie ,"Speech Analyzer in an ICAI System for TESOL ",M.Sc. Thesis, Computer science,
Victoria University of Wellington, 2004.
410
(ICoIT'17)
10thof April, 2017.
New Weights in Laplacian Smoothing on Triangular Mesh

Kawa M. A. Manmi
Mathematics Department, College of Science, Salahaddin University Erbil, Iraq
Kawa.aziz@su.edu.krd

Mesh smoothing is one of the basic procedures for
Article History: improvement of mesh quality. Most smoothing techniques
Received: 6 March 2017 move vertices of the mesh without changing topology of the
connectivity. Laplace smoothing is one of the simplest and
efficient algorithm, where in each step vertex of the mesh is
Published: 10 April 2017 move to the barycenter of its neighbors. The only problem
with Laplacian smoothing is surface shrinking when it is
Keywords: triangular performed iteratively. In this paper, three relatively simple
mesh, Laplacian weights proposed in Laplacian smoothing which has less
smoothing, Weight surface shrinking from the previous weights. The
factors, normal vector, performance and effectiveness of the presented weights are
discrete surface. demonstrated on two knowing simply and doubly connected
geometrical shapes respectively.
1. INTRODUCTION
T
he polygonal discretization, particularly triangular discretization are widely
used for representing discrete surface in computer applications for example
smooth shading [1, 2], surface reconstruction from cloud of points and in
numerical simulation methods such boundary and finite element methods. More
applications can be found in direct rendering of points in computer graphics [3]. This
discretization leads undesirable noise and small-scale oscillation. Therefore, the quality
improvement for the discrete surface is needed. The mesh improvement has been widely
used in many areas, such as mesh generation [46], simplifying discretization [7],
dynamic grid or mesh deformation [89], and other mesh processing procedures. Mesh
improvement approach can be basically divided into two main categories, topological
optimization and smoothing (also called geometrical optimization). Topological
optimization changes the topology of a mesh, that is, the node-element connectivity
relationship, while smoothing or geometrical optimization improves mesh quality by
simply moving or adjusting node positions without changing the topology of mesh. This
paper will focus on the latter, smoothing [10].
The most two popular approaches for smoothing and reducing the noises in triangular
surface are minimizing the total energy stored in the line segments that connects the
points in the mesh and Laplacian smoothing [11].
411
(ICoIT'17)
10thof April, 2017.
2. PROBLEM STATEMENT
Laplacian smoothing is a common, simple and efficient technique to improve of the
polygonal mesh in general and particularly in the triangular mesh. Most smoothing
techniques including Laplacian smoothing move the vertices of the polygonal mesh without
changing the connectivity of the faces and vertices. The smoothed mesh has exactly the same
number of vertices and faces as the original (see figure 1). Laplacian smoothing is an iterative
process, where in each step every vertex of the mesh is moved to the barycenter of its
neighbors. A great deal of mesh smoothing algorithms have been proposed in the literature.
The most common techniques are based on Laplace smoothing [12, 13]. Taubin [12]
introduced signal processing on surfaces that is based on the definition of the Laplacian
operator on meshes and developed a fast and simple iterative Laplacian smoothing scheme.
When Laplacian smoothing is applied to a noisy three dimensional triangular mesh without
constrain, noise is removed but the problem is that this procedure produce Shrinking.
2.1. LAPLACIAN SMOOTHING

Suppose a surface S is discretized into triangular elements and composed of a set of n
vertices r1, r2, , rn and a set of m faces f1, f2, , fn. The umbrella operator U for p is defined
by;
w r i i
U ( p) i
p, (1)
w i
i
where summation is taken over all neighbors of p, wi are positive weights. The Laplacian
smoothing for any point p is;
pnew pold U ( pold ), (2)
where 0 < < 1 is the weights. The geometric idea behind Laplacian smoothing algorithm
(2) is to reduce the high noise surface and tends to flatten every patch of the surface as shown
in figure 1.
FIGURE 1. Sketch of the surrounding triangular faces to the arbitrary node p.

The weights wi can be chosen in different ways. The simplest chose is;
wi 1. (3)
According to [14] the weights to the inverse distance between p and its neighbors produce
good results.
1
wi p ri .
(4)
412
(ICoIT'17)
10thof April, 2017.
Note that the Laplacian smoothing in (2) is applied repeatedly.

3. NEW WEIGHTS SCHEMES wi
In this section, we suggest three new weight factors wi, to reduce the shrinking effects
which occurs during the Laplacian smoothing. Suppose a typical node p on the mesh, and ri
(i=1, 2, , Mi) are neighbor nodes p. First, we introduce the angle between unit normal of p
and ri as wi,
wi arccos(np . nri ), (5)

where the unit norm arbitrary node n is calculated by averaging the normal of the surrounded
planar triangular elements,
k
A n i
e
i
n i 1
k
, (6)
An
i 1
e
i i
e
Where n is the unit normal of the face fi and Ai is the area of the triangle fi (see figure 1)
i
The second weight method wi is the sum of all length of the edges that passes the node Qi,
Mi
wi Qi rj , (7)
j
where rj, Mi are the surrounded nodes and the number of surrounded of Qi, respectively.
Third weigh factor is,
wi=vi (8)
where vi is the volume of the pyramid that generated by projecting the triangle fi on the plane
Z=0. Therefore, if we denote the vertices of the triangle fi in the local coordinates by ri, ri1, O
, then the base vertices of the pyramid are ri, ri1, ri, ri1, where ri, ri 1 are projected points of
ri, ri1, on the plane Z=0, respectively (see figure 2b).
a b
FIGURE 2. (a) Sketch of the surrounding triangular elements to the node p at the local
coordinates system O-XYZ. (b) Pyramid with origin O is apex and base vertices are
ri, ri1 , ri, ri1.
4. NUMERICAL EXAMPLES
Two geometrical shapes are destroyed randomly in this section, then the algorithm for
smoothing sections 2 and 3 are implemented. In the first one, suppose a unit sphere is
discretized into optimal triangular surface. Then it is randomly destroyed in all directions
(figure 3 (a)). In figure 3(b)-(f) the weights in equations (3)-(5), (7)-(8) are used in the
413
(ICoIT'17)
10thof April, 2017.
Laplacian smoothing (2), with =0.5 and 5 time repeated respectively, here the number of
triangle is 2000.
(a) (b) (c)
(e) (f)
(d)
FIGURE 3. The unit sphere surface with 980 triangular elements (a) Deformed the unit
sphere randomly in all directions. Laplacian smoothing applied 5 times, =0.5 and with using
weight functions (b) in equation (3); (c) in equation (4); (d) in equation(5); (e) in equation(7)
and (f) in equation(8).
In Figure 4 the same process of figure 2 repeated with the increasing number of iterations
into 20. It is clear that the surfaces are smoother in the previous cases. In the table 1, relative
changes in the volume have been presented, when the Laplacian smoothing applied with all
weights in sections 2-3. The same parameters in the case figure 3 used but for different
number of triangles. It is clear that the relative change of volume reduce with increasing
triangle number (m) on the surface and the volume changes smaller in the proposed weights.
(a) (b) (c)
414
(ICoIT'17)
10thof April, 2017.
(d) (e)
FIGURE 4. Laplacian smoothing applied 20 times, =0.5 and with using weight functions
(a) in equation (3); (b) in equation (4); (c) in equation(5); (d) in equation(7) and (e) in
equation(8).
TABLE 1. Relative change of Volume V when Laplacian smoothing implemented 20

times with all weights in sections 2-3 for difference number of elements m.
V %
m
Eq. (3) Eq. (4) Eq. (5) Eq. (7) Eq. (8)
500 0.6 0.54 0.54 0.3 0.28
980 0.37 0.35 0.31 0.28 0.28
2000 0.19 0.17 0.15 0.12 0.12
3920 0.1 0.09 0.08 0.07 0.07
The second example is a torus, which given by,

x(, ) (r1 r2 cos) cos
y(, ) (r1 r2 cos) sin 0 , 2, (9)
z(, ) r2 sin
where r1=0.8 and r2=0.4 as showed in figure 5(a). The points in on the torus can be
considered as a parabolic, elliptical or hyperbolic point. Therefore, this type of the surface is
good for investigation. Half of the torus perturbed randomly, then all weights in section 2
and 3 in the Laplacian smoothing implemented for 20 cycles wit =0.5. The number of
elements is 8214.
(a) (b) (c)
415
(ICoIT'17)
10thof April, 2017.
(d) (e) (f)
FIGURE 5. Laplacian smoothing applied 20 times, =0.5 and with using weight functions
(a) in equation (3); (b) in equation (4); (c) in equation (5); (d) in equation(7) and (e) in
equation(8).
TABLE 2. Relative change of Volume V when Laplacian smoothing implemented 20 times with all
weights in sections 2-3 for difference number of elements m.
m V %
Eq. (3) Eq. (4) Eq. (5) Eq. (7) Eq. (8)
600 0.93 0.93 0.92 0.90 0.70
1014 0.80 0.76 0.72 0.68 0.50
1944 0.58 0.52 0.56 0.52 0.45
4056 0.32 0.28 0.27 0.25 0.21
8214 0.19 0.15 0.13 0.10 0.06
5. CONCLUSION
Laplacian smoothing is a simple and efficient technique to improve mesh quality. The
only problem in the Laplacian smoothing is when it is implemented iteratively, the surface
shrinking. The weight factor may have rule for the amount of shrinking. In this paper, three
new weights proposed and numerical examples are considered for the comparison purpose.
6. REFERENCES
[1] H. Gouraud, Continuous shading of curved surfaces. IEEE transactions on computers,
100(6), pp. 623-629, 1971
[2] BT. Phong, Illumination for computer generated pictures. Communications of the
ACM, 18(6), pp. 311-317, 1975.
[3] G. Schaufler and H.W. Jensen, Ray tracing point sampled geometry. Proceedings of
the Eurographics Workshop on Rendering Techniques 2000 pp. ( 319-328), 2013
[4] J. Liu, Automatic triangulation of n-D domains, in Proceedings of the
CAD/Graphics91, pp. 238241, Hangzhou, China, 1991.
416
(ICoIT'17)
10thof April, 2017.
[5] E. A. Dari and G. C. Buscaglia, Mesh optimization: how to obtain good unstructured
3D finite element meshes with notso-good mesh generators, Structural Optimization,
8, (2-3), pp. 181188, 1994.
[6] P. L. George, Improvements on Delaunay-based three dimensional automatic mesh
generator, Finite Elements in Analysis and Design, 25( 3-4) pp. 297317, 1997.
[7] J. Li and G. Lu, Mesh simplification and optimization with edge collapse and mass-
spring model, Journal of Computer-Aided Design and Computer Graphics, 18(3), pp.
426432, 2006.
[8] N. Qin, X. Liu, and H. Xia, An efficient moving grid algorithm for large deformation,
Modern Physics Letters B, 19 (28-29), pp. 14991502, 2005.
[9] X. Liu, N. Qin, and H. Xia, Fast dynamic grid deformation based on Delaunay graph
mapping, Journal of Computational Physics, 211(2), pp. 405423, 2006.
[10] S. Sun, M. Zhang, and Z. Gou, Smoothing Algorithm for Planar and Surface Mesh
Based on Element Geometric Deformation. Mathematical Problems in Engineering
Volume 2015, Article ID 435648, 9.
[11] Y. Ohtake, A. Belyaev and I. Bogaevski, Mesh regularization and adaptive smoothing.
Computer-Aided Design, 33(11), pp.789-800, 2001.
[12] T. DeRose, T. Duchamp, H. Hoppe, J. McDonald, and W. Stuetzle. Mesh optimization.
In Proceedings of SIGGRAPH, pp. 1926, 1993.
[13] M. Gross and A. Hubeli. Fairing of nonmanifolds for visualization. In Proceedings of
IEEE Visualization, pp. 407414, 2000.
[14] T. Gabriel. A signal processing approach to fair surface design. In Proceedings of the
22nd annual conference on Computer graphics and interactive techniques, pp. 351-
358. ACM, 1995.
417
(ICoIT'17)
10thof April, 2017.
Proposed Secret Authentication Method

for E-government Purposes
Wasan Ala'a alhamami
Zanbaq H.Thenon
Amer S. Elameer
Soukayna Hassan

E-Government involves the use of information and
Article History: communication technologies in public administration. One of
Received: 20 March 2017 the most important features of the E-government is the secret
Published: 10 April 2017 and efficient access to the information in order to manage it
Keywords: e- in good way. The information amount in E-government
government, systems is a lot and every amount is under specific user
DNA coding. ID
controlling for example manger of specific system has good
authentication level, whereas the manger assist has less
generation
authentication level and so on. In this paper a proposed
algorithm, aes
authentication method is presented. The proposed method
encryption
provides three authentication levels in E-government systems
based on specific ID. The proposed method based on DNA
computing with AES algorithm to generate the ID and in its
process.
1. INTRODUCTION
-government readiness is an important indicator of the quality of a countrys
E technological and telecommunication infrastructure and the ability of its

citizens, businesses and governments to adopt, use and benefit from modern
technologies. With the increasing importance of trends such as cloud computing, open
(big) data, participation tools or social media, new indicators and approaches need to be
introduced in the measuring of the e-government development, and the existing indices
should to be updated, redefined and restructured[1]. Information security is an important
matter in communication systems such as internet communication, multimedia systems,
medical imaging, and especially military communication [2] and also in E-government
systems since the E-government includes all the previous systems so the need for
information security necessary. Cryptography is one of the most important fields in
computer security. It is a method of transferring private information and data through
open network communication, so only the receiver who has the secret key can read the
encrypted messages which might be documents, phone conversations, images or other
form of data. In the encryption process, information security uses cryptograph to shift the
information into the cipher form which does not allow it to be used by unauthorized
personnel [3]. Authentication is a service used to provide the identity of an entity [4].
418
(ICoIT'17)
10thof April, 2017.
DNA computing, also known as molecular computing, is basically a collection of

specially selected DNA strands whose combinations will result in the solution to some
problem, depending on the problem at hand, technology is currently available both to
select the initial strands and to filter the final solution [5].
2. E-GOVERNMENT
E-government refers to the delivery of national or local government information and
services via the Internet or other digital means to citizens or businesses or other
governmental agencies. E-government is a one-stop Internet gateway to major
government services. E-government facilitates provision of relevant government
information in electronic form to the citizens in a timely manner; better service delivery
to citizens; empowerment of the people through access to information without the
bureaucracy; improved productivity and cost savings in doing business with suppliers
and customers of government; and participation in public policy decision-making. E-
Governance refers to how managers and supervisors utilize IT and Internet to execute
their functions of supervising, planning, organizing, coordinating, and staffing
effectively [6].
3. DNA CODING
DNA sequences are used to encrypt information in the encryption of the communication
methods, mainly the ones that need a robust data encryption scheme to challenge
unauthorized access. DNA sequences, which consist of the following nitrogenous bases: (A, T,C,
G), complementary to RNA sequences, which consist of the following: (U, A, G, C). The non-
coding segments, called introns in DNA sequences, are removed by splicing and the remaining
segments that encode information for protein synthesis, called exons, are assembled in mRNA
[7].
4. CRYPTOGRAPHY
Cryptography is defined as the subdivision of cryptology in which encryption

/decryption algorithms are designed, to guarantee the security and authentication of
data. Cryptography can be classified as Symmetric key algorithm and Asymmetric key
algorithm. Symmetric-key algorithms also known as single- key, one key and private-
key encryption are a class of algorithms for cryptography, that uses a Private(shared
secret) key and a Public (non-secret) algorithm to execute encryption /decryption process.
Some popular and well-respected symmetric algorithms includes DES, TDES , Blowfish
, IDEA , TEA ,AES (Rijndael) , Twofish ,RC6, Serpent and MARS. Asymmetric key
algorithms also known as public key encryption is a form of crypto system in which
encryption and decryption are modern encryption technology mathematically performed
using different keys, one of which is referred to as public key and the other is referred to
as private key [8].
419
(ICoIT'17)
10thof April, 2017.
Figure (1): Different Symmetric & Asymmetric Cryptographic Algorithms
4.1 Advanced Encryption Standard

The Advanced Encryption Standard (AES) was specified in 2001 by the National
Institute of Standards and Technology .The purpose is to provide a standard algorithm for
encryption, strong enough to keep U.S. government documents secure for at least the next
20 years. The earlier Data Encryption Standard (DES) had been rendered insecure by
advances in computing power, and was effectively replaced by triple-DES. Now AES will
largely replace triple-DES for government use, and will likely become widely adopted for a
variety of encryption needs, such as secure transactions via the Internet. A wide variety of
approaches to implementing AES have appeared, to satisfy the varying criteria of deferent
applications. Some approaches seek to maximize throughput others minimize power
consumption and yet others minimize circuitry. For the latter goal, Rijmen suggested using
subfield arithmetic in the crucial step of computing an inverse in the Galois Field of 256
elementsreducing an 8-bit calculation to several 4-bit ones. Satoh et al. [3] further
extended this idea, using the tower field approach of Paar, breaking up the 4-bit
calculations into 2-bit ones, which resulted in the smallest AES circuit to date. Rijndael round
consists of four different stages
Sub Byte transformation: (S-box substitution) provides non linearity and confusion,
constructed by multiplicative inverse and affine transformation.
Shift Row: (rotations) provides inter-column diffusion where the bytes in the last three rows
of the states are cyclically shifted.
Mix Column: (linear combination) provides inter-byte diffusion where each column vector
is multiplied by a fixed matrix. The bytes will be treated as polynomials rather than
numbers
Add Round Key: (round key bytes XOR with each byte of the state and the round key)
provides confusion .
420
(ICoIT'17)
10thof April, 2017.
The encryption process begins with an Addroundkey stage, and followed by nine rounds of
Sub Bytes, Shift Rows, MixColums and Addroundkey transformation. The transformation
will be performed respectively, and iteratively (Nr times) depending on the key length. The
final round will only include 3 stages; Sub Byte, Shift Rows and Addroundkey. All of the
operations are byte- oriented. The encryption and decryption structure consists of several
transformation stages as shown in Figure (2). The decryption is essentially the same structure
as encryption, but Sub Byte, Shift Row and MixColums are replaced by their inverses;
InvSubBytes, InvShiftRows, InvMixColumns, and Addroundkey. It is the reverse order of
the encryption structure [9].
Figure (2): AES Encryption and Decryption
5. THE PROPOSED METHOD
One of the main features and responsibility of the E-government is that it must provide
secure access to the information. In this paper a proposed secure authentication method is
presented. In the proposed method DNA coding with AES is used together to provide the
authentication. The proposed method has three authentication levels based on specific ID
length that is used (if the ID length is 64 bits then the authentication level is the first, if the
ID length is 128 bits then the authentication level is the second, if the ID length is 192 bits
then the authentication level is the third).The ID that is based is given to every user under
the control of the specific E-government system. The ID is generated based specific steps
(Algorithm 1), after that these IDs are stored, and then these IDs are given to the users of the
E-government system. If the user want to use the system under the control of the E-
421
(ICoIT'17)
10thof April, 2017.
government he\she use the ID , then the ID is processed according the steps of the algorithm
(Algorithm 2).
Start
Read characters , key.
Convert characters ASCII values
Convert binary representation
Convert to DNA codes
Convert to decimal values based

on special convert
Apply AES encryption with

specified key according to the
length of the entered characters
Assign ID to user
Store the user , ID, key, characters

in specific database
End
Figure (3) :- ID generation flowchart
422
(ICoIT'17)
10thof April, 2017.
ID generation algorithm
Input: - Characters, key.

Output: - Encrypted ID.
Begin
Step1:- Read characters.
Step2:- Convert characters to ASCII values.
Step3:- Convert ASCII values to binary representation.
Step4:- Convert binary representation to DNA codes.
Step5:- Convert DNA codes to decimal values based on specific dictionary.
Step6:- Apply AES encryption method use the entered key with decimal values to get
ID.
Step7:- Store original characters, key, ID with the specific user information in
identification database.
Step8:- Inform the user with his\her ID.
End
423
(ICoIT'17)
10thof April, 2017.
Start
Input ID
Read key that is corresponding

to the entered ID form the
specific database in order for the
AES decryption.
Apply AES decryption
Convert to DNA codes
Convert to binary representation
Convert to ASCII values.
Convert to characters.
Is there any
matching with stored
Yes characters? No
Make the required Ignore Access and

Action. require another ID.
End
Figure (4) :- Proposed Authentication Method
424
(ICoIT'17)
10thof April, 2017.
Proposed authentication method algorithm
Input:- ID
Output:- Perform specific authentication
Begin
Step1:-Read the ID.
Step2:- Read the corresponding key from the specified database.
Step3:- Apply AES decryption method.
Step4:- Convert AES decryption result to DNA codes.
Step5:- Convert DNA codes to binary representation.
Step6:- Convert binary representation to ASCII values.
Step7:- Convert ASCII values to characters.
Step8:- Compare the resulted characters after the process with the stored characters if
there is match allow access and perform required action otherwise ignore the access.
End
5.1 Simulation of the proposed method process
In this section example of the proposed method will be explained in details in two
phases the first phase is the ID generation phase and the second phase is the
Authentication phase.
First phase
Step1:- Read the characters
abcdefgh
Step2:- Convert characters to ASCII values
097 098 099 100 101 102 103 104
Step3:- Convert ASCII values to binary representation.
0110000101100010011000110110010001100101011001100110011101101000
The result will be as follow:
Step4:- Convert binary representation to DNA codes using table (1)
425
(ICoIT'17)
10thof April, 2017.
Table (1) :- Binary to DNA converting

Binary Representation DNA code
00 A
01 C
10 G
11 T
TGAC TGAG TGAA TGTT CGCC CGCG CGCT CGGA
Step5:- Convert DNA codes to decimal values using table (2)
Table (2) :- DNA to decimal converting

DNA code Decimal value
A 0
C 1
G 2
T 3
32013202320032331211121212131220
Step6:- Apply AES encryption method on the decimal values using specified key
(64646464)
To get the ID
U2FsdGVkX18WeQCa34kWz3feWGYBrYFbvGQ46rSAzS2n4GAG8RgFY44Ix0
odpFMsvvFUB3QSyZICkxYU4LgTLA==
Step7:- Store original characters, key, ID with the specific user information in
identification database.
Step8:- Inform the user with his/her ID.
The first phase will be used once in the beginning of using the E-government system.
Second phase
Step1:- The user will enter his/her ID
426
(ICoIT'17)
10thof April, 2017.
U2FsdGVkX18WeQCa34kWz3feWGYBrYFbvGQ46rSAzS2n4GAG8RgFY44Ix0
odpFMsvvFUB3QSyZICkxYU4LgTLA==
Step2:- The corresponding ID key will be read from the specific database which is now
64646464
Step3:- Apply AES decryption using ID as ciphertext with the mentioned key. The
result will be as follow:-
32013202320032331211121212131220
Step4:- Convert decimal values to DNA codes using table (2).
TGAC TGAG TGAA TGTT CGCC CGCG CGCT CGGA
Step5:- Convert DNA codes to binary representation using table (1).
0110000101100010011000110110010001100101011001100110011101101000
Step6:- Convert binary representation to ASCII values.
097 098 099 100 101 102 103 104
Step7:- Convert ASCII values to characters.

abcdefgh
Step8:- Compare the resulted characters after the process with the stored characters if
there is match allow access and perform required action otherwise ignore the access.
6. Evaluation of the proposed method
The proposed authentication method has good security level since if the unauthorized
users want to access the specific system under the control of the E-government needs to
know the following information:
1. The entered characters that are used to generate the ID. Since there are 256
characters and the number of the accepted entered characters lengths are (8
characters, 16 characters, 24 characters) so the possibilities as follow (8256 ,
16256 , 24256).
2. The arrangement of the based DNA codes. Since there are four DNA codes , so
the number of the possibilities is (44=256).
3. The AES encryption key.
427
(ICoIT'17)
10thof April, 2017.
7. CONCLUSIONS
E-government security system is the more important issues today. In the proposed
method three levels of authentication are provided. The authentication level based on
the ID length that is used if the ID length is 64 (8charcaters) bits then the authentication
level is the first, if the ID length is 128 (16 characters) bits then the authentication level
is the second, otherwise the authentication level is third 192 (24 characters ). The based
ID generation based on DNA Coding with AES.
REFERENCES
[1] Renta Mchov, Martin Lnnika1 , Reframing E-Government Development Indices
with Respect to New Trends in ICT, REVIEW OF ECONOMIC PERSPECTIVES ,VOL. 15,
ISSUE 4, 2015, pp. 383411.
[2] Jinan N. Shehab 1, Hussien Y. Radhi 2, Ruwaida A. Ibrahim3 , Multimedia
Cryptography Based On Liu And Chen Systems , Diyala Journal of Engineering Sciences ,
Vol. 09, No. 04, pp. 24-35, December 2016.
[3] Mohammad A. Alia, Abdelfatah Aref Tamimi, and Omaima N. A. AL-Allaf ,
Cryptography Based Authentication Methods , WCECS 2014, 22-24 October, 2014, San
Francisco, USA.
[4] Menezes, A., P. Van Oorschot, and S. Vanstone, Handbook of Applied Cryptography,
CRC Press, pp.4-15, 516, 1996.
[5] Hasanen Samir Abdullah ,An Improved DNA Computing Approach Using Heuristic
Techniques , PhD Thesis ,computer science department ,University of Technology
,2008,Iraq.
[6] Palvia, S. C., Sharma, S. ,E-Government and E-Governance: Definitions/Domain
Framework and Status around the World , (2007). Hyderabad, India: IECG.,
[7] Dr.Alia Karim Abdul Hassan Huda nahi ali , Proposed Information Protection method
Using RNA
Computing , Iraqi Journal of Information Technology , Year: 2016,Volume: 7 Issue: E2
Pages: 118-136.
[8] Mansoor Ebrahim , Shujaat Khan, Umer Bin Khalid , Symmetric Algorithm Survey: A
Comparative Analysis, International Journal of Computer Applications (0975 8887)
Volume 61 No.20, January 2013.
[9] Dr. Alaa Kadhim , Sura Khalaf , Proposal New S-box for AES Algorithm Depend on
A.I Bee Colony , Eng. &Tech.Journal, Vol. 33,Part (B), No.1, 2015.
428
(ICoIT'17)
10thof April, 2017.
Security Issues in Social Media: Challenges and Solutions
Mardin A. Anwer
Software and Informatics department, College of Engineering, Salahaddin University - Iraq
mardin.anwe@su.edu.krd
Rina Dinkha Zarro
Software and Informatics department, College of Engineering, Salahaddin University - Iraq
rina.zarro@su.edu.krd
Kamaran HamaAli Faraj
Computer department, College of Science, University of Cihan-Sulamani KRG
Computer department, College of Science, University of Sulamani-KRG
kamaranfaraj@yahoo.com

The popularity of social networking becomes very important
Article History: for the societies. It makes life easy and modified the
Received: 15 March 2017 traditional society (TS) to electronical society (ES). Thus,
community by using of social networking in every day is
more than usual and becomes part of our daily life and
Published: 10 April 2017 without doubt sharing with our life and all around human
Keywords: Privacy, daily activity helped by Facebook and it is a new way of 21
Personally and business century communicates. The average use of social media by a
Information, Social normal person is seven hours per day. Social media users
Media, secure share ideas, information and communicate with different
architecture, secure hash people. This communication reveals a lot of information such
function, two-factor as private and business information that makes it visible to
authentication. anyone who wants to view it or it become cross information.
Cross information means seen and exchange both of users
information. Subsequently, other people are intending to take
advantage of this information. They can send spam or steal
identities to sell it or use it for other purposes. This paper
introduces new architecture to protect users accounts that can
be use in Facebook. We also introduce some techniques that
can be used in any types of social media networking. In
addition, our proposed system architecture is increasing the
security issue by using of two-tier architecture (2TA)
authentication especially enhances user privacy in social
website. The new contribution of our proposed system will
insure no fraud account or fake identity is created in the
network related to the business, as well as, people related to
this account, all data and information will be kept private and
429
(ICoIT'17)
10thof April, 2017.
less spam attacks will be guaranteed. In order to evaluate the

proposed architecture tests, some profiles of academic staff
and employees in the college of engineeringSalahaddin
University were inspected.
1. INTRODUCTION
nline social networks are a kind of virtual communication (VC) that allows
O
people to connect with each other. The importance of staying together as a
community is the main idea of social network [1]. Unlike traditional media
(TM), social media relies on user-generated content, which refers to any
content that has created by end users or the public as opposed to professionals.
TM such as radio, books, and network television (TV) is primarily design to
be a broadcast platform (one-to-many), whereas social media is design to be a dialogue
many-to-many (M2M) interaction [2]. There are many types of social network, which
depend on its features. These features are geographic location like Orkut in Brazil,
VKontakte in Russia or well-known globally, like Facebook and Twitter or business-
oriented like LinkedIn and Xing [3]. Online social media is fully electronic because of
participating users and publics and it is (M2M), but TV is semi electronic online social
networking because of (one2Many), the traditional for example books is a book for a
user. In 1997, the first social networking media emerged. It was called Sixdegrees.com.
This site permits the user to create profiles, list their friends and surf the friends list [4].
Starting from 1997 to 2010 there are some 1.5 billion users of social networking
websites [5].
This paper is included six sections; related work is outlined in section two, Section
three is presenting different kinds of social media-where some statistical data will be
introduced about the usage of these kinds, section four explains the privacy issue of
using the social media, also the proposed architecture and analysis of the solutions
related to security is detailed in section five, Section six is about finding and discussion
when applying the proposed design. The final section draws conclusions of the research.
2. Related Works
Many research and white papers discussed the security and privacy issues in social
media network. They concluded many solutions related to this concern. Most of them agreed
that the user should be aware of the information that posted in the social media. Kevin
(2013) did an experiment on 486 students from the University of Cape Town (UCT), South
Africa. The data also shows that UCT students perceive friends and total strangers to be
their main audiences on Twitter; the attitude of UCT students towards Facebook remained
positive, on the other hand, a less positive attitude was experienced from the students using
Twitter; and Facebook is a more popular method for communication between students. The
results obviously highlight the changes in usage, attitude and perception of Facebook [6].
Brad in 2011 inspected some of risks in three social media websites which are Facebook,
Twitter and LinkedIn and identify. He presented possible solutions to help protect the user,
users personal information and users company data [7]. Brand explained how is possible
to identify the legitimate messages from the hoaxes by doing simple steps. Using an up-to-
430
(ICoIT'17)
10thof April, 2017.
date email client and date antivirus/anti-malware software, also never open an attachment
unless it is from someone you know are the main advices from Brand. Giles in 2007
highlighted and addressed privacy and security risks associated with Social Networking
Sites. He provided recommendations on how to minimize these risks. He offered strategies
to improve privacy and security without compromising the benefits of information sharing
thus increasing the overall social value of Social Networking Sites. He promoted Stronger
Authentication and Access-control where appropriate Threats: Digital Dossier, Ease of
Infiltration, Squatting, Spear Phishing, Stalking, SNS Spam, and Social Aggregators.
Similarly, increase transparency of Data-handling [8].
In 2013, Bill et al discussed some of dangers when using social media and how to use
these sites more safely. For example, he explained how many social networking sites allow
user to use encryption called HTTPS to secure the connection to the site. Some sites like
Twitter and Google+ have this enabled by default, while other sites require user to manually
enabled HTTPS via account settings [9]. In the same year, Abhishek et al proposed an
architecture for secure request response exchange of data between users. This architecture
improves the customization of profiles. It suggests that only a proper knowledge of the
hacking strategies will prove the best defense in the war against cyber-attacks. According to
their architecture, the visitors or friends request for any information to the application
between the visitor and the user. The application requests to the user for the response then
the user can response from any one of the databases according to his trust on the person who
has requested for the information [10].
3. Types of Social Media
Although there are many brands of social media which have its own unique architecture
that shapes the types of interactions that can occur [11], they all have the same target, which
is sharing a mutual interest and gain a method of interaction or information sharing through
the service. Services vary in their scope, the pace of interaction and the type of content that
is shared such as videos, images and text. The services control the type of connection between
the users, items and data retention policies [12].
Indeed, small changes in the design of social media tools and policies around them can be
vital to their success and failure [13]. Table one list the top ten social media on 2016
according to ebizmba website.
TABLE 1. Top 10 social networking sites in the world [13]
Rank Site Estimated Unique Monthly visitors
1 Facebook 1,100,000,000
2 YouTube 1,000,000,000
3 Twitter 310,000,000
4 LinkedIn 255,000,000
431
(ICoIT'17)
10thof April, 2017.
5 Pinterest 250,000,000
6 Google Plus+ 120,000,000
7 Tumblr 110,000,000
8 Instagram 100,000,000
9 Reddit 85,000,000
10 VK 80,000,000
In this research will focus on Facebook, however Instagram is similar to Facebook

where the user can view posts, images but still cannot be posted from a Web interface without
a workaround [14].
4. Privacy Issues on Social Media
A main concern about using social networking sites is privacy. Privacy can be defined as
the process of respecting the desires of individuals where compatible with the aims of the
larger community. Privacy is not about what people expect but about what they desire.
Privacy is not merely an individual right; it is an important component of any flourishing
community [15]. When you are online, it means that nothing is private. The problem is that
all the information, posts and images that is posted online can last for very long time.
Consequently, these posts can be shared all around the world. In 2007, Pew Internet &
American Life conclude from their study that 66 percent of teens restrict their online profile
by making it private. They found 82 percent teens post their first name, followed by photos
of themselves (79%), name of city (61%) and email address (29%). The situation is different
for girls. They commonly do not post information that could help in finding their physical
location [16]. During the last ten years, academics explored the threat to privacy linked to
social media. Gross and Acquits in 2005 analysed 4,000 Carnegie Mellon University
Facebook profiles and outlined the potential threats to privacy contained in the personal
information [17].
There are privacy settings in every social media site. The user can control their privacy
by selecting the best options that suite the users situation. Figure 1 shows the privacy settings
in Facebook.
FIGURE 1. Privacy Settings in Facebook
432
(ICoIT'17)
10thof April, 2017.
An Instagram users can set their privacy preferences such that their posted photos and
videos are available only to the users followers that requires approval from the user to be
their follower. Such actions will appear in referenced users Updates page so that users can
keep track of likes and comments about their posts. Given these functions, Instagram is
regarded as a kind of social awareness stream [18]. Like other social media platforms such
as Facebook and Twitter [19]
In Twitter social media site, the user has the option of publicly displaying their tweets or
keeping them private. Figure 2 shows the privacy Settings in twitter.
FIGURE 2. The Privacy Settings in Twitter
In spite of, the security provided in privacy settings, there are lots of threat regarding this
matter. In July 2009, the wife of the chief of the British secret service MI6 posted highly
revealing details on her Facebook page. This information included Family details, personal
photos and location of their home. This led 200 million people to access this information.
This resulted to national security risk Sophos survey 2010 [20].
5. Privacy Solutions and Proposed Architecture
In this section, two-tier authentication for signing in a social network will be introduced.
Then some techniques that might be adopted to increase the privacy in social media
networks.
5.1 Proposed Architecture for social network

Toward ensuring that the user post correct information in their timeline or be sure that he
is the person that he claim to be, the proposal is to scan any ID that confirm this information.
Asking the user to scan his or her ID will prevent user to create fake account and pretend to
be someone else. Lets suppose that an employee who works in organization AB want to
create a Facebook page and introduce himself as an employee in AB Company. It is
important that the employee create his account while he is at his office and during week work
time. As it is obvious from the diagram when the employee start creating his account in social
media the IP address and the location of the company will be automatically discovered. In
this case, no user will claim working in specific company even if he steals or use one of
another employee ID. This will make the job easy for the government to trace any fake
account or any person that involved in blackmail. All the information in the ID will be
extracted using OCR engine. In addition, if the phone of the user was lost or stolen, the
433
(ICoIT'17)
10thof April, 2017.
person who has the phone cannot access the user account because he must use ID for logging
in as shown is figure3. When the user enters his username and password to the system these
data will be encrypted using secure hash function (SHA-2). A hash function or one-way hash
function can be defined as a function that takes an input as arbitrarily long string of bits (or
bytes) and produces a fixed-size result [21]. There are SHA-0, SHA-1, SHA-2. SHA-1 is
fast, but its speed is a weakness because brute-force attacks can perform faster. Therefore,
using SHA-256 (SHA-2) provides more protection against brute-force attacks in spite of
being a slower hash than SHA-1. Figure 3 present the proposed architecture for signing up
Facebook as an example.
FIGURE 3. The proposed architecture for signing up in social media (Facebook as an

example)
The same process is required for normal user. For normal user, it has to categorize them
as student or user (any other type except employee and student). If the user is a student, then
he should create his account while he is at school. In this case, the location of the school and
the ID will be officially identified. For other users, they can create their account at home and
use passport for introducing themselves and use their Gmail, yahoo, Hotmail or any ordinary
email account. The process of logging in required the user to enter the username, password
and his ID. The username and password will be hashed again using SHA-2 and will be
compared with the one that is saved in the social media database. If they are correct, then the
system will ask the user to scan or upload the ID that is used before in the signing up process.
The extracted information for the ID will be compared with the one is saved in the database.
If they are the same then the access is granted. When the user enters the work email, a copy
of his email should be directed to HR (Human Recourse) office or the manager. Using
business email confirm the identity of the user as an employee in this company. This means
the responsible is aware of the employees who use their work identity in their social media
account. Figure 4 shows the general two tier architecture of our proposed system. Database
as a backend is sql server and the middleware is C# for communicate backend and frontend.
The connected frontend is a scanner to scan ID and send it to database. Two tier architecture
is more than enough for the staff academy of college of Engineering- Salahaden University.
The new contribution of our proposed system is security on 2TA. Nevertheless, Figure 5
shows the process of signing in of users.
434
(ICoIT'17)
10thof April, 2017.
FIGURE 4. Two tier architecture
FIGURE 5. The process of signing in
5.2 Technique for Ensuring Security in Social Media
In Instagram, there is the problem of viewing all followers and following. Talking about
privacy term it can be said that Facebook introduce more privacy option for their users. Thus,
Instagram can provide options that make the account as private and private followers and
following. Figure 6 shows the recommend type of account as proposed.
Photos and videos that the user share in social network can give others a clear picture of
users life. It is easy to share photos that show your house number, street name, children's
names or office location. Yet, these pictures can be used for other issues. Many social media
network users claim that their pictures are downloaded and then uploaded by unknown
person or using these pictures as type of threating. Some of social media user upload their
pictures in many occasion and some of them might be embarrassing. Today many employers
look online profile before giving the job offer, because they do not want to see only one side
435
(ICoIT'17)
10thof April, 2017.
of employees personality. The president of a consulting company in Chicago decided to

check one of the candidates Facebook page, and found descriptions of marijuana, shooting
people and obsessive sex. Finally, the candidate was rejected for this [22].
FIGURE 6. The recommend type of account

The better solution for this issue is adding an option called Allowing download picture
or Allow screenshot the picture. This will restrict the public from using these picture for
unwanted matter. Figure 7 shows the setting after adding this option.
FIGURE 7. Social media settings after adding Allowing option
Also, a message can be added to notify the visitor that the account is protected from
screen shot as shown in figure 8.
FIGURE 8. Warning message about downloading picture
If the user want to control the process of allowing friend or public to save photo then an
option of allow saving would be added. Figure 8 illustrate this option.
To know the number of users that download the picture a button can be added for this
matter as shown in figure 10.
436
(ICoIT'17)
10thof April, 2017.
FIGURE 9. Permission setting for downloading and saving image
FIGURE 10. Notifying the user about the number of downloading picture
However, there are many other website which is similar to Facebook and Instagram where
they ask people to log in to their account to check new photos of their friends. When the user
opens these websites, it looks like a normal social media account such as Facebook. Then
after, the user enters the user id and password, without knowing that the account details are
being stolen by a hacker. In Instagram, if you have a public account, anyone can see your
images. With a private account, only your approved followers can see the photos and videos
you share. This option secures your picture from unauthorized users.
6. Finding and discussion

It was very important to select an organization or company to investigate the social media
account of their staff. Nonetheless, working in college of engineering and planning unit give
us the idea of selecting it as a case study. In addition, the fact of trusting a person to inspect
their account would not be easy in any other place. That is why the college of engineering is
selected. Although, there are many types of social media, but Facebook is selected because
it is widely used by both academic staff and employees. Before applying the proposed
architecture, all pages named as college of engineering Salahaddin Universitys Facebook
were located. Unsurprisingly it was found that there is more than one page. This means that
the college could not control the problem of creating other fake account created by
unauthorized persons. Moreover, it was found that every page claims that it is an official one
and users were confused about the official one. Figure 11 shows the different copies of the
existing college page.
437
(ICoIT'17)
10thof April, 2017.
After locating the official page of the college, the page information was checked. The
result was that not all the provided information is correct as shown in figure 12.
FIGURE 11. Different copies of college of Engineering- Salahaddin University Facebook page
FIGURE 12. Page information of college of Engineering- Salahaddin University

Facebook page
Next, after obtaining permission from some academic staff and employees of the college,
the information that they provided related to their identity, academic title, position and place
of work was investigated for authenticity and authority checking. While reviewing 65
accounts (academic and employees) the following are discovered:
Most of the employees (nearly 70%) do not provide work and personal information about
themselves. Where this was less in academic staff nearly 30% as shown in figure 13.
FIGURE 13. Lack of provided information in both academic and employees
Nearly 60% of provided information from academic staff is fake. Some entered academic
title which is not correct. There are lots of irrelevant information about their background an
example is shown in figure 14. While this occurred less in employees profile (nearly 20%).
438
(ICoIT'17)
10thof April, 2017.
FIGURE 14. Irrelevant information in Introduction

To start applying the proposed architecture, the employees in the deanery were asked to
open new account using their ID information. Each employee, provided a scanned copy of
their ID. Eventually, the authenticity of the information was guaranteed and accessing or
request made to the college site where more secure and controlled. The following figure
shows the differences between the existing Facebook architecture and proposed architecture
according to some security aspects.
As it is shown from the chart, the proposed architecture provides accurate information
about the user that want to create account associated to his business. It is crucial that all staff
in collage identify themselves with correct information, so it is easy for their colleagues and
students to find them. This is also true for any other search engine. As a result, the provided
architecture guaranteed two main security aspects. First, when creating accounts related to
any company or governmental establishment, the authentication using ID will insure no fraud
account is created in the network related to the business, as well as, people related to this
account. This will insure all the news and announcements will all be authentic.
The differences between the
current and proposed architecture
120%
100%
80%
60%
40%
20%
0%
The Current Architecture The Proposed Architecture
Privacy Identical Personal Work Irrelevant Spam
Name Information Information Information
Figure 15. The differences between the current and proposed architecture
Second, the architecture will guarantee the accounts want to associate to the main business
account will all be checked and authorized by IT administrator. In this case, all data and
439
(ICoIT'17)
10thof April, 2017.
information will be kept private and less spam attacks will be guaranteed. In addition,
following the above techniques will prevent fraud and hacker to download another users
picture and pretend to be them. By following the techniques, the user can give permission to
people who know to download their pictures and create private account.
7. Conclusion
Several seminar-evaluations were hold in Software engineering department regarding

social media and the entire seminar participant realized that will be very easy to create fake
accounts under specific name. As there is no authentication and authorization checking. That
is why it is difficult to be sure that the named account referred to the correct person or
business. This means that the security in social media is weak and vulnerable to hackers.
Hackers can use these accounts to send spam, spread malware, and steal identities in the
quest to acquire personal information for financial gain. In this paper, some techniques to
secure user account are suggested. Moreover, a different architecture is proposed for
handling security. In this architecture, two-tier authentication is required. When the user
signup to the social media site, user name, password and information in the ID card must be
entered. The information on the ID is extracted and saved in a specific database based on
users type which is employee or student or normal user. This means-costuming database
depends on the type of user to grant access depending on user credentials.
8. Acknowledgment
We would like to thank all who give us the permission to access their account and express
their fears of using social media. They inspire us to think deeply about writing a research
about providing secure privacy in social media network.
References
[1] Das B., Sahoo J. (2012) Social Networking Sites A Critical Analysis of Its Impact on Personal and
Social Life, International Journal of Business and Social Science Vol. 2 No. 14. www.ijbssnet.com
[2] Porter, J. (2008). Designing for the Social Web. Thousand Oaks, CA: New Riders Press
[3] West C.,(2010) The Risks of Social Networking Retrieved from

https://www.symantec.com/content/en/us/enterprise/media/security_response/whitepapers/t
he_risks_of_social_networking.pdf
[4] Boyd, d. and Ellison, N. B. (2007). Social network sites: definition, history, and scholarship. Journal of
Computer-Mediated Communication, 13(1). Retrieved from http://jcmc.indiana. edu/vol13/issue1/
boyd. ellison.html
[5] Kreutz, C. (2009). The next billion the rise of social network sites in developing countries. Retrieved
from http://www.web2fordev.net/component/content/article/1-latest-news/69-social-
networks .
[6] Johnston K., Chen M., Hauman M., (2013) Use, Perception and Attitude of University Students Towards
Facebook and Twitter ,The Electronic Journal Information Systems Evaluation Volume 16 Issue 3 2013,
440
(ICoIT'17)
10thof April, 2017.
(201-211) , ISSN 1566-6379 201 Academic Publishing International Ltd Reference this paper as: available
online at www.ejise.com
[7] Dinerman B.,( 2011) Social networking and security risks, GFI White Paper.
[8] Hogben G.,(2007),Security Issues and Recommendations for Online Social Networks. ENISA,
October 2007 E. Retrieved from http://www.ifap.ru/library/book227.pdf
[9] Wyman B., Scrivens W., Hoffman P., Spitzner L.,(2013) Social Networking Safely Retrieved from
http://www.securingthehuman.org
[10] Kumar A., Kumar G., Rai A., Sinha S. (2013) Social Networking Sites and Their Security Issues
International Journal of Scientific and Research Publications, Volume 3, Issue 4, April 2013 1 ISSN 2250-
3153 www.ijsrp.org.
[11] Lessig, L. (2006). (2009). 4 reasons why e-records are still a mess. Federal Computer Week. Code 2.0.
New York: Basic. Lipowicz, A. Retrieved from: http://www.fcw.com/Articles/2009/03/09/policy-email-
records.aspx.
[12] Hansen, D. L., Shneiderman, B., & Smith, M. A. (2011). Analyzing social media networks with
NodeXL: Insights from a connected world. Burlington, MA: Morgan Kaufmann.
[13] http://www.ebizmba.com/articles/social-networking-websites
[14] Danielle S. (2013), Moving on from Facebook Using Instagram to connect with undergraduates and
engage in teaching and learning
[15] Lenhart A.,Madden M.,(2009), Teens, Privacy and Online Social Networks
. Retrieved from http://www.it.cornell.edu/policies/infoprivacy/definition.cfm
[16] Lenhart A., Amanda B., Madden B., Mary M.. (2007). Teens, privacy &online social networks.
Retrieved from http://www.pewinternet.org/Reports/2007/Teens-Privacy-and-Online-Social-Networks/1-
Summary-of-Findings.aspx
[17] Gross, R. and Acquisti, A. (2005). Information revelation and privacy in online social
networks. Proceedings of WPES05 (pp. 7180). Alexandria, VA: ACM.
[18] Naaman M., Boase J., Lai C.-H. 2010. Is it really about me?: message content in social awareness
streams In CSCW.
[19] Hu Y., Manikonda L., Kambhampatihttp S., (2013) What We Instagram: A First Analysis of
Instagram Photo Content and User Types .Retrieved from : //149.169.27.83/instagram-icwsm.pdf
[20] Lewis J.,(2009) MI6 chief blows his cover as wife's Facebook account reveals family holidays, showbiz
friends and links to David Irving. Retrieved from http://www.dailymail.co.uk/news/article-1197562/MI6-
chief-blows-cover-wifes-Facebook-account-reveals-family-holidays-showbiz-friends-links-David-Irving.html
[21] STALLINGS, W.(2002) Cryptography and Network Security. Prentice Hall.
[22] Maloney-Krichmar, D., & Preece, J. (2005). A multilevel analysis of sociability, usability and
community dynamics in an online health community. Transactions on Human Computer Interaction, 12(2),
132.
441
(ICoIT'17)
10thof April, 2017.
SMS DISTRIBUTER BASED PATIENT APPOINTMENTS SYSTEM
Mohammad Salim Abdulrahman

IT Department, LFU University - Iraq
mhamditp@lfu.edu.krd
Rebin B. Koshnaw
Computer science Dept. College of Science, Salahaddin University- Iraq
rebin.saleh@su.edu.krd
Dr. Mazin S. Al-Hakeem
IT Department, LFU University Iraq
dr.mazin@lfu.edu.krd

Models with advanced access in general practice have been
Article History: introduced worldwide in order to enhance healthcare services
Received: 20 March 2017 throughout last decade. This has led to obtaining benefits such
Published: 10 April 2017 as eliminating waiting lists, enhancing patients timely access
to services and lowering no show. Despite, implementing
Keywords: SMS these models in Kurdistans public hospitals, more precisely,
distributer, in places with no internet access or low internet literacy, these
Appointment, SQL systems will be less used. The current paper work
Server, ASP.NET. appointment systems introduce one of the main problems
arise in hospitals which is missing appointments by the
patients due to the long queues given by the secretary of
doctors. In the high light of this problem, there have been
several web applications developed to show the exact date of
appointments, however, according to latest report by UNDP
report on Iraq development more than 50 % of patients in
Erbil city dont have internet access or they dont know how
to use internet. We are proposing an optimal solution for this
problem which is an SMS based patient appointment system.
Patients will be notified by an SMS a day before the actual
appointment after they have been to hospital and registered
themselves into an online registration system by the secretary.
These are carried out in practice using SMS distributer, SQL
Server database and ASP.NET and C# programming
languages.
1. INTRODUCTION
T
oday patient management process which includes patient appointment, health
forms, health records, and priority calls is considered important because it
supports health care sector and reflect its quality. For the purpose of improving
442
(ICoIT'17)
10thof April, 2017.
health care sector, its imperative to look at the process of patient management such as time
management and how to make more efficient. The biggest part of complaints made by
patients are concerned with the longtime of waiting between patient arrival to a clinic and
the actual starting time of consultation. One of the main problems of patients is the lengthy
time during patient appointment reservation till the time of actual meeting with the related
doctor or consultation [1] [2]. Moreover, appointments periods are not systematized since
there are no fixed number of appointments per day, and patient might miss or forget
appointment date which lead to cancelation of appointment and repeating appointment
reservation process for new period if the patient asked again for appointment. Therefore
many patients are not satisfied with current traditional system of appointments, especially
those having urgent needs. This problem caused by missing of appropriate system that could
re-arrange appointments and remind patients about their consultation time to reduce raised
delay and number of missed appointments [3] . Figure 1 shows the current general flow of
patients appointment to the consultation hour in Kurdistan region, and this flow needs
improvement [4]. Usually, patient will visit or contact the dentistry secretary to book an
appointment then waits for that particular date and time of appointment. As soon as
appointment date is approached, patients need to fill in registration forms in arrival. Next,
from the time of registration till consultation hour, the waiting time of patients could be
separated into these two parts:
Patients earliness Period spent from early arrival till the designated appointment
time.
Internal waiting time Time of waiting for the queue before starting the
consultation hour.
443
(ICoIT'17)
10thof April, 2017.
Appointment
Waiting Interal
Appointment or
Reservation Appointment waiting time
Confirmation Consultation
Time or Date
Time
Due to queue or
Request Geting dentist's delay
Appointment appointment date
and confirm it
Figure 1: Outpatient stream inside a dentistry
The common process within dentistry or health clinics is based on first come, first
served system, where two available methods of getting a consultation :
A- Visit the dentistry on the appointment day and wait if any queue is there.
B- Call the dentistry on the appointment day and ask about the suitable time to come to
avoid any internal waiting time such as queues time.
For the last two decades, IT has made big innovations which become available to
organizations and people, together with the health sector. In the past couple of years, there
are significant developments in health sector to assure the performance quality and standards
effectiveness of healthcare and its services [5]. In [6] it is stated that the Internet generally
holds a big influence toward developing the healthcare system as compared to all other
technologies that been adopted a few years ago. Nevertheless, because of some struggle
related to guarding confidentiality of the patients records which hinder the greatest
expectation of the Internet in health sector.
To avoid burden of queues and manual hand reservation for appointments, scheduling
appointment via phone or messages (SMS) is considered one of the best solution. Regarding
the appointment scheduling no reminder system is there on all platforms for appointments
which have been affirmed by the dentistry secretary [7]. Our proposed system is designed
especially for public clinics as there is huge traffic on them with visitors and the appointment
is not the same day, so they will be asked to comeback in few days ,and sometimes there will
be misunderstanding and patients forget the exact day. That is why we have designed the
reminder system specifically for clinics especially dentistry.
444
(ICoIT'17)
10thof April, 2017.
This paper organized as follows. Part 1 introduction about patient smart notifying system,
then Part 2 which state some literature about IT utilization in the health industry concerning
patient appointment. Part 3 shows the architecture of the proposed smart patient notifying
system, followed by part 4 on the implementation of our recommended system and its
operation. Part 5 provides the conclusion plus some future work.
In this section we are going to explore in brief on the adoption of various systems for
patients reminder on their appointments in health sector.
2.1. PAPER BASED APPOINTMENT SYSTEM

In Iraq and Kurdistan region still wide part of health sector not taking advantage from
technology development. It should be clear that in an outpatient clinic the process of
consultation involves registration process plus appointment scheduling procedure that is
paper-based in most clinics and dentistry especially the governmental ones and those situated
outside the city [8]. Usually patients health records and files are kept in tangible storage,
and will be passed by administrator or registrar to the office of doctor for appointment. This
traditional system or method is not efficient and taking a long time, and the possibility of
losing or mislaying records is obvious. The paper-based registration generally demands that
patients should fill the forms and give to the registers table then wait for their turns till
calling the names. One of the important issues what would happen if unregistered patient
came or a patients records are not to be found? Usually the registrar or admittance staff will
delay or postpone his/her consultation or appointment time. According to the survey report
in 2008 [9], majority of the complaints related to the time expended in the waiting hall and
19% of the outpatients criticized that they could not get their appointments within one week
time. Moreover, hundreds of outpatients appointments are postponed to another day more
than one time. Gratification of patients with health care process didnt associate only with
waiting time but then over one-third decided to not follow the appointment schedule as they
anticipated to have to wait for long times. This problem still considered as a challenge to
health care industry in Kurdistan region and worldwide. In our solution we propose as system
to deal with these issues, this system will enable the registrar to fill in patients information
into web form then these data will be saved into database . Later before one day of the
appointment the system will send SMS reminder to patients to remind them on the
consultation time.
2.2. ONLINE APPOINTMENT SYSTEM

Another type of systems being used worldwide is online registration and scheduling
system for instance (RegisterPatient.com). This website works 24 hours plus everyday per
week system and health care registrar is reported automatically on future schedules and
recorded patient via the online web system. This system enables patients to fill out their
registration forms at a suitable time to avoid a queue. After registration has been finished,
the patient will be notified about the nearest clinic. On the other hand, there are some issues
concerning this system. First of all, registration demand patients to fill in sensitive data such
445
(ICoIT'17)
10thof April, 2017.
as phone, ID, credit card numbers, and etc. As this system is a web-based, these sensitive
data can be stolen by hackers. Another issue, most of the scheduling appointment and online
registration system require monthly payments to the provider. Thirdly, some online
appointment systems are restricted to less noticed symptoms such as vomiting, abdominal
pains, and diarrhea [5]. Therefore when some of the symptoms are not available on the web
system, patients must call in for any available appointment periods and be waiting for
confirmation. For these urgent situations of needing consultation, this inadequacy might be
leading to troubles for healthcare staff and patients. Lastly, this system does not offer any
prioritization for scheduling the appointments. Nonetheless, it has been noticed to have such
shortcomings and ineffectiveness to the online system. Firstly, reserved patients have to show
their consultation cards which fixed with barcode for scanning to inform attendance. The
inconvenience will appear once the card unintentionally mislaid, then inpatients are to be
entered as an outpatient instead. Furthermore, in the emergency case, the patient is obligated
to book physically at whatever clinic or by calling in and waiting for the confirmation by
registrar for attainable periods. Another case arise if an interruption of consultation happened
because of delay to access to the electronic web system. To access records of a patient, one
must set and present his/her ID to the registrar. As a final point, priority levels might be set
manually by depending on status of registration and age, regardless the illness level. At this
time, there is no enough research accessible in the improvement of prioritization in any health
care clinics throughout the world [10]. However our proposed electronic system will deal
with urgent cases and priority levels by and will not have a problem with slow access to the
system because only one person (registrar) or administrator will create the records for the
patients.
2.3. MOBILE APPOINTMENT SYSTEM

Lately, health sector has started gradually moving towards employing cloud-based
systems, despite of security concerns which might happen or not. By the presence of these
systems, data size and capability to maintain data and records will be handled in a central
manner. This system will improve the effectiveness of shared data among health care
specialists nationwide, in addition to decrease the operating expenses of the health sector
mostly since there will not be any repetitions. On the other hand, cloud networking are still
rare where just a few IT experts have the ability to manage well the structure of database
besides security issues. Application of Intelligent Agents in Hospital Appointment
Scheduling System uses an intelligent agents that excludes the human ones to perform the
related search actions in scheduling processes aim to improve the workflow and accordingly
saving the effort and time of healthcare staff [5]. This system offers the consultation and
arranging of the appointments by adopting mobile applications, Android 2.2 Froyo. The
mobile system relied on setting the precedence level of patients into appointment booking.
The developed agents are working using fuzzy orientation, to collect information from
patients and schedules appointment with the health clinic. The testing of this system was
performed by the use of JADE-LEAP in Android 2.2 Froyo mobile OS. There are some
disadvantages discovered in this system. Initially, unregistered patients are demanded to
book an appointment before twenty four hours or more before the planned time. After
approval, these patients will automatically became a registered patients. In urgent cases, this
method certainly not useful. Moreover, it only deals with one central health center,
446
(ICoIT'17)
10thof April, 2017.
nevertheless of clinics nearby. Finally, there are no automated system calls as a reminder
before the scheduled date. Our proposed reminder system will tackle these issues by sending
SMS automatically to patients based on the data entered by the administrator or clinics
secretary.
3. METHODOLOGY
The aim of designing appointment reminder system is to reduce waiting time and make
scheduled appointments well-arranged by setting fixed number of appointments per-day. The
reminder system will be able to send reminder SMS to patients one day before the
consultation, by fixing the number of appointments per day so each day will have a group of
patients and their number can be set by the clinic administrator or doctors secretary.
Regardless of deep research from previous work in enhancing operations of healthcare, some
problems are still unsettled. There are still concerns for walk-in patients; sometimes they
must wait for a long time to reach their turns. Patient may schedule the appointment; but,
they still need to wait for their appointment to be affirmed by administrator of a clinic. The
process of waiting for scheduled appointments may take weeks or even months for any
available periods. Both of the mentioned scenarios are not appropriate for all parties: patient,
healthcare personnel and the healthcare clinic. This proposed reminder system would help as
a queuing system to access patients appointment information which will alert patients before
12 hours of the actual appointment by an SMS message. This system focusing only on the
basic outpatient standard system such as registration, and scheduling appointment. Figure 3
below shows the proposed system layers.
Layer 1: User Interface

Security
Layer 2: Server
Layer 3: Database Server
Layer 4: SMS Distributer KOREK Telecom
Patients Phones
Figure 3: Four basic layers of patients reminder system
3.1. LAYERS OF THE PROPOSED SYSTEM
For the purpose of creating the system that we proposed to remind patients on their
appointments, the researchers have developed the system by working on these four layers:
447
(ICoIT'17)
10thof April, 2017.
1- User interface: this is considered the first step of using the system and it is represented
by a web-based screen. This screen will enable the user whether he/she is an admin
or a secretary to enter the data of the patients, figure 2 is a snapshot of this screen.
2- Server: This is a web server which host our web-based application to make available
for users which are admins and secretaries of clinics. We a host on a public server
which supports our web application.
3- Database server: this layer is responsible for saving all the data of the patients, so the
proposed system can use these data to automatically remind the patients on their
appointments before a specific number of hours. However, next layer (database
distributer) will work on notifying the potential patients.
4- SMS distributer: The main job of this layer is to send notifications to the patients
through SMS messages which is considered one of the most efficient way of
communication these days especially in rural areas where internet service is still
rarely used. This layer or service will utilized with Korek Telecom Bulk message,
because its the most popular telecommunication network in Erbil city of Kurdistan
region and most of the patients are clients with Korek Telecom.
3.2. USE CASE DIAGRAM OF THE PROPOSED SYSTEM

Figure 4 shows the use case diagram of the proposed web based application. There are
three main actors playing this application who are Admin, Patient and Secretary. The admin
has the capability of inserting new users for the system as well as modifying them and
deleting one when required. Admin can also for patients and modify them plus delete them
when needed. The patients will receive bulk messages from GSM companies for notification
purpose of the appointment. The secretary can insert new patients information plus a search,
edit and delete old patients.
Figure 4: Use case diagram for the proposed system
4. RESULTS AND CONCLUSION

This sections has three main parts, the first one will explore the results of implementing
the proposed system, second will present some feedback from involved parties to validate
this paper work, and third part which focus on drawing conclusions and gives potential future
work for researchers and industry people.
448
(ICoIT'17)
10thof April, 2017.
4.1. SYSTEM IMPLEMENTATION RESULTS
After the system has been implemented, its database is connected with a GSM company
using SMS distributer so that through Bulk messages, the notification message is sent to
patients based on the information presence in the database about the date and phone number
of the patient.
As it can be seen from figure 4, the secretary will be filling out a form containing all
required information on the patient specially the name and phone number. The form is
validated such that the user should fill all the fields. As soon as the form filled out, the data
will be stored on a public server on the internet. The database is connected to SMS distributer.
When the day of the appoint approaches, the patient will receive a notification message as it
can be seen in figure 5.
Figure 5: Patient Insert User Interface
Besides having internet-less notifications to patients, hospital visitors can access online
web page and search for their appointment by Name, visiting Date and Phone number. The
figure 6 shows that, using our web based application, the secretary can search for patients
too.
Figure 6: Patient Search by Phone number User Interface
449
(ICoIT'17)
10thof April, 2017.
Figure 7: Mobile SMS appointment reminder
As it can be seen from figure 7, the patient will receive a bulk message from the GSM
Company about the appointment before 12 hours. Through this way the appointment missing
rate will be reduced and patients arrive to the hospital on exact date and time.
4.2. VALIDATION OF THE SYSTEM
After the application implemented, we connected it the appointment database with Korek
Telecom using SMS distributer. We subscribed to Korek GSM using 2000 Bulk messages.
The application installed on the secretary work station and put under test for one week.
After seven days, the opinions on the application and the solution provided for appointment
procedure from several personals who used the system.
The general director of the Shaqlawa believes that the solution will reduce the
appointment problems to a great extinct which were causing troubles before. He thinks that
the application is very suitable for a place like Shaqlawa city since around %50 of people
dont have internet access or they dont know how to use the internet.
The general director of the hospital says that from now, our data is saved in a safe place
and we can retrieve it whenever needed in a very easy way. He believes that this system will
reduce the cost of paper in future.
The Doctors mentioned that the application have had the troubles facing with wrong or
missed appointment being solved. Doctors found the search feature for patients very useful
to know the number of patients they are going to treat in an exact day.
The secretaries believe that this system will reduce the force, work and miss-
understandings between patient and hospital in appointment issues.
The patients look comfortable with the system since they get the notification in a very
convenient way without having concern on internet and emails plus a correct day for
treatment.
450
(ICoIT'17)
10thof April, 2017.
5. CONCLUSIONS
1. The missing appointments by the patients due to the long queues considers the real
main problems arise in hospitals.
2. Using the proposed appointment management system, the percentage of missed
appointments by patients is decreased tremendously as most of the population in rural
areas have mobile devices regardless being a smart phone while no need for internet
access and knowledge on using the internet.
3. During this research, it is found the proposed SMS distributer based appointment
system is the optimal and the most efficient way to remind patients about their
appointment regardless having internet access such as in rural areas.
4. Using this system helps the clinics' secretary to manage the workflow of appointments
since the system can specify fixed number of daily appointments then remind them
automatically via SMS.
6. REFERENCES
]1[ F. P. Mardiah and M. H. Basri, "The Analysis of Appointment system to reduce
outpatient waiting time at Indonesias public hospital," Human Resource
Management Research, vol. 3, no. 1, pp. 27-33, 2013.
]2[ Z. Zhu, B. H. Heng, and K. L. Teow, "Analysis of factors causing long patient
waiting time and clinic overtime in outpatient clinics," Journal of Medical Systems,
vol. 36, no. 2, pp. 707-713, 2012.
]3[ A. Aeenparast, S. J. Tabibi, K. Shahanaghi, and M. B. Aryanejhad, "Reducing
outpatient waiting time: a simulation modeling approach," Iranian Red Crescent
Medical Journal, vol. 15, no. 9, pp. 865-9, 2013.
]1[ C. R. Anthony, M. L. Hansen, K. B. Kumar, H. J. Shatz, and G. Vernez, "Building
the Future: Summary of Four Studies to Develop the Private Sector, Education,
Health Care, and Data for Decisionmaking for the Kurdistan RegionIraq (with
Arabic-language version)," Rand Health Quarterly, vol. 3, no. 2, 2013.
]8[ Y. Symey, S. Sankaranarayanan, and S. N. binti Sait, "Application of smart
technologies for mobile patient appointment system," International Journal, vol. 2,
no. 4, 2013.
]7[ R. S. Sadasivam, R. L. Kinney, S. C. Lemon, S. L. Shimada, J. J. Allison, and T. K.
Houston, "Internet health information seeking is a team sport: analysis of the Pew
Internet Survey," International journal of medical informatics, vol. 82, no. 3, pp.
193-200, 2013.
]6[ W. Cao et al., "A web-based appointment system to reduce waiting for outpatients:
A retrospective study," BMC health services research, vol. 11, no. 1, p.2711 ,315 .
]5[ T. K. Al Hilfi, R. Lafta, and G. Burnham, "Health services in Iraq," The Lancet, vol.
381, no. 9870, pp. 939-948, 2013.
]6[ L. R. LaGanga and S. R. Lawrence, "Service Appointment Scheduling with Walk-
In, Short-term, and Traditional Scheduling," in Decision Sciences Annual
Conference, 2008.
]17[ D. Gupta and B. Denton, "Appointment scheduling in health care: Challenges and
opportunities," IIE transactions, vol. 40, no. 9, pp. 800-819, 2008.
451
(ICoIT'17)
10thof April, 2017.
Survey and New Security methodology of Routing

Protocol in AD-Hoc Network
Shahab Wahhab Kareem
Department of Information System Engineering, Technical Engineering College, Erbil Polytechnic
University - Erbil, Kurdistan Region, Iraq
shahabwk@yahoo.com
Dr.Yahya Tareq Hussein

Department of Information Technology, Lebanese French University - Kurdistan Region, Iraq
Dr.yahya@lfu.edu.krd.

Because of many reasons routing is the complex task in ad
Article History: hoc sensor network. The self-organizing network which is
Received: 17 March 2017 Mobile ad hoc network (MANET) its established
automatically via wireless connections by a combination of
portable nodes without the support of a centralized
Published: 10 April 2017 management or static infrastructure. The portable nodes
redirect packets among these nodes, enabling connection
Keywords: wireless between nodes outside the range of wireless transmission hop
sensor network (WSN); by hop capabilities are necessary. So it important is increasing
Multipath routing to provide suitable routing protocol plus security. This paper
protocol; WLAN; endeavors to present a full survey about secure routing. At the
Minimum Spanning beginning investigates the reason which makes ad hoc
Tree; Mobile Ad hoc network is susceptible to attacks. After that, it shows the well-
network. known traditional secure protocols then compares them.
Finally, the proposed algorithm depended on the Dynamic
Source Routing (DSR) to remove malicious nodes, minimum
spanning tree for routing path plus coding-encryption
technique of a chaos-based adaptive arithmetic for
encrypting, compression plus decompression, decrypt a
message.
1. INTRODUCTION
I
n every wireless networking environments in today's abundant use of communication
devices, nodes communicate by two methods either shortly with their planned
communication partner or in the form of base stations which is called networking
infrastructure as well as a backbone network. Example, the using 802.11 in ad hoc networks,
shows the components also the networks inside the Wireless Network Infrastructure. The
self-organizing network which is Mobile ad hoc network (MANET) its established
automatically via wireless connections by a combination of portable nodes without the
452
(ICoIT'17)
10thof April, 2017.
support of a centralized management or static infrastructure. In the portable ad hoc networks

every nodes provided by a receiver plus transmitter. This wireless receiver and transmitter
enable the nodes at the same radio communication range to contact with each other.
Habitually nodes participate with the same physical media; they transmit and collect signals
at the same frequency band, and follow the same step sequence or broadcasting code [1].
An attacker may produce malicious nodes that forward only certain messages and drop
others. In order to grab the route forcibly, a Black Hole attack can execute via only one node
which counterfeits the hops count and sequence number of a routing message. Then the Black
Hole node will snoop, or immediately discard the received data packets. The type of Denial
Services attack is a Gray Hole. Here the node forms wrong routing information in the
network. A Gray Hole discard just a part of the packets do not drop all the packets [2]. In the
proposed algorithm the malicious removed (black hole and gray hole) by using DSR routing
protocol in the Ad hoc network. After removing all malicious by using the minimum
spanning tree to select the shortest path routing protocol for sending and receiving a message,
finally Using Chaotic System and Variable Model Arithmetic Coding to make Compression
and Encryption Scheme.
This paper contains: Section 2 describes some algorithms for routing protocol, section 3
comparison between algorithms and designing proposed, in section 4 plan of work is discuss,
in section 5 the conclusion.
2. SECURE ROUTING
Protected routing protocols overcome malevolent nodes that be able to obstruct the
accurate operation of a routing protocol by adjusting routing information, via imitating
other nodes and by inventing false routing information. This Protected routing protocols
of ad hoc networks are either combinations of security mechanisms within the existing
protocols or entirely novel independently protocols. In general, the proposed secure
routing protocols was classified into two classes, routing protocols which use hash chains
and routing protocols that require predefined dependence relationships in order to operate
[1].
hash chains is A one-way hash function H(.)to be more secure in routing

protocol is to follow these properties:
1-H(.) can take a message of arbitrary length as input and produce a message
digest of a fixed-length output.
2- Given x, it is hard to compute H1 (y) = x given y. However, it is easy to
compute H(x) = y.
3- Given x, its arithmetically not practicable to determine x x such that H(x) =
H(x) [3].
453
(ICoIT'17)
10thof April, 2017.
Basic of Chaotic Systems Properties its designed via few simple function f
which is repeated on som set X. Especially, the maping f: XX is sid to be
chatic on X if
o Periodic points are thick in .
o f is topologically transitive;
o f has sensitive reliance on initial conditions;
In intuitive way, a map owns sensitive reliance on primary situations if there exist
points arbitrarily near to x which finally separate from by at least > 0 under
repetition of f, while a topologically transitive map has points that ultimately
transfer in repetition from one random minor neighborhood to any other [4].
In this paper, we presented a survey for secure routing and proposed secure routing based
on Paillier public key for encrypt and decrypt the Message and also used DSR to remove
these malicious nodes, and minimum span tree for routing path are used. The following
are some of the routing protocol discussed.
Methodology for Securing Multipath Routing

The Ad-Hoc Demand Distance Vector(AODV)
A Simulation Study of Security Performance Using Multipath Routing in Ad
Hoc Networks DSR.
A Novel Method of Secure Routing Protocol: for Improve Network Life.
2.1. THE AD-HOC DEMAND DISTANCE VECTOR (AODV)

This algorithm also can be called a pure on-demand route acquisition system; Nodes
does not give false data to active paths either share or exchange information about the
routing table nor keep this information. Additional, a node does not have to maintain
and discover a route to another node until the communicate between two nodes needed,
the fig-1- below shown the routing discover process unless the former node which is an
intermediate routing station provides its services to maintain communication between
two other nodes[5].
454
(ICoIT'17)
10thof April, 2017.
FIGURE 1. Route discovery process of AODV
The importance connectivity of the local mobile node makes each mobile node uses
multiple techniques to become more aware of neighboring nodes, one of these technique
includes local broadcasts known as hello messages (not system-wide). The routing tables
are designed to optimize response time of the nodes within the neighborhood to local
movements, and due to the requests of new routes, this technique provides fast response
time. The primary objectives of the algorithm are:
Discovery packets were broadcasting when it needed only.
To differentiate between management of local connectivity and maintenance of
general topology.
To distribute information to mobile nodes that needs the information about
changes in local connectivity.
2.2. MEHODOLOGY OF SECURING MULTIPATH ROUTING

A protocol is responsible for the protection which implemented by using a digital
signature, this digital signature generated by applying the encryption algorithm and MD5
hash function [2]. The correctness of data, nonrepudiation, and authentication is security
ensures. The protocol supports Byzantine attacks; sink hole, selective forwarding and
data tampered or altered routing. Ad hoc On-demand Multipath Distance Vector
(AOMDV) routing protocol is regularly applied in Wireless Sensor Network [6].
In AOMDV, the network is inactive while waiting for a fresh connection is required.
When any node requires communicating with others, it transmits a request for
connection. Another AOMDV nodes for a second time transmits this message, then keep
the nodes log that they received from; each node remains this process till target node get
[7].
455
(ICoIT'17)
10thof April, 2017.
FIGURE 2. Flow Diagram of Multipath Routing

The figure 2 has shown the process flow diagram of the algorithm [8]. Nodes transmit
messages back over a momentary route to the requesting nodes which are desire to send
message to the wanted node after receiving this message. The node chooses the shortest
route for communication that which begins the process (i.e. just the route that having the
minimum number of hop count). Determined the behavior plus characteristic of the
attack from several parameters. Broadcasting messages are to maintain the capacity of
the wireless network. A rapid substantiation is complete to assess and confirm the
dubious attack [9].
2.3. A Novel Method to Improve Life of Network for Secure Routing Protocol
The procedure for of working start in login to account (for new user create an account)
after this authenticate it by username, password and secret question to authorize. Then
generate path step and choosing file /data to encrypt file/data by using Advanced
456
(ICoIT'17)
10thof April, 2017.
Encryption Standard (AES) algorithm and send it to the desired node through selected
path. At the destination node, the user verifies whether the received data for reliable/pure
or malicious if data after verification goes to reliable the data is pure otherwise, data is
not pure (The intruder is attacking data). Recognize the attack data and reject it and
finally negative acknowledgment sent to the source [10]. The following pseudo code of
the algorithm:
1. Start and login process
2. if register then login and authenticate and authorized
3. Else create a count and repeat step 2.
4. generate secure path.
5. send file (encrypted file) at source and receive file (decrypt
file) at destination
6. verifies data (if pure data ) its save it
7. Else identify the node that send data by vampire attack.
8. reject data and send negative acknowledgement to the source.
2.4. A Simulation Study of Security Performance

The SPREAD scheme reduce the message interruption which is the result of
eavesdropping or compromising nodes by combines a multipath routing and secret
sharing [11]. The secret message is dividing to several (N) parts (called shares). This
dividing is by utilizing a threshold secret sharing scheme (T, N), it can simply retrieve
the message so that from every T or more shares. It is impossible to retrieve the message
while from any shares (T-1) or less. By applying link encryption for this scenario where
SPREAD is implemented. To negotiate between the two neighboring nodes, each link
uses a dissimilar encryption key (.g. using th Diffi-Hllman ky exchange protocol).
For composition the messages, the opponent can either overhear the sending and
receiving of all network then trying brute-force decryption or compromise the nodes
these interrupt all the secrets transmitted [11].
457
(ICoIT'17)
10thof April, 2017.
FIGURE 3. Illustration of multipath routing
From network viewpoint, an opponent may interrupt all the essential information, if all
message follows a single route to its destination, to retrieve that message with
compromising any node on the route [11]. With the SPREAD scheme, excepting the
destination and source nodes, the opponent should conciliation some nodes inside
network on a number of independent routs to find the least required (T) shares.
Considering the multipath routing and the scenario of (3, 3) secret sharing as shown in
Figure3.
2.5. Dynamic Source Routing (DSR)

DSR is an on-demand protocol designed to limit the bandwidth occupied by control
packets in ad hoc wireless networks by dropping the periodic table update messages
which required in the proactive routing protocols. In the network, to connect node with
another node, initially, need to discover a proper path in order to follow while
transmitting packets to the destination node. While the requirements stay without change,
this path should then continue to work for as long as it is needed. For the benefit of the
intermediate nodes, in this protocol every potential information extracted from the source
path and stored within route cache that contained in a data packet. DSR also allows
iggy-backing of a data packet [12].
458
(ICoIT'17)
10thof April, 2017.
FIGURE 4. Route establishment in DSR.
In the Maintenance procedure when an intermediate node in the network is moving wy

or drop making a break in the wireless link, an error message is created from the node
adjacent to the broken link to inform the source node in a whole network. The source
node reinitiates establishment procedure for the route. When a Route Error packet is
received the source node is removed and the cached entries at the intermediate nodes.
Figure 4 illustrates the working of DSR [12].
3. Comparison Between The Algorithms

The following shown the comparison in table 1 depended on similar type routing
protocol, routing protocol, advantage and disadvantages of the above algorithms
[1][2][3][4][13][14][15][16][17][18].
TABLE 1: comparison types of routing protocol
Type Techniques
Routing Disadvantag
Name Routing advantage
protocols e
protocol
capable of Gets single path
both unicast from the available
funded path, and reduce the increase
and
the source sends all control traffic latency in
AODV REACTIVE multicast
its packet through message finding new
routing with
this one funded overhead. routes.
loop free
path.
routes
459
(ICoIT'17)
10thof April, 2017.
Gets multipath allows the

from the route
intermediate
multicast funded path increased
AOMDV REACTIVE nodes to
routing between each flooding.
source and
reply to
destination RREQs.
Each node starts as exploit
its own group of not depend
simple
size one, with a on any
Novel broadcastin properties of
REACTIVE virtual address 0. protocol,
Approach g protocol like
Nodes broadcasts design
form groups with distance
properties.
their neighbors vector.
combines Divide the secret aggregating
the secret message into N bandwidth, Severe
SPREAD REACTIVE sharing and parts that used to reducing collisions at
multipath find multi secure blocking MAC layer.
routing path. probability.
Divided into two the link is
types routing broken or a
Route
discover and change is
DSR PROACTIVE broadcast maintenance
routing noticed
mechanism
maintenance another path
is chosen
4. METHODOLOGY OF PROPOSED STUDY

The mechanism of an efficient security is established to protect the communication inside
the network (connection between nodes) and to prevent attacks of Black hole and Gray
hole by using DSR. In this mechanism, when several nodes configured the network is
from this network are examined and checks whether there are malicious nodes to remove
by using the use of advanced DSR protocol mechanism.
Then a spanning tree creates to calculate the least distance between every/each node
which can cover network without making a cycle. Then we select the route with
minimum distance. For security and compression we use the ciphers based on chaotic
systems and arithmetic coding, which uses the secret information (the key which Based
on the initial condition/parameters for the chaotic map). The system proposal is shown
in Figure 5.
460
(ICoIT'17)
10thof April, 2017.
4.1. A-DSR ROUTING PROTOCOL

In the development method, the stirrer is transmit a Route Request packet, recognizing
the destination to which the route is required. In general, any node upon receiving the
Route Request, in case that if it has not already forwarded a copy of the Route Request
it will retransmits the request; when the request received by target node the, a Route
Reply backward to the stirrer, rather than forwarding the request, listing the route taken
by the Request. Route reply for each copy of the route request that it receives will return
by target node; Each Route Request packet carries a sequence number generated by the
source node and the path it has traversed. A node upon receiving the Route Request
packet, checks the sequence number on the packet before forwarding it. The sequence
number on the packet is used to prevent loop formations and to avoid multiple
transmissions of the same Route Request packet by an intermediate node that receives it
through multiple paths [12].
Construction of network model
Discover malicious
Malicious node
nodes during route
detection
protocol process using
DSR
Minimum spanning tree

construction
chaos-based adaptive arithmetic

coding
Encryption process
FIGURE 5. Proposed System Architecture
4.2. SPANNING TREE CONSTRUCTION

The minimum distance of spanning tree is calculating between every / each nodes which
can cover all the nodes without forming a cycle. Spanning tree holds security tie-ups only
with neighbor. Spanning Tree Protocol is a network protocol which maintains plus
461
(ICoIT'17)
10thof April, 2017.
establishes this network by connecting a group of portable node in the wireless ad hoc
network. the process of calculating minimum span tree shown in Figure 6 [14].
FIGURE 6. Construction of spanning tree

4.3. SECURITY AND COMPRESSION MECHANISM
Chaotic and arithmetic coding mechanism is used to ensure security. Each and every
node from the mobile ad-hoc network has its own system. The model that execute zeroth-
order adaptive arithmetic coding. Figure 7 shows that the encoder block that uses the
secured information that depends on the modified arithmetic coder (this key is based on
the first condition/parameters for the chaotic map) this secured information is to coding
the initial file for the plaintext, so, the output will be encrypted and compressed. The
insecure channel that transmit the data that are coded to the encoder block will be secured
by the information that resulted from the encoder block, while the key for the chaotic
map will transferred by a private, secure channel. Figure 7 shows the process for encode,
encrypt and decode and decrypt message from sender to receiver.
Modified Modified Decoded

Insecure
Input Arithmet Arithmetic plaintext
channel
plaintext decoding output
ic coding
Decoder
Encoder
Seed for Secure Block
Block Seed for
logistic channel
logistic
function function
FIGURE 7. Chaotic and arithmetic coding

5. CONCLUSION
system
Mobile Ad hoc Network requires the high level of security as compare to the regular wired
networks Security issues which is neglected while designing routing protocols for ad-hoc
networks. Through DSR protocol, it is easier to infract the security of wireless ad-hoc
network. DSR protocol is susceptible to various attacks including Black hole and Gray hole
462
(ICoIT'17)
10thof April, 2017.
attacks. Efficiently finds short and secure route to the destination. A secure ad hoc network
has to meet different security requirements. First privacy concerns require Confidentiality in
the use of data. Second Encryption and Decryption by using chaotic cryptosystem is one of
the mechanisms used to enforce confidentiality. Third the integrity of data (cipher text) arise
to destination must not be modified by maliciously only the authorized user can be decrypted
and recover the original data. Fourth the data compression is also an issue for several reasons
in Authentication and Non-repudiation refer to the inability to deny the performance of some
action falsely. The DSR scheme is depends on the idea to distribute a secret through multiple
independent paths while it is transmitted across the network.
6. REFERENCES
[1] Dongbin Wang, Mingzeng Hu and Hui Zhi, A survey of secure routing in ad hoc
networks, The Ninth International Conference on Web-Age Information
Management, 2008 IEEE.
[2] S. Marti, T. J. Giuli, K. Lai and M. Baker, "Mitigating Routing Misbehaviour in
Mobile Ad Hoc Networks", Mobile Computing and Networking. (2000), pp., 255-
265.
[3] Yipin Sun, Rongxing Lu, Xiaodong Lin, Jinshu Su and Xuemin (Sherman) Shen,
NEHCM: A Novel and Efficient Hash-chain based Certificate Management Scheme
for Vehicular Communications.
[4] F. Bevitelli, E. Di Cola, L. Fortuna and F. Itnlia, Multilayer Chaotic Encryption for
Secure Communications in Packet Switching Networks , Communication
Technology Proceedings, 2000. WCC - ICCT 2000. International Conference on 21-
25 Aug. 2000
[5] CHARLES E. PERKINS, ELIZABETH M. ROYER, AD-HOC ON-DEMAND DISTANCE
VECTOR ROUTING, MOBILE COMPUTING SYSTEMS AND APPLICATIONS, 1999.
PROCEEDINGS. WMCSA '99. SECOND IEEE WORKSHOP.
[6] Parul suneja , Anil Kumar and Annu Soni , System scenario based investigation of
AODV and AOMDV Routing Protocol in MANET , 2015 International Conference
on Soft Computing Techniques and Implementations- (ICSCTI).
[7] Sunita Gupta and Ghanshyam Prasad Enhanced load balancing and delay
constraint AOMDV routing in MANET, 2016 Symposium on Colossal Data
Analysis and Networking (CDAN).
[8] Sukiswo and Muhamad Rifqi Rifquddin ,Performance of AOMDV Routing Protocol
Under Rushing and Flooding Attacks in MANET, 2015 2nd International
Conference on Information Technology, Computer, and Electrical Engineering
(ICITACEE)
[9] Vipin Bondreand Sanjay Dorle Design and Performance Evaluation of AOMDV
Routing Protocol for VANET , IEEE International Conference on Computer,
Communication and Control (IC4-2015).
[10 ] Sunil Bhutada, ranthi Kumar.K,Manisha.K, A Novel Approach for Secure Routing
Protocol: To Improve Life of Network, International Conference on Contemporary
Computing and Informatics (IC3I), 978-1-4799-6629-5/14 2014 IEEE.
463
(ICoIT'17)
10thof April, 2017.
[11] Wenjig Lou,Wei Liu,Yuguang Fang, A Simulation Study of Security Performance

Using Multipath Routing in Ad Hoc Networks, 0-7803-7954-3/03 2003 IEEE.
[12] Loay Abusalah,Ashfaq Khokhar,Mohsen Guizani, A Survey of Secure Mobile Ad
Hoc Routing Protocols, IEEE COMMUNICATIONS SURVEYS & TUTORIALS,
VOL. 10, NO. 4, FOURTH QUARTER 2008.
[13] Rabab Mohsin, John Woods, Performance Evaluation of MANET Routing
Protocols in a Maritime Environment 6th Computer Science and Electronic
Engineering Conference (CEEC) university of Essex ,2014.
[14] Sisily Sibichen1, Sreela Sreedhar2 ,"An Efficient AODV Protocol and Encryption
Mechanism for Security Issues in Adhoc Networks", (ICMiCR-2013).
[15] Mohsin Ur Rahman and Sohail Abbas Simulation-based analysis of MANET routing
protocols using group mobility model ,International Conference on Inventive
Computation Technologies (ICICT),2016.
[16] Faria Farjana Khan , Anindya Jana , Tahesin Samira and Kazy Noore Alam Siddiquee,
Performance of Agro-Sensors: Assessment of Optimality in Routing Protocols of
MANET in Wireless Sensor Networks, International Conference on Intelligent
Control Power and Instrumentation (ICICPI) , 2016.
[17] Yue Yang , Siyuan Hao and Haibin Cai , Comparison and Evaluation of Routing
Protocols Based On A Collaborative Simulation Using SUMO and NS3 with TraCI,
International Conference on Information System and Artificial Intelligence , 2016.
[18] Guruprasanna and R Sujatha M A novel Approach to avoid malicious attack to
enhance network in WSN, IEEE International Conference On Recent Trends In
Electronics Information Communication Technology, 2016.
464
(ICoIT'17)
10thof April, 2017.
The Estimation of Wind Velocity Using Data Mining

Techniques
Sattar Nabee Rasool
Software Engineering, Technology Faculty, Firat University-Turkey
star81sat@gmail.com
Ahmet Koca
Mechatronics Engineering, Technology Faculty, Firat University-Turkey
akoca@firat.edu.tr
Karwan Hussein Qader
School of Computing, Faculty of Technology, University of Portsmouth- United Kingdom
Karwan.qader@port.ac.uk

Estimation of wind velocity in real time is very essential as it
Article History: can provide valuable information to people of different
Received: 10 March 2017 domains such as agriculture, aviation and tourism to mention
few. Since climate data is growing exponentially it is hard to
Accepted: 1 April 2017 analyze it manually. Therefore, machine-learning techniques
Published: 10 April 2017 such as unsupervised and supervised learning methods are
used to mine voluminous data and discover valuable
Keywords: Climate knowledge. Predictive modeling in data mining is required to
parameter estimation, estimate climate parameters. In this paper, we proposed a
data mining, predictive framework that exploits data mining techniques such as J48,
modeling, clustering, KNN, Neural Networks, SVM and Linear Regression. The
classification framework takes climate dataset as input, completes training
phase and makes different models using data mining
algorithms. Finally, it ends by exploiting linear regression,
which models the relationship between a dependent variable
and an exploratory variable. The framework results in
estimating wind velocity and finding prediction error rate. A
prototype application is built based on Weka, which is used
to demonstrate proof of the concept. The empirical results
applied on all data are obtained from the Turkish Government
Meteorology Services for summer months of 2013. It
revealed that the proposed framework is useful to have a
predictive model with respect to estimation of climate
parameters.
465
(ICoIT'17)
10thof April, 2017.
INTRODUCTION
D ata mining techniques became indispensable in the contemporary business world

where enterprises need to analyze huge amount of data. Data mining provides
both predictive and descriptive models. The predictive models include
techniques pertaining to classification, regression, time series analysis and
prediction. Descriptive models include clustering, summarization, association rules, and
sequence discovery. Clustering is an unsupervised learning method used to group similar
objects while classification is known as supervised learning method, which needs training
samples for prediction of class labels for some test objects. Similarity measures like
Euclidean distance are widely used to know similarity between two objects in the given
dataset. Predictive methods are used in many real time applications. For instance weather
forecasting, forecasting of stock values, and prediction of possible increase or decrease of
values of certain items to mention few.
In this paper different data mining algorithms are analyzed, trained, and models are created
to have utility in wind velocity of given dataset. After experimenting with different models,
five different datamining tools is used in order to predict wind velocity. An application is
built based on Weka tool in order to demonstrate proof of the concept. The remainder of the
paper is structured as follows. Section 2 reviews related literature. Section 3 presents the
proposed framework for data mining. Section 4 presents experimental results while section
5 presents conclusions and recommendations for future work.
RELATED WORKS
Aziz and Yosof [1] proposed a classification model to analyze graduates employment. They
employed different algorithms like J48, k-Nearest Neighbor (KNN), Multilayer perception,
Logistic Regression and Naive Bayes. They found that logistic regression resulted in highest
accuracy 95.2%. Balle and Lisboa [2] focused on machine learning methods and models of
data mining. They include monograms, rule induction, graphical models, and data
visualization. Minh, Hue, Dzung, and Toan [3] made a simulation study of equity valuation
model. Langone et al. [4] proposed a machine learning method known as Least Squares
Support Vector Machine (LS-SVM) for fault detection online with industrial machine data.
It is supervised learning method, which helped to know early detection and classification of
faults.
Claesen et al. [5] proposed a supervised learning model using SVM. The model makes use
of positives and unlabeled objects. It actually makes use of an ensemble of SVM models in
order to have an intuitive approach for appropriate learning model. It has features like
supervised learning, positive and unlabeled learning, and learning with false positives. Frandi
et al. [6] propose another SVM classification approach along with an algorithm for large-
scale benchmark data classification. Alaiz et al. [7] proposed a graph based classification
method for robust classification of data presented in the form of graphs. Chae et al. [8]
proposed a data mining approach known as feature selection for intrusion detection using
NSL-KDD dataset. They used the concept of information gain and gain ratio. Chen et al. [9]
emphasized the need for considering big data and big data eco-system for many real time
enterprise applications.
466
(ICoIT'17)
10thof April, 2017.
Huang et al. [10] proposed an SVM based classifier known as pin-SVM for finding pinball
loss in SVM. Mehrkanoon et al. [11] proposed a semi-supervised algorithm that is based on
kernel spectral clustering. Kernel here holds prior knowledge about labels. Drumetz et al.
[12] focused on dimensionality estimation algorithms for finding intrinsic dimensionality in
datasets.
Olaiya and Adeyemo [13] investigated the utility of data mining techniques for predicting
weather information like wind speed, evaporation, rainfall and maximum temperature. They
used two algorithms namely decision tree and artificial neural network for achieving this.
They found the usefulness of data mining techniques for weather forecasting and study of
climate changes. Krishna [14] reviewed many weather forecasting models available. They
include models based on classification techniques, clustering techniques, and artificial neural
networks. Schmidhuber [15] made a review of data mining techniques that are used for deep
supervised and unsupervised learning. Especially they focused on neural networks. Cortest
et al. [16] proposed a framework known as AdaNet for analyzing artificial neural networks.
The framework followed adaptive structural learning. Siddiqui and Muhammad [17]
performed clustering based data mining operations on KDD Cup 99 dataset. They used k-
Means clustering algorithm provided by Oracle Data Mining tool.
Schubert et al. [18] proposed a methodology for clustering uncertain data. Uncertain data is
the data which has some noise and the data is not precise. They used a general framework to
cater to the needs of analysis of uncertain data. They named the framework as ELKI which
has many implementations of algorithms that made use of distance measures, pruning
techniques, visualization components and evaluation measures. Agrawal and Gupta [19]
explored usage of C4.5 algorithm with certain optimization for healthcare domain with
additional rules. They studied the difference between classification functionality of original
C4.5 and an optimized C4.5 in terms of efficiency and complexity. Denoeux et al. [20]
explored a rule known as evidential K-nearest neighbor rule for making a clustering
procedure. Their procedure is named as EK-NNclus, which takes scale parameter and k of
neighbors. In addition, it does not need the number of clusters to be generated. It was able to
get both model-based and density-based procedures with having knowledge of number of
clusters priori. Sanghavi et al. [21] explored three kinds of algorithms related to feature
selection. They are known as Filter, Wrapper and Embedded. They built logistic regression
algorithm in order to identify disease with medical data mining. In this paper a framework is
proposed to explore data mining algorithms and produce a prediction model ultimately using
linear regression technique. The proposed model shows that such climate can be predicted
using the data mining techniques considering the error rate calculations. Therefore, this work
filled those gaps, which it has been conducted to gather fifth different tools into a model
includes cluster, classification and regression techniques. The proposed model shows the
comparison evaluation results that obtain by the model execution. It is simply give clear
results of all used techniques and guide the user, which one will be the most suitable tool
according to the nature of work.
467
(ICoIT'17)
10thof April, 2017.
PROPOSED METHODOLOGY
In this paper, we proposed a methodology for the exploration of data mining algorithms for
prediction of wind velocity parameter of climate dataset. In other words, the methodology
throws light into building various models using different data mining techniques. The
methodology takes climate dataset as input and performs training phase, so as to get training
data ready for exploration of different classification algorithms. The data mining algorithms
used in the exploration of the models include kNN, Neural Networks, J48, SVM, and linear
regression. The methodology is illustrated in Figure 1.
FIGURE 1. Proposed methodology

KNN is meant for pattern recognition method meant for regression and classification. Neural
networks are the networks used in computer science that model human brain and nervous
system. It is a computational approach, which is based on collection of artificial neurons that
mimic how biological neurons work. J48 is a decision tree algorithm based on ID3. SVM is
a supervised learning model which can have associated learning algorithms for both
regression and classification tasks.
Linear regression is a data mining approach to modeling the relationship between a
dependent variable and one or more independent variables. Simple linear regression makes
use of one independent variable. A linear model is built using linear regression approach
which models relationship between model parameters linear predictor functions.
In this paper, the kNN, J48, SVM, neural networks and linear regression model are used as
a prediction models. The individual methods are employed to have wind velocity prediction.
Then the results pertaining to actual wind velocity and predicted wind velocity are presented
468
(ICoIT'17)
10thof April, 2017.
for all algorithms in the next section. Error rate is computed for all algorithms to know
prediction accuracy of different algorithms.
EXPERIMENTAL RESULTS
A Dataset that contains weather data with attributes is taken as a case study for this research
work. First column of the file is station number, every station represents a Turkish city which
is five stations. These cities chosen by their geographical position and include INEBOLU,
SINOP, AMASRA, SILE and AKCAKOCA. Every station place is near the sea. So that their
wind potential is higher than other places. All data are obtained from the Turkish
Government Meteorology Services for summer months of 2013.
The wind potential of a place is related with geographical position, humidity, atmospheric
pressure and the ambient temperature. All of these are datasets parameters. And it has 456
instances covering data of five stations for summer months. An excerpt of the dataset is
shown in Table 1. Weka is the environment used to explore a climate dataset that is very
useful and the mining of such data provides useful information to various domains in the
world. Weka tool is used to exploit all existing methods. Apart from Weka, the environment
includes JDK 1.7 (Java programming language), Net Beans 7.2 (Integrated Development
Environment) are used in Windows platform. JDK is used to have basic tools like compiler,
application launcher and Java language support. Net Beans is used to have rapid application
development as it supports developing applications faster. Weka is integrated with Net Beans
in order to reuse existing methods and other support functionality like loading files, viewing
data and results. Swing API in JDK is used to have Graphical User Interface (GUI) while
java.io package is used to work with input and output files. Different kinds of data mining
techniques as presented in Figure 1 are used to perform mining on given dataset. The dataset
is related to climate data that were captured in a one of the region in Turkey within first
fifteenth days of the June. The attribute of pressure, Humidity and temperature and actual
wind speed resented in the datasets. To predict the wind velocity using all datamining
techniques the data inputted into the proposed model. In supervised simulation the predefined
classes achieved through a training phases of the existing datasets. In NN phase, the
backward propagation used based on excerpt of dataset is presented in Table 1 (except the
last column which is the predicted wind velocity using Linear Regression), which is used as
input to the proposed framework.
As shown in Table 1, it is evident that the input dataset contains different attributes like
station number month, day, and pressure, humidity, and temperature. The proposed
framework is applied to have training samples and different models. All these experiments
are done using Weka, which is one of the widely used data mining tools.
TABLE 1: An excerpt from generated output of Regression
Station Pressure Humidity Temperature
Month Day Wind Velocity (m/s)
Number (hPa) (%) (oC)
1 6 1 1008.7 67.7 19.7 19.52310330013237
1 6 2 1005 80.3 19.6 19.60674819378439
1 6 3 1010.1 69.3 17.3 19.69039308743642
469
(ICoIT'17)
10thof April, 2017.
1 6 4 1011.1 72 19.4 19.77403798108844

1 6 5 1013.4 82.7 17.4 19.85768287474047
1 6 6 1014.6 80.7 18.8 19.941327768392494
1 6 7 1013.6 72 19.5 20.02497266204452
1 6 8 1010.1 77.7 20.4 20.108617555696544
1 6 9 1012.3 81.3 19.2 20.19226244934857
1 6 10 1012.9 77.7 19.2 20.275907343000593
1 6 11 1008.4 75.7 20.6 20.35955223665262
1 6 12 1005.3 80.3 21.2 20.443197130304643
1 6 13 1006.5 80.3 21.9 20.526842023956668
1 6 14 1009.8 88.7 20 20.610486917608696
1 6 15 1010.7 85.3 20.8 20.694131811260718
As shown in Table 1, the outcome of the proposed methodology is presented in the last
column. It is wind velocity, which is predicted using linear regression model. This kind of
prediction models has impact on many applications in the real world such as weather
forecasting, prediction of dependent variables for making well-informed decisions. Same
scenario applied on the datasets based on all other techniques and then Error rate is computed
for all by comparing predicted wind velocity and actual wind velocity.
TABLE 2: Error rate for Regression method
Actual
Station Predicated Wind
Month Day Wind Error Rate
Number Velocity (m/s)
Velocity
1 6 1 17.4 19.52310330013237 12.2017
1 6 2 15.2 19.60674819378439 28.9918
1 6 3 15.5 19.69039308743642 27.0348
1 6 4 15.5 19.77403798108844 27.5744
1 6 5 15.8 19.85768287474047 25.6815
1 6 6 18.2 19.941327768392494 9.56773
1 6 7 16.6 20.02497266204452 20.6324
1 6 8 15.5 20.108617555696544 29.733
470
(ICoIT'17)
10thof April, 2017.
1 6 9 19.4 20.19226244934857 4.08383
1 6 10 19.1 20.275907343000593 6.15658
1 6 11 18.4 20.35955223665262 10.6497
1 6 12 17.2 20.443197130304643 18.8558
1 6 13 15.9 20.526842023956668 29.0996
1 6 14 16.2 20.610486917608696 27.2252
1 6 15 18.4 20.694131811260718 12.4681
As presented in Table 2, the results show the actual wind velocity and predicted wind velocity.
Error rate is computed as follows.
Error Rate= (Predicted Wind Velocity-Actual Wind Velocity)*100/Predicted Wind Velocity (or)
Error Rate= (Actual Wind Velocity - Predicted Wind Velocity)*100/Predicted Wind Velocity
Error rate is presented in the last column of the Table 2, which reflects the difference between
actual wind velocity and predicted wind velocity. In the same fashion, the results of other
algorithms are presented as follows.
TABLE 3: Error rate for kNN
Actual
Velocity
1 6 1 17.4 17.566 0.95402
1 6 2 15.2 17.725 16.6118
1 6 3 15.5 18.112 16.8516
1 6 4 15.5 18.345 18.3548
1 6 5 15.8 18.365 16.2342
471
(ICoIT'17)
10thof April, 2017.
1 6 6 18.2 18.75 3.02198
1 6 7 16.6 19.12 15.1807
1 6 8 15.5 17.56 13.2903
1 6 9 19.4 18.25 5.9278
1 6 10 19.1 18.28 4.2931
1 6 11 18.4 18.56 0.86957
1 6 12 17.2 18.85 9.59302
1 6 13 15.9 18.88 18.7421
1 6 14 16.2 18.54 14.4444
1 6 15 18.4 18.56 0.86957
TABLE 4: Error rate for J48
Actual
Velocity
1 6 1 17.4 15.466 11.11494
1 6 2 15.2 15.723 3.44079
1 6 3 15.5 16.12 4
1 6 4 15.5 16.35 5.48387
1 6 5 15.8 16.58 4.93671
1 6 6 18.2 15.95 12.36264
1 6 7 16.6 17.26 3.9759
1 6 8 15.5 16.58 6.96774
472
(ICoIT'17)
10thof April, 2017.
1 6 9 19.4 16.95 12.62887

1 6 10 19.1 17.56 8.062827
1 6 11 18.4 17.85 2.98913
1 6 12 17.2 17.86 3.83721
1 6 13 15.9 18.12 13.9623
1 6 14 16.2 18.2 12.3457
1 6 15 18.4 18.15 1.358696
TABLE 5: Error rate for SVM
Actual
Velocity
1 6 1 17.4 13.46 22.64368
1 6 2 15.2 13.72 9.736842
1 6 3 15.5 14.25 8.064516
1 6 4 15.5 14.35 7.419355
1 6 5 15.8 14.58 7.721519
1 6 6 18.2 14.35 21.15385
1 6 7 16.6 15.25 8.13253
1 6 8 15.5 14.68 5.290323
1 6 9 19.4 15.28 21.23711
1 6 10 19.1 15.45 19.10995
1 6 11 18.4 15.85 13.8587
1 6 12 17.2 16.12 6.27907
1 6 13 15.9 16.05 0.9434
1 6 14 16.2 16.35 0.92593
1 6 15 18.4 16.15 12.22826
473
(ICoIT'17)
10thof April, 2017.
TABLE 6: Error rate for Neural Networks (BPNN)
Actual
Velocity
1 6 1 17.4 13.56 22.06897
1 6 2 15.2 12.78 15.92105
1 6 3 15.5 12.25 20.96774
1 6 4 15.5 12.35 20.32258
1 6 5 15.8 12.58 20.37975
1 6 6 18.2 14.12 22.41758
1 6 7 16.6 13.54 18.43373
1 6 8 15.5 13.24 14.58065
1 6 9 19.4 15.86 18.24742
1 6 10 19.1 14.75 22.77487
1 6 11 18.4 14.76 19.78261
1 6 12 17.2 13.89 19.24419
1 6 13 15.9 13.95 12.26415
1 6 14 16.2 13.98 13.7037
1 6 15 18.4 14.86 19.23913
As shown in Table 3 through Table 6, the results of different algorithms such as kNN, J48,
SVM and neural networks are presented along with the error rate in wind velocity prediction.
TABLE 7: Error rate for all methods
Average Error
Method Name
Rate
KNN 10.349264
J48 7.1644882
SVM 10.98300233
Neural
18.68987467
Networks
Linear
19.33040933
Regression
In addition, Table 7 shows the average error rate of all instances for each method is
computed. The results revealed the performance of different methods for predicting wind
474
(ICoIT'17)
10thof April, 2017.
velocity. The results stated that J48 obtained the lowest error rate while Linear Regression
got the biggest average of the error rate. Likewise, the error rate of other used techniques are
shown in the table 7.
CONCLUSIONS AND FUTURE WORK

This paper focuses on data mining techniques used to build different modes using training
samples. The paper explored different data mining techniques using Weka environment. The
techniques employed include J48, kNN, Neural Networks, SVM and Linear Regression.
Climate dataset is considered as input. A framework is proposed to have data mining applied
on the given dataset. Different algorithms aforementioned are applied for making different
models. The framework functionality ends with application of linear regression model, which
predicts wind velocity. The dependent variable considered as wind velocity, which is empty
in the given dataset. A prototype application is built to demonstrate the proof of the concept.
The experimental results revealed the utility of the proposed framework in predicting wind
velocity. The error rate of Linear Regression is between 4.08383 and 29.0996. The error rate
of KNN is between 0.86 957 to 18.7421. The error rate of J48 is between 1.358696 and
13.9623. The Error rate of SVM is between 0.92593 and 22.64368 while the error rate of
neural networks is between 12.26415 and 22.77487. J48 shows low average error rate. In
future, this research can be extended to ensemble methods to have more quality predictions
for datasets from different domains.
REFERENCES
[1] Mohd Tajul Rizal Ab Aziz and Yuhanis Yusof , Graduates Employment Classification
using Data Mining Approach", ICAST, pp. 1-8, 2016.
[2] Vanya Van Belle and Paulo Lisboa. Research directions in interpretable machine
learning models. ESANN, pp. 533-541, 2013.
[3] Nguyen Dang Minh and Nguyen Thi Minh Hue , Equity valuation model of vietnamese
companies in a foreign securities market a simulation approach, IEEE, vol. 66, pp. 1155-
11, 2012.
[4] RoccoLangone , CarlosAlzate , BartDeKetelaere , JonasVlasselaer , WannesMeert and
JohaA.K.Suykens,LS-SVMbasedspectralclusteringandregressionforpredicting
maintenanceofindustrialmachines, Elsevier Ltd, vol. 37 , pp. 268-278, 2014.
[5] Marc Claesen, Frank De Smet, Johan A. K. Suykens and Bart De Moor, Arobust
ensemble approach to learn from positive and unlabeled data using svm base
models. neurocomputing, SI on Advances in Learning with Label Noise, pp. 1-34, 2014.
[6] Emanuele Frandi, Ricardo anculef and Johan A. K. Suykens, A partan-accelerated
frank-wolfe algorithm for large-scale svm classification, pp. 1-19, 2015.
[7] Carlos M. Alaz, Michael Fanuel and Johan A. K. Suykens. Robust classification of
graph-based data , pp. 1-8, 2016.
[8] Hee-su Chae, Byung-oh Jo, Sang-Hyun Choi and Twae-kyung Park, Feature selection
for intrusion detection using nsl-kdd, isbn, pp. 184-187, 2013.
475
(ICoIT'17)
10thof April, 2017.
[9] Min Chen , Shiwen Mao and Yunhao Liu. big data, A Survey.Springer, vol. 19 , pp.
171-209, 2014.
[10] Xiaolin Huang, Lei Shi, and Johan A. K. Suykens. Solution path for pin-svm classifiers
with positive and negative values, pp. 1-10, 2012.
[11] Siamak Mehrkanoon and Johan A. K. Suykens. Non-parallel semi-supervised
classification based on kernel spectral clustering, pp. 1-8, 2014.
[12] Lucas Drumetz, Miguel Angel Veganzones, Ruben Marrero, Guillaume Tochon, Mauro
Dalla Mura, Giorgio Licciardi, Christian Jutten and Jocelyn Chanussot, Hyperspectral
Local intrinsic dimensionality, HAl, pp. 1-16. 2016.
[13] Folorunsho Olaiya and Adesesan Barnabas Adeyemo . (2012). Application of Data
Mining Techniques in Weather Prediction and Climate Change Studies . Information
Engineering and Electronic Business, p51-59.
[14] G.Vamsi Krishna (2015) A Review of Weather Forecasting Models-Based on Data
Mining and Artificial Neural Networks 06 , p214-22.
[15] Jurgen Schmidhuber,Deep learning in neural networks: an overview, pp. 1-88, 2014.
[16] Corinna Cortes, Xavi Gonzalvo and Vitaly Kuznetsov, Adanet: adaptive structural
learning of artificial neural networks, pp. 1-20, 2016.
[17] Mohammad Khubeb Siddiqui and Shams Naahid, Analysis of kdd cup 99 dataset
using clustering based data mining, International Journal of Database Theory and
Application, vol. 6, (1-5), pp. 23-34, 2013.
[18] Erich Schubert, Alexander Koos and Tobias Emrich , A framework for clustering
uncertain data , pp. 1976-1979, 2012.
[19] Gaurav L. Agrawal and Hitesh Gupta, Optimization of c4.5 decision tree algorithm
for data mining application International Journal of Emerging Technology and Advanced
Engineering, vol. 3, (1-3), pp. 341-345, 2013.
[20] Thierry Denoeux, Orakanya Kanjanatarakul and Songsak Sriboonchitta, Ek-nnclus:
A clustering procedure based on the evidential k-nearest neighbor rule, pp. 1-39, 2016.
[21] Dhaval Sanghavi, Hitarth Patel and Sindhu Nair, Logistic regression in data mining
and its application in identification ofdisease, International Journal of Current
Engineering and Technology , vol.4, (1-6), pp. 3837-3839, 2014.
476
(ICoIT'17)
10thof April, 2017.
Unbreakable Digital Watermarking Technique

Abdulkareem O. Ibadi
Computer Engineering Department, Baghdad university college for economic
sciences, Baghdad - Iraq

In this research, a new concept of watermarking method is
Article History: introduced; it is suitable for any type of information file. It
Received: 8 March 2017 doesn't make any change on the watermarked file. The
concept is to store secure watermarking file (SWF) in a
reliable third party (RTP). SWF is generated by locating the
Published: 10 April 2017 logo file in the watermarked file. SWF is formatted in special
compressed form and can be compressed and encrypted using
Keywords: any suitable application.
Digital Watermarking,
Image Watermarking
1. Introduction
T
he threats of copyright are grown up on the owner's mind, so they always try to
secure their data from these threats by merging the owner ID with digital media.
This process is called Watermarking; the proof of ownership is done by extracting
the owner ID from the digital information file.
The protection of ownership of digital information becomes important concerns with the ease
of editing and reproduction. Digital watermarking, a scheme to embed special labels in
digital sources, has made considerable progress in recent years. There are several categories
of watermarking schemes [1]. Watermarking can be robust or fragile. In robust watermarking
technique, the modification to the watermarked content will not affect the watermark. But
fragile watermarking is a technique in which watermark gets destroyed when watermarked
content is modified or tampered with [2].
Among them, fragile watermarking is a technique to insert a signature for image

authentication. The signature will be altered when the host image is manipulated.
An effective authentication scheme should have desirable features that the embedded
authentication data is to be invisible under normal viewing conditions; and to allow the
watermarked image be stored in lossy compression format.
477
(ICoIT'17)
10thof April, 2017.
In other hand, the patent must be registered to save the owner rights. The registration itself
means a patent protection. Although patent registration protect the intellectual rights, some
patent can't be registered like cars company industrial secrets, military weapon industries,
secure communication applications and techniques, ..,etc.
Besides, the ideation in some fields can't be protected from being stolen like literatures and
arts. Therefore a new concept or type of watermarking is needed.
In this paper, a new concept is discussed for watermarking any type of information
files with the condition that the size of the watermarked file must be greater than the size of
the logo file in mostly small ratio.
2. Image Watermarking
Digital image watermarking is derived from steganography where content is hiding

with other content for secure considerations. The difference between them is that in
steganography the hidden data is in a highest priority while in watermarking the cover and
the hidden data are both in highest priority [3].
Watermarking must have four properties to reach its goal they are:
a) Effectiveness
b) Host signal Quality
c) Watermark Size
d) Robustness
Effectiveness means the watermarking process must be detective. The second property
insures that watermarking process will minimize the host changes. The third property
provided that the watermarking should be of minimum size. The robustness property is to be
withstanding against threats [4].
3. Applications of watermarking
Some of the watermarking applications are listed hereby [5]:
a. Copy Protection: this is done by integrating the copying device with the
watermarking detection.
478
(ICoIT'17)
10thof April, 2017.
b. Broadcast Monitoring: broadcasting and TV channels are monitoring
by watermarking.
c. Medical applications: using watermarking to protecting patients
documents from being altered.

d. Fingerprinting: detecting the legal fingerprint using watermarking
techniques.
e. Data Authentication: watermarking is used with data transmission for
hast authenticity.
f. Owner Identification: watermarking is used to identify the ownership
of a digital media.
4. Watermarking models
Watermarking process can be modeled in two classifications they are

[6]:
a. Communication-based models: in this type watermarking is a process

of communication; a secret message is embedded in a cover in the sender
side and extracted in the receiver side as shown in figure 1.
Communication-based watermarking models can be divided into two
sub-categories also. Using side-information to enhance the process of
watermarking and without using side-information at all.
479
Figure 1. The communication-based model
(ICoIT'17)
10thof April, 2017.
b. Geometric models: in geometric model, watermarking can be viewed as

high-dimensional vectors called the media space and using a number of
regions of desirable properties of the image like embedding region and
detection region.
5. Types of watermarking
Watermarking can be classified Based on the type of document as [7]:

Image Watermarking
Video Watermarking
Audio Watermarking
Text Watermarking
Based on processing method used, watermarking can be classified as:
Spatial-domain techniques
Transform-domain techniques
On the basis of necessary data for extraction, watermarks can be divided
in to two categories:
a. Blind
b. Informed
In blind watermarking original document is not required during watermark
detection process. But in informed, original document is required during
watermark extraction process.
6. The Proposed Authentication Method

The proposed authentication method can be discussed as two phases: the
watermarking phase and the extraction phase which are explained in the next
sections.
480
(ICoIT'17)
10thof April, 2017.
6.1 Watermarking Phase
The idea behind the proposed method is to take the logo file bit by bit
and looking for the equivalent bits in the watermarked file and storing the
positions in another separate file, so it is not a hiding operation therefore the
hidden file will be called logo file and the cover file will be called the
watermarked file.
For example the logo byte in figure 2, need to be located in two bytes of
the watermarked file.
0 1 1 0 0 1 1 1 Logo byte
0 0 1 0 0 1 0 1 1 0 0 1 0 0 1 1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 watermarked content
1 3 6 7 10 12 15 16 Bits Positions of the logo byte in the watermarked file
01 10 11 01 11 10 11 01 Secret Watermarking file
Figure 2. The proposed authentication

method
The first bit of the authority byte is located in the first location in the
cover file, the second is located in location number 3, the third bit in location
six, and, so on the last bit of the hidden byte is located in location number 16.
These locations must be stored in a secret file to be retrieved in the extraction
operation.
The locations file is written in a special new designed format called the
Secret Watermarking File (SWF). Each location in SWF is represented by two
481
(ICoIT'17)
10thof April, 2017.
bits which explain the difference between each two adjacent locations. The first
two bits represent the location of the first bit of the logo file in the watermarked
file, the second two bits represent the difference between the second location
and the first, the third two bits represent the difference between the third
location and the second, and so on the ith two bits is the difference between the
ith location and the previous location.
If the difference between locations is greater than three, the 00 state is

used to mean that add 3 to the pointer as shown in figure 3.
1 0 1 logo bits
0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Watermarked file
6 7 15 Bits locations of authority byte in the authorized file
00 11 01 00 00 10 SWF
Figure 3. Locations treatment of SWF

The first location is 6, which is represented in SWF as "00","11" and
means add 3 to the binary value "11" which is equal to 6, the second location
add "01" to the last location which is equal to 7, the difference between 15 and
7 is 8, which is greater than 3, "00" is inserted and subtracting 3 from 8 which
equal to 5. 5 is greater than 3, "00" is inserted and subtracting 3 from 5, by
subtracting 3 from 5 the value 2 is less than 3 so the binary value "10" is
inserted.
482
(ICoIT'17)
10thof April, 2017.
6.2 The Extraction Phase
SWF file must be securely stored in the owner database or any reliable
application to be used in the extraction process. As in the previously discussed,
SWF contains the locations of the logo file contents (bits) in the watermarked
file, and once getting SWF the user can retrieve the logo file from the
watermarked file.
The extraction method starts by reading SWF which begins by the

location of the first logo bit in the watermarked file and storing it in a new file
called the extracted file, then reading the location of the second bit and
appending it to the extracted file, and so on, until reaching the end of SWF. The
last 21 bits in the extracted file are three characters represent the extension of
the logo file, so the extracted file will have this extension to be viewed in the
viewer application.
7. Secure Authentication Files Management
The output of the proposed authentication method is the secure

watermarking file (SWF). SWF size is as the double of the hidden file or
greater, when a suitable compression method is used SWF size might be tenth
of the logo file size.
SWF is needed to be store in a secure database located in a Reliable Third

Party (RTP) which may be a global agency for this specific activity. Once a file
ownership prove is needed, SWF will be used to rebuild the logo file from the
watermarked file.
If someone needs to save the ownership of digital information, he must

generate its SWF and store it in RTP without saving the original information
483
(ICoIT'17)
10thof April, 2017.
file. RTP will save the name of the owner and send an index number to the
owner which must kept for future watermarking proof as shown in figure 4. The
index number represents the index of information SWF in the RTP database.
The smaller authority number will win the ownership of specific digital
information if it has more than one index number.
Anyone has to send the original information file and its authority number to the
RTP to prove his ownership of this information. ATP saving other extra
information to prove the sender is the owner of the authority number.
RTP .
.
SWF .
Database ANi: namei, swfi
ANi+1: namei+1, swfi+1
Server .
.
.
Figure 4. Reliable third party
Anyone has to send the original information file and its index number to
the RTP to prove his ownership of this information. RTP saving other extra
information to prove the sender is the owner of the index number.
8. Evaluations
o The proposed watermarking method is a new type of authentication

concept it is suitable for any type of digital information files. The watermarking
process performed without inserting or editing the watermarked file.
o The logo of the owner will play a great role in this method. The logo
file must be unique, reserved and clearly describe the owner. The experiments
484
(ICoIT'17)
10thof April, 2017.
show that the size of the information file must be greater than the double of the
logo file size. This condition can be easily performed by extending the size of
the information file by repeating the information or by inserting some remarks.
o RTP must be one of the greatest companies in the world like Google,
Yahoo... etc. because it is reliable, well-known and has infrastructure enough
to perform this authentication method.
o The method is called "Unbreakable" because it looks like a document
registration in a secure place (RTP), and breaking the method means attacking
RTP.
o By using this method anyone can authenticate any unnecessarily
published or secret information.
9. References
[1] F. Mintzer, G. W. Braudaway, M. M. Yeung: "Effective and Ineffective

DigitalWatermarks", ICIP,1997.
[2] Jaseena K.U., Anita John, "Text Watermarking using Combined Image and
Text for Authentication and Protection ", International Journal of Computer
Applications (0975 8887), Volume 20 No.4, April 2011
[3] Lalit Kumar Saini, Vishal Shrivastava,"A Survey of Digital Watermarking
Techniques and its Applications",
(IJCST) Volume 2 Issue 3, May-Jun 2014
[4] Prabhishek Singh, R S Chadha ,A Survey of Digital Watermarking
Techniques, Applications and Attacks, International Journal of Engineering
and Innovative Technology (IJEIT), Volume 2, Issue 9, March 2013
[5] Mei Jiansheng, Li Sukang, A Digital Watermarking Algorithm Based On
DCT and DWT, Proceedings of the 2009 International Symposium on Web
485
(ICoIT'17)
10thof April, 2017.
Information Systems and Applications (WISA09) Nanchang, P. R. China,

May 22-24, 2009, pp. 104-107
[6] Melinos Averkiou, " Digital Watermarking", paper, 2015.
[7] M. Chandra, S. Pandey, R. Chaudhary, Digital Watermarking Techniques
for Protecting Digital Images, IEEE, 2010.
486
(ICoIT'17)
10thof April, 2017.
Using RSA Digital Signature to Solve Pollution Problem in

Single-Source Network Coding
Rami S. Youail
Department of Information System Engineering, Erbil Polytechnic University-Iraq
r.rami.samir@gmail.com
r.rami.samir@epu.edu.krd
r.rami.samir@lfu.edu.krd

Pollution problem in network coding becomes a serious
Article History: problem, where no receiver can be reached by a source it does
Received: 15 March 2017 not need, in other words, the intermediate nodes may send
polluted or faked messages into the network. This injection
will prevent the sinks nodes from recovering the original
Published: 10 April 2017 messages correctly. Additionally, a malicious node can inject
garbage into the distribution network, if undetected; the
Keywords: Network garbage will pollute the whole network. The most difficult
coding, pollution problem in pollution propagation is the expansion of
problem, RSA pollution. If a small number of polluted packets/messages is
encryption algorithm, not detected at the early stages of the network; these small
digital signature, polluted messages will be expanded and used by downstream
authentication. nodes and will affect the entire network. Therefore, the
polluted or forged messages should be detected and filtered
as early as possible before it grows up and distributed overall
the network. To overcome this problem, a signature-based
authentication system is proposed, which can detect the
polluted or forged message at each intermediate/sink node.
The system only needs to transmit the private key of the
source node to each intermediate node.
1. INTRODUCTION
N
etwork coding (NC) is a new technique proposed by Ahlswede et.al.[1] to maximize
the throughput of multicast networks. The main idea behind NC is to allowing
intermediate node to encode the incoming data and generates a new encoded packet.
This technique is different from the traditional networking which is capable of
duplicating and forwarding only. NC has various applications in computer networks, such as
wireless networks and P2P systems. Network coding produces new challenges of security,
for example, the pollution free problem, where no receiver can be reached by a source it does
not need. If the pollution does not detected as early as possible, the pollution will grow up
487
(ICoIT'17)
10thof April, 2017.
since the polluted messages could be used by all down-stream nodes, this kind of pollution
will prevent the sink nodes from recovering the source messages correctly.
2. Literature review
Many attempts have been done to overcome this problem. In [2], the pollution problem of
network coding is solved by using a network of tagged pipes; where each link is partitioned
into tagged pipes and each tag is a subset of the source messages. The packets outflowing in
each pipe can only be a function of the sources mentioned in the tag. Authors in [3] can detect
or filtered the polluted messages in sink nodes only. The work in [4] presents a security
scheme for network coding that reduces the cost of verifying blocks on-the-fly while
preventing the probing of malicious blocks. Digital signature, which is adopted in this work,
is used to prevent this problem. The source node generates its own signature and attached it
to the messages, the intermediate nodes check whether this signature is valid or not. The
source node needs only to transmit its secret key to the intermediate nodes in order to
regenerate a signature after packets being network coded.
3. INFORMATION SECURITY
In any information security system, there are several objectives need to be achieved by
the system itself. If any of these objectives is absent; there will be a defect in the system and
will not function properly [5]. RSA signature, by itself, can achieve these objectives. Some
of these objectives which are achieved by the proposed system are listed below:
TABLE 1: ACHIEVED OBJECTIVE IN PROPOSED SYSTEM
confidentiality Only authorized persons can access the secret information.
data integrity Ensuring information has not been changed during the transmission.
Message authentication
It ensures that the sender is the real one, not other person claiming.
signature Binding information to a source and the source cannot deny.
4. DIGITAL SIGNATURE
A cryptographic process which is considered as the backbone in message authentication,
authorization, and non-repudiation is the digital signature. The purpose behind a digital
signature is to provide a method for a signer to bind its identity to a piece of information. In
easy way, the digital signature of a message is a number resulted from a function of some
secret information belongs to the signer, and on the content of the message [5].
A. Preliminaries or Basic Terminologies and Concepts

In the following, some terms and basic concepts are defined [5].
M is a set of message space which can be signed, M consists of strings of symbols from
an alphabet, each element belongs to M is called plaintext.
K is a set of key space.
488
(ICoIT'17)
10thof April, 2017.
S is a set of elements called signatures.

SA is a transformation function from M to S, and is called a signing transformation for
the signer. SA is kept secret by the signer and will be used to create signatures for
messages from M.
VA is a transformation from the set MS to the set {true, false}. VA is called a
verification transformation for the signers signatures, which is publicly known, and is
used by other entities to check the originality of the signature created by the signer.
The signing procedure (phase two in the proposed scheme) consists of two steps. In the
first step, the signer computes a signature s for each message m M , and then transmits the
pair (m,s) at the second step.
The verifier (phase three) needs to verify that a signature s on a message m is created
(signed) by the real signer, so he\she needs to perform the following:
1. Obtain the verification function VA of the signer, which is publicly known.
2. Calculate u=VA(m,s).
3. If u=true; the signature is correct and accepted, otherwise reject it.
B. The Proposed Model

The proposed model consists of three phases:
Phase one: the source node chooses the security parameters and calculates the public and
private keys, send the secret key to intermediate nodes via secure channels. The
source node sends the public key via unsecure channel or attaches it with the
message.
Phase two: the source node calculates the signature which will be attached with its
messages.
Phase three: the intermediate/sink nodes verify the received message.
RSA digital signature scheme is adopted in this work, the source node performs the
following steps at phase one [5], [7]:
1.Generate two distinct random numbers n1 and n2. Note that n1 and n2 must be prime and
large enough.
2.Calculate n= n1 n2 and (n1 1)(n2 1) .
3.Select a random integer e, which represents the public key, such that 1 e and gcd
(e, ) 1.
4.Calculate the unique integer d where 1 d , such that ed mod (n1 1)(n2 1) 1
5.The sources public key is (n,e) and its private key is d.
Taking in consideration that n should be long enough (for example; n=1024 bit length)
and the difference between the primes n1 and n2 (which are half length of n, i.e.: each one
of them is 512 bits of length) should not be small. If n1 and n2 is small, then n could be
factorized and the secret key could be determined [5].
At the second phase, the source node calculates the signature for each m M as follows:
1. Compute the signature s=md mod n, 0 m <n.
489
(ICoIT'17)
10thof April, 2017.
2. Attach s to m and send the packet (m,s), downstream to the nodes.

At the third phase, each node will verify the received signature by using the verification
transformation, if the received packet has a valid signature and it does not contain any forged
message, it will be accepted, otherwise ignored. The transformation verification is given by
m=se mod n.
5. SYSTEM MODEL
There are X1,X2,,Xm information source observed at source node S. The intermediate
nodes are capable of encoding the incoming messages Xi ,i=1,..,m into encoded message.
Hence, the encoded messages could be written as:
E (1 , 2 ,...., m ) ( X 1 , X 2 ,...X m )T mod (1)
Where i is a coefficient picket up randomly from a finite field (Galois Field). Every
intermediate node performs network coding operation on its incoming packets and generates
the vector E, then computes the signature for E. If S denotes the signature function, then the
signature will be calculated by using the, multiplicative property which is sometimes referred
to as homomorphic property [5], [6], [7] as:
d
r
S ( E ) ( X i i ) mod mod n (2)
i 1
Where r represents the number of incoming links (packets to be network coded) at any
node , then append S(E) to E and forward it downstream, the packet format will be [E + S(E)],
where the first part represents the encoded message and the second represents its signature.
The intermediate/sink node verifies the received message (packet) according to the source
(signer) public key and the message signatures. Simply, the verifier check whether
S ( E ) e mod n E (3)
If the above equation satisfied, then the packet will be considered as valid one, otherwise
its polluted and will be discarded. Since each intermediate node or sink node will receive r
packets from its r incoming links, so we need to compute (3) for r times in order to verify all
the incoming packets which will take long time.
To increase the calculation speed of verification process; we support the batch verification
[8]; which will speed up message verification. The verification process in 3 will change to
be:
e
r r
Si mod n mi mod n (4)
i 1 i 1
To see the proof of the above formula the reader may refer to [7].
490
(ICoIT'17)
10thof April, 2017.
6. SECURITY ANALYSIS
The message is defined as polluted by the inequality (5):
S(E)e mod n E (5)
which means that the content of the message is incompatible with the signature attached with.
To prevent unauthorized persons/nodes from polluting the encoding message; each signature
should be calculated based on the secret key and the verified message.
In this system, determining the sources private key from its public key is equivalent to solve
the problem of integer factorization, which is a hard problem to solve and its hardness
depends on the length of n.
7. CONCLUSION
In this work, RSA digital signature scheme is adopted to solve the problem of pollution
in single-source network coding. Each intermediate node will receive many signed packets;
this will slow down the verification process. To overcome this slowness; the batch
verification is used to speed up the verification process. The only disadvantage in our system
is that we need to transfer the secret key to all the intermediate nodes via secure channels.
8. REFERENCES
[1] R. Ahlswede, N. Cai, S. Li, R. Yeung, Network information flow, IEEE Transaction
on Information Theory, vol. 46, no. 4, pp 1204-1216, 2000.
[2] Y. Wu, On constructive multi-source network coding, in IEEE International
Symposium in Information Theory, Seattle, July 9-14, 2006.
[3] S. Jaggi, M. Langberg, S. Katti, T. Ho, D. Katabi, M. Medard, Resilient network coding
in the presence of byzantine adversaries, IEEE Proceeding INFOCOM, 2007.
[4] Gkantsidis, P. R. Rodriguez, Cooperative security for network coding file distribution,
IEEE Proceedings INFOCOM, 2006.
[5] A. Menezes, P. V. Oorschot, S. Vnastone, Handbook of Applied Cryptography, 5th. ed.
Florida: CRC Press, 2011.
[6] J. Talbot, D. Welsh, Complexity and Cryptography, 1st. ed., Cambridge, Cambridge
University Press, 2006.
[7] L. Harn, Batch verification multiple RSA digital signature, IEEE Xplore, vol. 10, Issue
12, pp. 1219-1220, 11 June, 1998.
[8] M. Bellare, J.A. Garay, T. Rabin, Fast batch verification for modular exponentiation
and digital signature, Proceedings of Advances in Cryptography, LNCS, vol. 1403,
1998
491
(ICoIT'17)
10thof April, 2017.
Virtualizing Network Services: Practical Perspectives
Emad H. Al-Hemiary
Baghdad - Iraq

The work in this paper focuses on creating virtualized
Article History: network services for datacenters based on open-source
Received: 18 March 2017 software and explains the preliminary and mandatory
requirement associated with. Datacenters are necessity for
organizations to maintain their workflow and processed data
Published: 10 April 2017 as well as fast and secure data retrieval. The current and future
design emerge virtualization as key technology in efficient
Keywords: and manageable datacenters. Therefore, the design proposal
Virtualization, is based on virtualizing physical servers to the required
virtualized network services following infrastructure as a service. The designed
services, open-source and implemented network use open-source operating systems
datacenters. and software exclusively. Different types of services like
web, email, eLearning, databases, and Cloud ready are
included and tested. The runtime for the implemented
datacenter clearly supports this proposal and confirms the
direction of use open-source software over closed-systems for
building highly customized network services and reduce
overall cost.
1. INTRODUCTION
D
atacenter is a combined electrical, mechanical, and civil infrastructures
within an organization, city, country or worldwide level. Whether it is
centralized or distributed, many requirements are listed by specialists for
efficient and powerful datacenters [1,2]. These requirements include specialized rooms
containing networking devices and servers secured by special equipment and maintained
operational most of the time using backup electrical systems supporting power failure
situations. The main function of datacenters is the store and process of data as well as
network services. Every organization; small, intermediate, or enterprise have their own
datacenter where it keeps and centralize its information technology (IT)
communications within one or more physical geographical area. Storing data coming
from within the organization or remotely from other connected sites is associated with
492
(ICoIT'17)
10thof April, 2017.
security measures. Therefore, IT specialists link data storage servers with security
access [3].
When datacenter network requires more servers to be added in order to increase
capacity or add new type of service, physical hardware servers increase and large
number of servers becomes an issue which need more management, control and
operational power. This where virtualization [4] takes place. Virtualization is the
process of exploiting servers power through creating smaller virtual servers, services,
or functions. Imagine a hundred service and each counted service requires a standalone
server with ( , , ) where the latter symbols represents (processing speed, dynamic
memory, permanent storage). If one physical server have (, , ), then approximately,
a new virtual machine (VM) could be created as long as:
> |, (1)
Where is the available memory given by:

(2)
and is the required memory for the operating system running virtualization and the
symbol | represents conditional relation on and . As long as the and (central
processing unit supports virtualization and the permanent storage is dynamically
allocated and adequate) are satisfied, the virtual machine is related to the available
memory. According to that, the number of VMs can be counted approximately as:

(3)

Thus, the major requirement for creating virtual machines within physical sever is the
capacity of its dynamic RAM (random access memory). For the previous calculations,
the one hundred servers example could be distributed over five, ten or twenty physical
servers depending on the servers power and capacity. Whenever the cost and space is
a problem, virtualization is the solution. Many management software called Hypervisors
[5] allow many guest operating systems to run on the same physical server machine.
Kernel Virtual Machine (KVM) [6], VirtualBox [7], Xen [8], and VMWare [9] are
examples of widely used hypervisors. Since the focus of this work is based on open-
source, Linux-based software are used exclusively. The rest of the paper is organized as
follows: Section 2 discusses network services and their practical implementation
requirements and section 3 gives the implemented virtualized network layout.
Conclusions are given in section 4 and finally references are listed in section 5.
493
(ICoIT'17)
10thof April, 2017.
2. NETWORK SERVICES AND THEIR PRACTICAL PERSPECTIVES

The aim of this work is to highlight the practical requirements to establish an open-
source datacenter supporting small business organization and based on virtualizing network
services. Basically, Domain Name System (DNS), Web, and Email are main services
required to establish datacenters. The following sections explains these services in details.
2.1. DNS
The naming strategy followed since the Internet becomes globally wide helps end users
to use names instead of numbers in web browsing and servers connections. This is better
remembered since numbers; referred to IP (Internet Protocols) addresses, are difficult to
remember and might change according to organization network updates. Therefore the
naming strategy (DNS) is essential whenever web, email, multimedia or other types of
services are operational. Shortly, DNS functionality is to map names to IPs and vice versa.
But this is not the only functionality it possesses! DNS is actually the key to other services
especially when ownership and authentication is required for the domain it serves. To
highlight this function, figure 1 shows a typical DNS network services or protocols hosted
by an organization denoted by local site. As the figure shows, an authoritative-only DNS
(responds only to queries about domains or subdomains configured inside by the network
administrator) with master-slave configuration to ensure continuous operation and this
should be physically separated and not in the same geographical area and also not logically
addressed by the same subnet to isolate any power or network failure and keep service
operational. The DNS server consists of two zones: forward and reverse. The forward zone
configures the domain name like anydomain.com and lists all the required records within a
file known to the DNS process called forward zone, while the reverse zone is required for
IP-Name verification, this is explained later on. As shown in figure 1, the master (primary)
DNS service updates its slave (secondary) with zone records periodically or when the zone
file serial number is changed, and in this way the slave becomes an updated copy from the
master.
Local Site
Web
Mail Remote Site

Secure update
Master Slave
INTERNET
E-Commerce Authoritative-Only Authoritative-Only
Database
FIGURE 1. Master-Slave practical implementation for reliable DNS operation.

When a request from the Internet targeted one of the services (Web, Mail, etc.) in the local
site, the DNS receives a query for the IP address of the server hosting this service from the
requester (the operation is actually a series of requests and replies [10]). The DNS server
through its service contacts its local database and reads the forward zone A record (The A
records maps domain or subdomain names to their assigned IP address) as shown in figure
494
(ICoIT'17)
10thof April, 2017.
2. It is important to mention here that the local site may have its own DNSs for serving local
clients and these DNSs are not what figure 2 shows. As the reader recalls, the main IP
version 4 TCP/IP settings requires four IP numbers to work: IP address of the machine,
subnet mask, default gateway (first router the client connects to) and the primary DNS to
contact for web requests. The latter one is called non-authoritative and responds to any query
through its local cache. On the other hand, DNS holds authentication records for verification
purposes. As an example, the google site verification is a TXT record issued by Google for
a specific domain to verify its ownership and this is used typically for verifying domain mail
server as it will be shown later on. Table 1 shows authentication records and their
functionality.
A query from the requester
DNS reach Nameserver
(ns1) for anydomain.com 1 What is the IP address for
ns1.anydomain.com 2 www.anydomain.com?
DNS Requester: Open page for

INTERNET
www.anydomain.com
3
4
4
ns1 IN A 1.2.3.4 6
ns2 IN A 5.6.7.8
www IN A 1.2.3.5
mail IN A 1.2.3.6
... 5
Forward zone
Web Server
www.anydomain.com
1.2.3.5
FIGURE 2. DNS response to queries. All messages are exchanged over UDP port 53.
2.2. MAIL
One of the main services required by an organization for exchanging messages is the
electronic mail. E-Mail service is composed of many protocols for sending and receiving
messages. The E-Mail server uses Simple Mail Transfer Protocol (SMTP) [11] either
unsecured (port 25) or secured (Secure Socket Layer - SSL port 465 or Transport Layer
Security TLS port 587) to send messages to another server. The receiving protocols are
mainly Post Office Protocol or simply POP3 (Unsecured port 110 or SSL port 995) and
Internet Message Access Protocol or simply IMAP (unsecured port 143 or SSL port 993).
Due to the fact that email is a source of Spams and Viruses, ownership and authentication of
the sending mail server is required by the receiving mail server. The process involves a trace
back from the recipient Mail Server to the sender as shown in figure 3. When an email is
send from some mail server to another one, the recipient mail server contacts the ISP (Internet
Service Provider) of the senders mail server asking for reverse lookup of its IP address to
its name. If a match is found then the email is accepted, otherwise the recipient mail server
495
(ICoIT'17)
10thof April, 2017.
establish a reject policy regarding the senders mail server and its IP address. The case
becomes worse when the recipient mail server is Yahoo, Gmail, Hotmail or other free web-
based global mail servers. This is due to the SPAMHAUS Project [12] where IP addresses
are stored as black listed. Therefore, for practical perspectives, any mail server should contact
SPAMHAUS database to ensure and verify senders.
As figure 3 reveals, the role of DNS is clearly mandatory in the process of sending and
receiving email. Without the authentication mechanisms shown in table 1 and figure 3, the
email sender cannot sustain his eligibility and thus considered as Spam.
TABLE 1: Authentication records required for practical authoritative-only DNS servers
Record Type Typical form record Function

Identifies which mail servers are
permitted to send email on behalf of
"v=spf1 +a +mx +ptr +ip4:1.2.3.6 anydomain.com. The purpose of an SPF
SPF TXT
include:mail.anydomain.com ~all" record is to prevent spammers from
sending messages with forged from
addresses at your domain.
Unique security token provided in the
Google Admin console's verification
Google google-site-verification=-
instructions. The token is a 68-character
domain TXT Aqw1jSo8lsOZwbwd-cQ5bQvtm-
string that begins with google-site-
verification M1DqVEHEIKXZ1z6U
verification=, followed by 43 additional
characters.
Define DomainKeys policy and public
Domain encryption keys for a domain name.
TXT "v=DKIM1; p=domain public key
Keys Each email message will be signed by
domains public key.
The goal of DMARC is to build on this
system of senders and receivers
_dmarc.anydomain.com IN TXT collaborating to improve mail
DMARK TXT
v=DMARC1; p=none authentication practices of senders and
enable receivers to reject unauthenticated
messages.
496
(ICoIT'17)
10thof April, 2017.
emad@anydomain.com hasan@yahoo.com
To: hasan@yahoo.com 1
From: emad@anydomain.com SMTP TLS
Message: ... connect to
mail.yahoo.com
Accepts Connection
2 2
Receives Message
Mail server 3
Hold message
mail.anydomain.com INTERNET Read sender s IP address
1.2.3.6
4 Is 1.2.3.6 listed?
Answer: No -> Perform 5
5
ISP s DNS Answer: Yes -> Return rejection
Is 1.2.3.6 point to mail.anydomain.com? response to sender with 550-5.7.1 code
Yes: Receive email and consider sender Authentic
No: Reject email and notify sender
...
1.2.3.6 IN PTR mail.anydomain.com
... SPAMHAUS Project
Reverse Zone Database
FIGURE 3. Senders mail server authentication mechanism. PTR refers to pointer as it

points IP address to name
2.3. WEB, DATABASE AND OTHER SERVICES

Most of the Internet services requires web access using Hypertext Transfer Protocol
(HTTP or HTTP over SSL HTTPS). The process involves a mechanism of requests and
responses messages between the Web client and Web server to download and view the page
contents using an application layer software like Internet Explorer. The pages forming the
dynamic website normally store its contents in databases for future retrieval. Many websites
can be hosted on the same Web server using software like Apache2 which is an open-source
platform for hosting web contents. The websites themselves are written using server-side
scripting languages like PHP5 combined with HTML and Javascript and data are stored
within the servers permanent storage using language like SQL (Structure Query Language);
Finally, datacenters include many other services like multimedia streaming, voice managers,
Cloud for private use, workflow, human and financial systems and other useful related
services.
3. NETWORK SERVICE VIRTUALIZATION USING OPEN-SOURCE

In the previous sections, network services are explained with highlights on their
practical implementation. All of these services can be virtualized using one or more physical
server entity characterized by high dynamic RAM as it has been explained in section 1.
Therefore, this work considers an HP ProLiant DL380 G8 Server with 32GB DDR3
Registered Dual in-line Memory Module (DIMM) and Intel(R) Xeon(R) CPU E5620 @
2.40GHz, 16 cores. This physical server runs UBUNTU 14.04.3 Linux operating system and
virtualized using VirtualBox hypervisor and controlled by the front end web interface called
PHPVirtualBox and divided into VMs hosting the network services shown in table 2. The
network interface card of the server is configured in bridged mode to assign different IP
addresses to each VM. The detailed network diagram is shown in figure 4 where multiple
sites cooperates to maintain DNSs databases as previously explained.
497
(ICoIT'17)
10thof April, 2017.
In the network shown in figure 4, the master DNS located in the college of Information
Engineering / Al-Nahrain University is synchronized with its slave DNS installed in the
department of Research and development at the Ministry of Higher Education and Scientific
Research. Both of these servers run NSD (Name Server Daemon) software and encrypt their
synchronous updates with private key. Whenever a request reaches the Master DNS
(identified by its parent DNS), the master DNS responds back with query answer unless it is
unreachable and consequently the slave DNS responds back on behalf. Figure 4 shows also
the reverse DNS location in the network of SatGate ISP for proper email server ownership.
The whole network is operational without problems and it is completely built using open-
source software as indicated in Table 2. Maintaining virtualized services running requires
periodic monitoring and backup and these are also managed using available open-source
software like Nagios [13] and Amanda [14].
TABLE 2: List of virtualized network services using open-source software
Service Name Software VM

512 Mbyte of RAM, Single CPU,
Master DNS NSD [15]
4GByte dynamic storage
Postfix [16], Dovecot IMAP/POP3
[17], MySQL Database [18],
Mail Server SpamAssassin Mail Filter [19],
100 GByte dynamic storage
amavis [20], Roundcube [21],
Apache 2 [22], PHP5 [23]
Web Server Apache 2, PHP5, MySQL Database (10 VMs) Depends on Application
E-Learning Moodle [24], Apache 2, PHP5, 2000 Mbyte of RAM, Single CPU,
Server MySQL 100 GByte dynamic storage
Workflow Apache 2, PHP5, MySQL
100 GByte dynamic storage
File Sharing and 4000 Mbyte of RAM, Single CPU,
ownCloud [25]
Data Storage 500 GByte dynamic storage
Other services - Depends on Application
4. CONCLUSIONS
In this paper, practical perspectives in network service virtualization are highlighted.
Different types of services like DNS, Web, and Mail are explained and critical issues running
them are pointed out. The work is implemented using open-source software features no cost
(low cost when support is needed), reliability and high availability. It has been shown that
physical servers can be divided into multiple VMs using hypervisors software like
VirtualBox to exploits its resources, and each VM is considered as a dedicated server with
assigned amount of memory that is considered the key to virtualization. Therefore, the
amount of memory decides how many running VMs a hypervisors can produce as it has been
shown by the approximate equations in section 1. The work shows practical implementation
of multiple site-to-site cooperation to achieve the required ownership, authentication and
verification for domain and mail services. Finally, the use of virtualization easily expands
network services as the process involves adding virtually and not physically.
498
(ICoIT'17)
10thof April, 2017.
Slave
Authoritative-
Only DNS
Security
Slave DNS
Measures
Part of Virtualized Network @

RRD / MOHESR INTERNET
Reverse
DNS
SatGate Iraq
Security ISP
Master Measures
Authoritative-
Only DNS
...
Local Network
Virtualized Servers @ Al-

Nahrain University / College
of Information Engineering
Physical Servers
HP DL380 G8
FIGURE 4. Network diagram showing virtualized services distributed over multiple

sites
5. REFERENCES
[1] M. Rahman, and A. Esmailpour, A Hybrid Network Architecture for Data Centers, in
IEEE First International Conference on Big Data Computing Service and Applications
(BigDataService), 2015, pp. 7-13.
[2] G. Schulz, The Green and Virtual Data Center, Boston MA, USA:Auerbach Publications,
2009.
[3] M. Ammar, M. Rizk, A. Abdel-Hamid, and A. Aboul-Seoud, A Framework for Security
Enhancement in SDN-Based Datacenters, in 8th IFIP International Conference on New
Technologies, Mobility and Security (NTMS), 2016, pp. 1-4.
[4] N. Jain, and S. Choudhary, Overview of virtualization in cloud computing, in
Symposium on Colossal Data Analysis and Networking (CDAN), 2016, pp. 1-4.
[5] A. Kivity, Y. Kamay, D. Laor, U. Lublin, and A. Liguori, "kvm: The linux virtual
machine monitor", Proc. Linux Symp., vol. 1, 2007, pp. 225-230,
[6] Kernel-Based Virtual Machine (KVM), https://www.linux-kvm.org, (Accessed: 17th
March 2017).
[7] Oracle VM VirtualBox, https://www.virtualbox.org, (Accessed: 17th March 2017).
[8] The Xen Project (Xen), https://www.xenproject.org, (Accessed: 17th March 2017).
499
(ICoIT'17)
10thof April, 2017.
[9] Virtual Machine Software (VMWare), http://www.vmware.com, (Accessed: 17th March

2017).
[10] RFC1034,1035 - Domain Names - Implementation and Specification IETF,
https://www.ietf.org/rfc/rfc1034.txt (rfc1035), (Accessed: 17th March 2017).
[11] RFC 5321 - Simple Mail Transfer Protocol - IETF Tools,
https://www.ietf.org/rfc/rfc5321.txt, (Accessed: 17th March 2017).
[12] The Spamhaus Project, https://www.spamhaus.org, (Accessed: 17th March 2017).
[13] Network Monitoring, Nagios, https://sourceforge.net/projects/nagios/, (Accessed: 17th
March 2017).
[14] Amanda: Open Source Backup, http://amanda.zmanda.com, (Accessed: 17th March
2017).
[15] Name Server Daemon (NSD), https://www.nlnetlabs.nl/projects/nsd/, (Accessed: 17th
March 2017).
[16] Mail server, Postfix, http://www.postfix.org, (Accessed: 17th March 2017).
[17] Secure POP3 and IMAP server for Linux, Dovecot, https://www.dovecot.org,
(Accessed: 17th March 2017).
[18] Open source database, MySQL, https://www.mysql.com, (Accessed: 17th March 2017).
[19] Apache SpamAssassin Project, http://spamassassin.apache.org, (Accessed: 17th March
2017).
[20] Amavis Virus Scanner and Checker, https://www.ijs.si/software/amavisd/, (Accessed:
17th March 2017).
[21] IMAP client, Roundcube, https://roundcube.net, (Accessed: 17th March 2017).
[22] The Apache HTTP Server Project, https://httpd.apache.org, (Accessed: 17th March
2017).
[23] PHP scripting language, http://php.net, (Accessed: 17th March 2017).
[24] Moodle, https://moodle.org, (Accessed: 17th March 2017).
[25] OwnCloud, https://owncloud.org, (Accessed: 17th March 2017).
500
)(ICoIT'17
10thof April, 2017.
E. Banking
.

Haval.aqrawi@epu.edu.krd
.

luqmanruadzy@epu.edu.krd

1661
:
1662 :
:
:
:
, ,
,

,
.

( )Font

( )
.
501
)(ICoIT'17
10thof April, 2017.

( )Software

( )Interface

.

.
.4 :

.

. .

. .
502
)(ICoIT'17
10thof April, 2017.
.4.4 :
""Electronic Banking
" "Internet Banking "Remote Electronic
"Banking ""Home Banking ""Online Banking
""Self Service Banking " "Web Banking

"
".

) (pc
PME Personal Financial Management (Microsoft's
)Money ). ( Ntuits Quiken
:
,
.

.
.4.6 :

.
1685 American Express
Bank Americard 1675 Visa
Carte Bleue .
1657 France Telecom
Cartes mmoire 1662
Cartes puce .

.

()Foundation Sidssadol and Aokachir

.
503
)(ICoIT'17
10thof April, 2017.
( )Tom Bloomfield ( )Mondo

.
( )Mondo
67
.

.
( )UAE

%68 .
.4.1 :

E-
Loan

.

-
-

.

(
)

.
504
)(ICoIT'17
10thof April, 2017.
.6 :

.
:
INFORMATIONAL .6.4

.

.
COMMUNICATIVE .6.6

( )Email (.)Phone
TRANSACTIONAL .6.1

.

.

.
.1 :
.1.4

. "" " "
505
)(ICoIT'17
10thof April, 2017.

.
.1.6

.
.1.1 :
( )

.
.1.1 :

.
.1.3 :

.
.1 :
.1.4

,
.

.
506
)(ICoIT'17
10thof April, 2017.
.1.4.4

.
.1.4.6
-:
.
.
.
.

.1.4.1

268
1
.
.1.4.1

.
.1.4.3

507
)(ICoIT'17
10thof April, 2017.
( )Sony 15

.

.
.1.6

.

,

:
.1.6.4

.
,

,

.
.1.6.6

508
)(ICoIT'17
10thof April, 2017.

.
.1.6.1

.
.1.6.1

- -
.
.3 :

( ) ICT

.

.

.
509
)(ICoIT'17
10thof April, 2017.
.3.4 ( )I.T.
:
.3.4.4 (:)MICROSOFT OFFICE

( )Microsoft Access, Microsoft Excel

.
.3.4.6
:
.
.3.4.1

.
:
.1 ()PHP-MY SQL SERVER
.2 ()ASP.NET
.2 :

:
.2.4

.

.
.2.6

,
:
510
)(ICoIT'17
10thof April, 2017.
.2.6.4
( )Servers
.
.2.6.6

( )Load ( Account
.)Users
.7 :
.7.4
.1
.
.2 .
.3
.
.1
.
-8

.
.7.6
. .1
.2
.
. .3
. .1
. .8
. .7
.6
.
.5
.
511
)(ICoIT'17
10thof April, 2017.
-17 ,
.
-17
.
-11 .
-12
.
" " -1
( )17 . 2717
. .2711 , -2
.
" " -3
5-6 . 2717
.2777 -1
-8
.2777
-7
.2777
-6 " "

.2002
-5
.2002
-6 " " .2777
-17 " (
) " .2775
512
(ICoIT'17)
10thof April, 2017.

1- Neil B.Harrison and Paris Avgeriou, Pattern-Based Architecture Reviews, IEEE,
2010.
2- Hui Li, Jingiun Zhang and Yuejuan Chen AspectOriented Modeling in Software
Architecture Pattern Based on UML, IEEE, 2010
3- Julie Street Fant,Building Domain Specific Software Architecture from Software
Architectural Design Patterns, ACM, 2011
4- B.Fernandez and Rouyi Pan, A Pattern Language for Security Models, Eduardo,
Plop Conference, 2001
5- Peter S. Rose, "UBank Management & Financial ServicesU" , Mc graw Hill,
International Edition, Sixth Edition,2005.
6- Fraser +Gup and kolari, "UCommercial Banking"U the UManagement of RiskU ,
South- Western Publishing (U.S.A), 2001
7-

1- www.centralbank.ae
2- www.ityarabic.org/e-businers.
3- www.bank.org/arabic/period
4- www.Bank.Of.cd.com
Abstract
Since the uprising spring in 1991 and the formation of the first government in the
Kurdistan Region in 1992 heighten attention a lot of information technology and its divisions
opened all the existing universities, including public and private universities, this opened the
large prospects in use in all areas, and a whole generation of sons appeared region have
become experts in the software to work on ready-made applications relying on computers
that are designed by private companies, whether they are in Kurdistan or abroad and solving
the problem of the quality of the writing (Font) in and databases which use in Kurdish
language Sorani dialect, Where previously the problem of Kurdish letters a major obstacle
in the use of computers in departments (Official office ) and private companies, and during
this period has been a set of software applications designed on PCs for a number of
government departments in interfaces multi-lingual, and there are applications only use the
Kurdish language.
With such a large number of specialist software engineering, banks and have access
to advanced degrees and their participation in the sessions and conferences outside the
province, but the government did not benefit them in the construction of application software
particularly in this area so that it is able to absorb the most important financial business in
513
(ICoIT'17)
10thof April, 2017.
the region and three interfaces multi-lingual, to see the problems this meet and we do not
benefit in a scientific discreet in this area, which will facilitate access to perfection and a goal
that please him.
In the developed countries and the Gulf in particular countries and the growing e-
commerce operations has become a great need for a new type of non-conventional banks in
terms using technology beyond the normal pattern of performance does not adhere to a place
or a specific time important, and as a result of the accelerating growth of information and
communication technology electronic banking, which contributed In came effective in
providing a variety of services, low-costs time and place.
514
)(ICoIT'17
10thof April, 2017.

( )
The Effect of Using Tow Approaches of Distance Education
( Computer & the Internet) on Students' Achievement at Open
Education College and Their Motivation for Learning Mathematics

Dr.ahmed_Auun67@lfu.edu.krd

:

:
( ) . :
:
( ) ( )08
( )18 :
( ) 43 ,
.. ,
. ,:,

.

.

.
.

.
( )11-6 2772
( )
(
6886.)6
.
515
)(ICoIT'17
10thof April, 2017.
( ) calculus
.
( )1
2773-2772 .
()4

6883-6881 6881-6881 6886-6884 6884-6888
%33.7 %33.23 %34.11 %31.70

.
( ) %27
) )*(.
.

( 6881 .)4

) (Distance Education .

.

.

) ( Hessemer , 1986 (
)2771 ) ( Gaeddert , 2001( ) 2773 ( ) 2773 (
) 2771 ( ) 2777 (
)2772 ) ( Shiratuddin , 2001) ( Heo & Chow , 2005

) (Douglas 1999 ( ) 2773 ( ) 2778

.
) (Wegner &Others , 1999 ( Favretto & Others ,
) 2005
( . ) 2771

. (*)
516
)(ICoIT'17
10thof April, 2017.

( ) .

.
-:
(( " "
))

.

(. 6886
)6

( 2773 .) 1

( . 6883
)01

- -1 ((
))
( 17 - 18 2771 )
- 2 (
( 27 - 16 2778 )
- 3 26 26 2777

.
( 6881 )66-41

517
)(ICoIT'17
10thof April, 2017.

.

).(kosakowski ,1998 , EDO-IR-98-04 ( )2771

( ) ( 6884 .) 133

. ( Binder , 1993,p.39) .

( . , 6881 ,.) 43-41
) ( World Wide Web

.
) ( internet explorer ) ( e mail ( 0n- line
(Corl , 1996 ,ED402998 ) .

(cotton,1996,ED400577.
.

( 4111 .) 04

( . 2771.) 17

( . 6883 .) 11 - 16

( .
6881.)41 ) (Kuhn,2001

.
) . (Kuhn,2001,p.1 ) ( Charp,2000
518
)(ICoIT'17
10thof April, 2017.
( ) 67

) . (Charp ,2001,p.10-14

. ( )2773

( 6881 .)7-2
( )2771

( . 6884 ) 41-44

.

( 4117 . ) 47-43

( . 4101 . )1

( 6884 .)2

( 4101. )6

. ( )1668 "
"
( .
4113 40 . )68
.

( 6888.)42-43

( 6888 . ) 473
" "
519
)(ICoIT'17
10thof April, 2017.
(
4112. )46

.)Lawson & Chinnappn, 2000, P 26( .

.

( 4113.)613-611

.
( )1656

-1 .
-2
-3 .
-1 ( . 4101 ) 164-168

( 4102 . ) 660-667

(Johnson , 2000 ,ED .
.)441302

( 4111 .) 616
(
6888 .)41 ( ) 2778

(6883 .)131 ( ) 2777

520
)(ICoIT'17
10thof April, 2017.

(. 6888)673-671
( )1656

( 4101 . )462( )1661

) 4111 . ) 1

(Middleton .
) &Spenias ,1999,p.67

)(Miltiadoa&Wilhelminac,2003,p.5
-:
-1
( )
.
-2
.
-3
.
-1
.
-8
.
-7
.
-6
.
:
( )
.
:

-1
()7.78
521
)(ICoIT'17
10thof April, 2017.
-2
.
()7.78
:
-:
-1 ( )
-2 /
-3
:
-1 ): ) Distance Education
): (Simonson & other ,2000
"
".
. (Simonson & other , 2000 , p.7). ) ( (USDLA , 2004
) : "
" (USDLA , 2004 ,on-line) .
"
( )
" .
-6 ( ) Computer
( : )2777 "
" ( . 6888 ) 7
( : ) 2771 "
) (Data ) ( Instruction
) (Information " ( . 6881 )61
"-: (Pentium 4
) ) (Mouse ) (Keyboard
) ( Display Monitor ) (CD-D
) (Speakers ) ( Soft Ware ( Windows
) xp pro. SP2 "
-1 )( Internet
) ": (Costro ,1996
".
) (Costro ,1996,p.2 ( ": )2777

( 6888 "
) 680 ( ": ) 2771
" ( .
6884 ) 1 -:
522
)(ICoIT'17
10thof April, 2017.

( ) 2777-2778
"
-1 )(Achievement (: )Chapln,1971
" (Chaplin , 1971 , .
) p.5 ) : ( Good ,1973 " " .
) ( Good ,1973 ,p.7 ( )1655 -: "

" ( . 4100 ) 44 "-:

"
- 3 ) ( Motivation of Learning
( ) 1666 ":

" ( . 4111 ) 42
) ": McMinn &Bulter,2003

"( McMinn &Bulter,2003 ,p.1). -:
"
" .
.

) : ( Distance Education
-: :

.

( . 4110 ) 6
( ) 1667 -:
-: ( 1587 ) .
-: ( ).
-: ( 1676 - 1677 ).
-: .
-: ( . 4112 ) 4
523
)(ICoIT'17
10thof April, 2017.

.

.
( ) 2777 :
-1 .
-2 .
-3
.
-1 .
-8
( . 6882)4
-:
( )2773
-:
-1
1517
-2 (1587)
( ).
-3 ) (Latter Day saints ( 1621)
.
-1 1618 ) ( Iowa
) (Etv .1687
-8 ( 1658) ( 1667)
) (NSF ( . 6881) 1
) (Valentine,2002 ( ) 177

524
)(ICoIT'17
10thof April, 2017.

(Valentine , 2002 , p.1) .
-: :
( ) 2771( ) 2777 :
-1 .
-2
-3
.
-1 .
-8 .
-7 .
-6 .
-5 .
-6 .
( 6881 ( (0-7 6882) 1-0
-:
( )2771 -:
-1

-2 .
-3
.
-1 .
-8 .
-7 .
-6 .
-5 ( . 6881) 434-438
( )2771 -:
- 1 .
- 2 .
- 3 .
- 1 .
- 8 ( . 6881 ) 6
525
)(ICoIT'17
10thof April, 2017.
: -:

( .
6882 ) 0 -:
-1

( Shale &Gomes , 1997 , p. 5 ( .
-2

( Garrison ,1997.p.8).
-3
(Feenberg,1999,p.337) .
4
( Johnson &others , 2000 , p. 5 ) .
-5
( ) 21 .
) ) Levin &others , 2001 , p.2
-6
(Twigg , 2003 , p.5 ) .
- 7
(Brown&others,2003,p.3) .
-8 .
)( Shachar&Neumann ,2003 , p.2
-: :
) ( James Taylor , 1995
(-: )1
( )1

-4
The corresesponde model
526
)(ICoIT'17
10thof April, 2017.
/ -: -6
-4
-6
-1
-1
-3
/ -: -1

/
the flexible learning model -1

/ -3
the intelligent flexible learning

, 6883 ) 00 - 03 (

:
-: ) ( Web/Computer-Based Training
:

) . ( Interactivity
( 6882 ) 41

) (Knowledge Transferring
:
.com www.bitlearning -1
www.mindleaders.com -2
)(Rovai , 2002 ,p.1 www.ajde.com -3
-:
) (Web/Electronic Performance Support Systems

.
527
)(ICoIT'17
10thof April, 2017.
( 6881 .) 2 -3

()
Spice land &others ,2002 , p.1) ( .
-: ) :(Web / Virtual Asynchronous Classrooms

) . ( Garrison & Anderson ,1999 ,p.51

) ( e-mail /
) (Chatting ) (forums (Rovai ,2002 ,p.1 ) .

) ( Black Board ) ( www.blackboard.com )( WebCT
) (www.webct.com ( ) Power Point ( ) Word art
( ) Photo Editor ( Levin &&others , 2001 , p.2 ).
-: ) :(Web / Virtual synchronous Classrooms

( ) ( Rovai , 2002 , p.2 ) .
) ( Feenberg , 1999 -:
-1 ) . ( Chatting Rooms
-2 ) . ( White Boards
-3 ) . ( Audio Conferencing
-1 ) . ( Video - Conferencing
-8 ) . ( Sharing into Programs
) ( Feenberg , 1999 , p. 340

on
line
( McDonald , 1998 , p. 244 ).
) ( The computer
-: :
528
)(ICoIT'17
10thof April, 2017.

.
) ( Preston ,1990

. ( Preston , 1990 ,ED 320 569 ) .

( 1663 ) 227

( . 4117 ) 71

.

.

.
. . (Anderson , 1998 ED421481 ),

( . 2777 .) 325

) ( Feed back ( . 2777 ) 211

) (Zeszotarski ,2000 ,ED438010 .

( 6886 )42 .
( ) 2771
- :
-1 .
-2 .
-3 .
-1 .
-3 ( . 6881 ) 7
529
)(ICoIT'17
10thof April, 2017.
( ) 2778

( .
2777 . ) 21

..
-: :
( )2777
-:
-1 .
-2 .
-3 .
-1 .
-8 .
-7 ) .( Multimedia
( 6882 ) 421-421
-: :
) ( Sherwood ,1983 ( ) 1667
-:
-1 ) :( Managing Learning With Computer

()Data Bases )(Library Retrieval Systems
-2 ) : (Learning About Computer

). ( Computer Literacy
-3 ) :( Learning About Thinking With Computer
) (Processes
.
-1 ) : ( Learning From Computer

.
-8 ) : ( Learning With Computer
)(Simulation
) (Educational Game .
530
)(ICoIT'17
10thof April, 2017.
) ( ( Sherwood ,1983 , ED241479 4112 ) 341 387

( )1668 ( : )1666
-:

-1 -:
.
-2 -:

- ) (Computer Managed Instruction : CMI

.
- ) : (Computer Assisted Instruction : CAI

.
) : ( Computer As Tutee -1
.
(4113 ()413-4144117)618-663
-: :
-4 ) : ( Drill & Practice

) (Godfriey &Sterling , 1982 , p.144 .

-:
-1 .

-2 .
-3 .
-1 .
).).Goktep & Others , 1989 , p.167

( .
4111 ) 78
) : ( Educational Game

.
531
)(ICoIT'17
10thof April, 2017.
) ( Ellington & Adinall , 1982 ,p.223

-3 ) : ( Dialogue

( . 4113 ( ) 117 4117
.) 667

( . 4111 ) 0
-1 ) : ( Tutorial

( . 4111 ) )74

( . 6886 )47
-3 ) : ( Problem Solving

.
) , .(Alessi & Trollip 1985 ,p.52

( .
4102 ) 616-614
-2 ) : (simulation

(Alessi & Trollip , 1985 ,p.199).

( . 4111 ) 1
-: ( ) :

.
) (Hannafin & Hooper ,1989
-:
532
)(ICoIT'17
10thof April, 2017.
-1 -:
.
-2 :
.
-3 -: .
) (Hannafin & Hooper , 1989 , p.155-156
( ) 2773 -:
-1
.
-2 .
-3 .
-1 .
-8
.
-7 .
-6 .
-5 ( . 6881 ) 48
) : ( The Internet
-:

.

) . (Tony ,1998 , p.1-3

(Haury & Milbourne , 1999 , ED 433218 ) .
) ( Williams ,1995
-:
-1 .
-2

.
533
)(ICoIT'17
10thof April, 2017.
-3 .
-1
( . 6888 .) 1

.
( 6884 . ) 16-14

)(Spector & Michael , 2001, ED456841

)(Maylene , 2003,p.1
..
( ) 1665
"

"( . 6883 ) 11

.
- : :

( )2777 ( )2771
-1 .
-2 .
-3 .
-1 .
-8 .
534
)(ICoIT'17
10thof April, 2017.
-7 .
-6 ) ( Intranet ) ( Internet
.
-5 .
-6 .
-17 ). (CD
-11 .
-12
.

-13 .
11 ) ( Classrooms Without Walls
.
( 6888 ( ) 1-4 6881 ( 1-6
-: :

( .
) 1666 -:
-1 .
-2 .
-3 ( . 4111 ) 3-1
( ) 2777 -:
-1 -:
.
-2 -:
.
-3
.
-1 -:
.
-8 .
-7 ( ( Lycos , yahoo , Google
.
-6
( . 6888 ) 42-41
( ) 2771
-:
-1 .
-2 .
535
)(ICoIT'17
10thof April, 2017.
%57 . -3
( ). -1
-8
.
( . 6881 ) 41-40 -7
: ) ( Motivation of Learning
) ( Christophel ,1990

.
) (Christophel , 1990 , p.323-340

) ( Borophy ,1987

( 6888 .) 1

( .
6883 ) 121
:
) (Weiner, 1980
-1 .
-2
.
-3 (Weiner, 1980 , p.19).
( )1652
-1
.
-2
. ( . 4106 )603
( )1651
-1 .
-2 .
-3 .
-1 ( . 4101 ) 410

536
)(ICoIT'17
10thof April, 2017.

( . 6888 ) 612-613
:

.

( Kloosterman & Gorman 1990, .
) .p. 375

.
) . ( Middlton & Spanias , 1999, p. 69

.
:
-1 :

( . 4101 ) 114

(
4116 . ) 461
( )
()2
( )6

-6 :

( 6884 .) 401
(
) ( ) 57 .
537
)(ICoIT'17
10thof April, 2017.

( )43
( )43
( .
4112 .) 41
( )37

( )18 . ( )18
.
()3
( )3
-3 -:

( . 1658 ) 211

.

-:
/ :
()1
()1

2 (*)
2
44 1 43
1.01 8.411
41 6 43
(*) 2 ( .) Yate's distribution
538
)(ICoIT'17
10thof April, 2017.
2 ) ( Chi square Independence Test 2

2 ( ) 7.78 ( ) 1
.
/

( ) 183.7
( ) 113.2
) ( One Way ANOVA () 7.78
( ) 25 1 F F

( . ) 8
()3

F F

044.6 044.6 4
1.6 8.16
4112.12 148747.14 60
/ :

( ) 11.5 ( 12.13
) ( ) 25 1 ( ) 7.78

( . ) 7
( ) 2

F F

8.00 8.00 4
1.6 8.84
24.30 4761.411 60
/ :
( ) 6
( ) 7

539
)(ICoIT'17
10thof April, 2017.

2 2

1 3 2
3.114 4.61
1 0 1
2 ( ) 2 ( ) 7.78
.
/ :

. 2 ( ) 1 (
) 7.78
( . ) 5
( ) 0

2 (*)
2

2 1
1.014 8.61
0 7
/ :
) ( Raven ) (Progressive Matrices
. ( )1653

( . 4101 .)11 -67
( )11.8
( )18.2 ( ) 25 1
( ) 7.78
(.)6
( ) 1

F F

1.0 1.0 4
1.6 8.82
73.74 6468 60
(*) 2 ( .) Yate's distribution
540
)(ICoIT'17
10thof April, 2017.
/ :

( ) 115.8
( 118.77
) ( ) 25 1 () 7.78

(.)17
( ) 48

F F

00.72 00.72 4
1.6 8.11
623.2 7112.201 60
- 5 :
/ :

-:
-1 ) .) Function
-2 ( .) Limit
-3 (. ) Continuity
-1 ( .) Differentiation
-8 ( .) Application of Differentiation
/ :

.
( )27

(
) .

.
/
" "
( 4110 ) 40
( ) 8
541
)(ICoIT'17
10thof April, 2017.

-:
-1 ) .) Function
-2 ( ) Limit .
-3 ( ) Continuity .
-1 ( ) Differentiation.
-8 Differentiation) Application of

( )

( 4101 . )12
(
)
( )
( )177
%57
.
( ) ()11
()44

11 48 41 43
43 7 6 2
7 6 1 6
61 2 0 1
42 0 1 3
488 11 18 17
/ :

-:
-1 .
-2
. ()1
542
)(ICoIT'17
10thof April, 2017.

(
)

( )
()4
.
( )12
.
( ) 46 ( )

7
27 ( )
71 ( )
12 ( )
16 ( )
38 ( )
163
( ) ()13
543
)(ICoIT'17
10thof April, 2017.
( )41 ( )

07 ( )
28 ( )
18 ( )
448 ( )
18 ( )
117
( :) 2

()2

544

)(ICoIT'17
10thof April, 2017.

)(n ()6 ()4

()1

545
)(ICoIT'17
10thof April, 2017.
( ) ( )772 ( )328
( ) ( )336
.
2 :
/ :

( )
- :

.
- :
( )

( ) 87

.
:
( )41

488 () 11 () 18 () 17

%488 %11 %18 %17
42 3 3 2 %11 0
9 1 3 1 %47 1
3 4 4 1 %0 6
41 1 1 3 %63 2
1 1 1 1 %47 1
38 42 16 18 %488 61
- :
( )87
(
) ( )87
546
)(ICoIT'17
10thof April, 2017.
-: ( )38 1 .
( 1667 .)68

. ( ) 11 - 37 , ( - 12
) 87 .
- :

.
:

( ) 87
- :

( ) 3

%56
- :
( )

( )87 (
)%26 ( ) ( )%26
( ) . ( )11
( )11
:
1 -:
( 0.357 ) 0.714

547
)(ICoIT'17
10thof April, 2017.
2 -:
( 7.386 ) 7.68
.
1 : (
)

1 :

( )67
.
3 :
.
-:
( ) %56
.
2 :
- :
) (Split Half Method .
( )
.
( ) 7.65
( ) . .
( ) 7.55
- :
( )Cronbach - Alpha
( . 4110 )133
( )7.61
.
/ :
( )
( )1666
548
)(ICoIT'17
10thof April, 2017.

:
- : ( )37 ( )21 ( )18
( :
) (37
)111
- :

%67
- : ( ) ( )87

( )%55
( 4101 )468 .
- 6 :
-1

-2 :

.
-3 :

( )
)

.
-1 :

.
549
)(ICoIT'17
10thof April, 2017.
- 8 :

.
) :
/
. -:

( ." )7.78 :
( )43
t
t
(*)

16.316 16 43
( )
4.784 4.016

14.67 17.11 43
()
( )12
( )36.63

( )1.532 ) ) t
( )1.671

.
. -:
"

(" . )7.78 :
()42
t
(*)
t

8.717 411.21 446.1 43
4.784 ( )
t 1.701 ( ) 28 . 0.05 (*)
t 1.701 ( ) 28 . 0.05 (*)
550
)(ICoIT'17
10thof April, 2017.

610.41 480.1 43
()
()446.1
( )480.1

( )8.717 ( ) t ()4.784

.
/ :
- :

-:
-1
( )
( )727

) ( E- books

( )Pentium - 1

) (Feed Back .
(
)Microsoft Office PowerPoint 2003
.

( )17

.

551
)(ICoIT'17
10thof April, 2017.

.
.
-2

.
) ( Calculus
) ( ON LINE

.

( )

.
( )
-:
-1

.
-2

.
-3

-1

.
552
)(ICoIT'17
10thof April, 2017.
-8

.
-7

.
-6

.
-5

.
-6

.
-17
TI-85 Graphing
Calculator Live Math Notebook
.

( )
() 1
-:
-1
( )17

.
-2

.
553
)(ICoIT'17
10thof April, 2017.
-3 (
)
.
-1
.
-7 ( )

( 2771 ( ) 57 1666 ( )61
1666 ( )81 2778 )Keller, 1983, P. 603( )373
.

:
-1

.
-2

.
-3 ) ( Tutorial

.
.
-1

-1
2773
" " -2
( )44-1
. 2772
-3
( ) ( )66-68
. 2773
-1
( .2771 )
-8
.2777
554
)(ICoIT'17
10thof April, 2017.
-7 .1666
-6 1 .2777
-5 1 3 . 2777
-6
( )2 . 2771
-17 .1658
-11
( .2771 )
-12 .1661
-13
1 2 . 1666
-11 .1656
-18
( . 1656 )
" -17
: " ( ) 7 0
. 2771
-16
. 1668
-15 " " ( )44-1
2771
-16
.1653
-27 " "
7 2 . 2771
-21
.2771
-22
( )44-1 . 2772
-23
( .1656 )
-21 .2771
-28 ( )4 . 1668
-27
. 2773
-26 " "
88 . 1668
-25 . 1667
-26
( .2778
)
-37 " "
61-67 .2777
1 .1666 -31
-32 1 . 2771
-33 "
. 2773
555
)(ICoIT'17
10thof April, 2017.
-31 ( ) WebCT
/
( . 2772 )
-38
2777
-37 "
" 6 1 .2777
-36 1653
-35
( )44-1 2771
-36 1 2777
-17
. 2771
-11 .1653
-12
.1658
-13 " " PC Magazine 1666
.
-11 "
" 61-67
.2777
-18 . 1668
-17 1 .1667
-16
( .1662 )
-15 1667
-16 "
( )
"
. 2773
-87 . 1656
-81
( .1666 )
-82 " " 13
( )3 , 1666
-83 1
.1666
-81 . 1651
-88 " ( : )"
44-1
. 2772
-87
.1653
-86
1656
-85 . 1667
-86 " "
17 . 2771
556
)(ICoIT'17
10thof April, 2017.
-77 7 1665
.
-71 . 2771
-72 3 2778
-73
( .1666 )
-71 " "
11.1666
-78 " " 26
61656
-77 " " ( ) 1/6 - 6/ 67
. 2773
-76 ( )
" 3 1
.2772
-75 1 .1655
-76 2777
.
-67 . 3
.1651
-61. 1 .1657
-62
. 2773
-63 :
21 . 1666
-61 " "
82 2.1663
-68 1 .1656
-67 1 .2777
-66 " " ( )
27 2 .1663
-65 " "
. 1665
-66 :
( ) 62 61 . 1665
-57 Microsoft Visual Basic 0.6 Professional
.1666
-51 " "
( )62-61
.1665
-52 .1652
-53
.2773
-51 " "
. 2778
-58 ( 6888 6888 ) 1
. 1666
557
)(ICoIT'17
10thof April, 2017.
-57 " "

6 . 2773
-56 . 1666
-55 " "
: ( ) 68- 41 . 2778
-56 " "
23 18.1667
-67 " : "
. 1666
-61 ( ( MATHEMATICA
( . 2771 )
-62 " " 12
. 1668
-63 " "
1 .1662
-61 " " ( -66
) 61 . 1666
-68
. 2771
-67 2771
.
-66
. 2777
-65
( .1666 )
-66 " "
( ) 26 2 . 2777
-177 1 .1666

101- Aiken , Lewis R , Psychological Testing And Assessment , 3rd , Allyn & Vacon Inc
,U.S.A , 1979 .
& 102- Alessi, Stephen M. & Trollip, Stanley, Computer Based Instruction (Methods
Development), Univ. of Iowa, 1985.
103- Anderson, Jeffrey, " Service - Learning and Teacher Education ", ERIC Digest, ERIC
Identifier (ED 421481), U.S.A , 1998.
104- Binder, C., " Behavioral Fluency A new Parading ",(in) Educational Technology, Vol.33,
Englewood Cliffs , U.S.A, 1993.
105- Brown , Gary, & Others , " Formal Course Design and the student Learning Experience
" Journal of Asynchronous Learning Networks , 7(3) , 2003 .
106- Costro , Paulo M.S, Internet in Mechanical Engineering Education & Practice ,
University of California , U.S.A , 1996 .
107- Chaplin, J.P., Dictionary of Psychology, 4th ed, Dell, New York, 1971.
108- Charp ,S. , " Internet Usage Education " , Technology Horizon in Education , ( The ) ,
27(10) , U.S.A , 2000.
109- Christophel, D. M. , The relationship among teacher immediacy behaviors, student
motivation, and learning ".(in) Communication Education, 39, 1990 .
558
(ICoIT'17)
10thof April, 2017.
110- Cooper, Hohn, Measurement and Analysis of Behavioral Techniques, Chio Charles, E.
Merill, Columbus, 1974.
111- Corl, S.F., " Novices on the net : An introduction to education class uses e-mail and the
Internet ", ERIC Digest, ERIC Identifier (ED 402998), U.S.A ,1996.
112- Cotton, Eileen Giuffre, " The online Classroom : Teaching with the Internet ", ERIC
Clearing House on Reading English & Communication, ERIC Identifier (ED 400577),
U.S.A , 1996.
113- Crawford ,C ., The Art of Computer Game Design , , Mc Graw- Hill , Berkely , 1984 .
114- Douglas , Masakoo , Individualized Learning ,Utilizing the Internet and JWPce
Computer Program: A Case Study of Heritage Japanese language learners , University
of California , U.S.A , 1999.
115- Ebel, R.L., Essentials of educational measurements, 2nd ed, Englewood Cliffs,
Prentice - Hall, New Jersey, 1972
116- Ehrmann , Stephen C. , " Asking the Right Question : What Does Research Tell Us
Technology and Higher Learning? ",(in)Change,(27) 2, 1997.
117- Ellington ,H. and Adinall , E, A Handbook of Game Design , London, Kogan Page ,
1982 .
118- Favretto ,Giuseppe &others , "E-learning measurement of the learning differences
between traditional lessons and online lessons " , European Journal of Open, Distance
and E-Learning , no.(2) ,on-line ,2005.
119- Feenberg , A. , "Reflections on the Distance Learning Controversy" , Canadian Journal
of Communication , no.24 , 1999 .
120- Finney ,Ross L., Thomas , George B. , Calculus , Addison-Wesley Publishing Company
, Vol. ( 1) ,U.S.A, 1989 .
121- Flowers , Jim , " Online Learning Needs in Technology Education " , Journal of
Technology Education , 13(1) , Fall , 2001 .( on-line )
122- Freund, John.e, Modern Elementary Statistics, 5th ed, Prentice / Hall International
editions, Englewood Cliffs, New Jersey, U.S.A, 1979.
123- Gaeddert ,Terri J. , Using Accelerated Math. To Enhance student Achievement in high
School mathematics Courses , Friends University , U.S.A , 2001 .
124- Garrison ,D. R. , "Computer Conferencing : The Post Industrial age of Distance
Education " , (in) Open Learning ,12( 2 ) , 1997 .
125- Garrison, D. R., & Anderson, T., " Avoiding the industrialization of research universities:
Big and little distance education" , The American Journal of Distance Education, 13(2) ,
1999 .
126- Godfrey , D.& Sterling ,S.; The Elements of CAL , , Reston Publishing Co. Reston
,1982 .
127- Goktepe, Must & Others, " Design And Implementation of a Tool For Teaching and
Programming ", Computer And Education, Vol. 13, No. 2, 1989.
128- Good, G.V., Dictionary of Education, 3rd ed, McGraw - Hill, New York, 1973.
129- Hannafin, M. J & Hooper, " An integrated framework for CBI S.screen design and layout
". Computer in Human Behavior , 5(3) ,1989 .
130- Haury, David L. " Milbourne, Linda A. Using the Internet To Enrich Science Teaching
and Learning " , ERIC Digest , ERIC Identifier (ED433218),U.S.A , 1999 .
131- Heo ,Misook & Chow, Anthony , "The Impact of Computer Augmented Online
Learning and Assessment Tool " , Educational Technology & Society , 8(1) , 2005 . ( on-
line )
559
(ICoIT'17)
10thof April, 2017.
132- Hessemer, S.J., " The Effect of Computer Assisted Instruction on Motivation and
Achievement in Fourth - Grad Mathematics ", Dissertation Abstracts International, Vol.
47, No. 10, April, 1986.
133- Johnson, Dana T., " Teaching Mathematics to Gifted Students in a

Mixed - Ability Classroom ", ERIC Digest, ERIC Identifier (ED 441302), U.S.A ,2000.
134- Johnson, Scott D.,& others , " Comparative Analysis of Learner Satisfaction and Learning
Outcomes in Online and Face to Face Learning Environments. " , Journal of Interactive
Learning Research , 11(1) , 2000 .
135- Keisler , H. Jerome , Elementary Calculus An Infinitesimal Approach , Second Edition
,University of Wisconsin , U.S.A , 2005.
136- Keller, Bonnie & Reigeluth, Charls, A Comparison of Three Instruction
presentation Format, Syracuss Univ. School of Education, No. 6, New York, 1982.
137- Kloosterman P., & Gorman J., " Building Motivation in the elementary mathematics
classroom ", School Science and Mathematics, Vol. 90, U.S.A., 1990.
138- Kosakowski, John, " The Benefits of Information Technology", ERIC Digest, U.S.A ,
EDO - IR - 98 - 04, 1998.
139- Kuhn ,Tomas, " Internet Teaching :Introduction , (in) EESE ,Vol.5 ,Berlin , 2001.
140- Lawson, Michael J., & Chinnappan, M., " Knowledge Connectedness in Geometry
Problem Solving ", JRME Online, Volume 31, Issue 1, January, 2000.
141- Levin , Sandra R. &others " Highly Interactive and Effective Online Learning
Environments for Teacher Professional Development. International Journal of
Educational Technology " , 2(2) ,2001.
142- Maylene Y., Damoense , " Online learning: Implications for effective learning for higher
education in South Africa " , Australian Journal of Educational Technology , 19(1), 2003
. ( on-line )
143- McDonald.. " Interpersonal group dynamics and development in computer conferencing
:The rest of the story". In Proceedings of the Annul conference on Distance Teaching and
Learning, Madison, Wisconson. 1998 .
144- McMinn , N. &Butler S. , " Linking Assessment and Motivation in the Classroom
Environment " , (in) Learning Consortium Conference , Retrieved Feb. 9 , 2003 .
145- Middleton, J.A., & Spanias, P.A., " Motivation for Achievement in Mathematics: Finding,
Generalizations, and Criticisms of the Research ", JRME Online, Volume 30, Issue 1,
January, U.S.A., 1999.
146- Moore , M.G. , " Editorial : Distance Educational Theory " ,The American Journal of
Distance Education , 5(3) , U.S.A ,1990 .
147- Morgan, Nancy A. - Sprague, Carolyn , " An Introduction to Internet Resources for K-12
Educators. Part II: Question Answering, Electronic Discussion Groups, Newsgroups,
Update 2000 " , ERIC Digest , ERIC Identifier (ED 438808) ,U.S.A , 2000 .
148- Preston, Nancy R., " Trends and Issue in Educational Technology: 1989, ERIC Digest,
ERIC Identifier (ED 320569), U.S.A ,1990.
149- Roblyer, M.D., " The Impact of Microcomputer - Based Instruction on Teaching and
Learning: A review of Recent Research ", ERIC Digest, ERIC Identifier (ED 315063), U.S.A
,1989.
560
(ICoIT'17)
10thof April, 2017.
150- Rovai ,Alfred P. , " A Preliminary Look at the Structural Differences of Higher Education
Classroom Communities in Traditional and ALN Courses " , Journal of Asynchronous
Learning Networks , 6(1) , 2002 .
151- Shachar , Mickey & Neumann ,Yoram , " Differences Between Traditional and Distance
Education Academic Performances :A meta- Analytic Approach " , International Review of
Research in Open and Distance Learning , 4(2) , 2003 .
152- Shale , D. &Gomes , J. , " Performance Indicators and University Distance Education
Providers " , Journal of Distance Education ,13(1) ,1997.
153- Shiratuddin , Norshuhada , "Internet Instructional Method: Effects on Students
Performance" , Educational Technology & Society , 4(3) , 2001 . ( on-line )
154- Sherwood, Robert D., " Computers in the Schools: Methods of Utilization and Action
Recommendations ", Thesis at New York University, ERIC Digest, ERIC Identifier (Ed
24147), 1983.
155- Simonson ,M.S. , &others , Teaching and Learning at a Distance : Foundation of

Distance Education , Upper Saddle River, NJ., Merrill &Prentice Hall , 2000 .
156- Sparks , John C., Calculus Without Limits , 2nd Edition, Author House, , U.S.A , 2005
.
157- Spector, J. Michael - de la Teja, Ileana , " Competencies for Online Teaching " , ERIC
Digest , ERIC Identifier (ED 456841),U.S.A , 1999
158- Spiceland, J. David & Hawkins, Charlene P. " The Impact on Learning of an
Asynchronous Active Learning Course Format ", Journal of Asynchronous Learning
Networks , 6(1) ,2002
159- Tennant, Roy , " Internet Basics: Update 1996 " , ERIC Digest , ERIC Identifier (ED
392466),U.S.A , 1996 .
160- Tony, Greening , " WWW support of student learning: A case study " , Australian
Journal of Educational Technology ,14(1) , 1998. ( on-line )
161- Twigg ,Carol A. , " Improving Learning and Reducing Costs : New Models for Online
Learning " , (in) Educause Review , 38(5) , 2003 .
162- University of Akron, E-Calculus, University of Akron , D. P .Story , ,U.S.A ,2005 . ( on

line )
163- USDLA,US American Distance Education Association, www.usdla.org , on-line , 2004.
164- Valentine ,Doug , " Distance Learning : Promises , Problems , and Possibilities " , Online
Journal of Distance Learning Administration , 4(2) , Fall , State University of West
Georgia , Distance Education Center ,2002 .
165- Wegner ,Scott B. &others , " The Effects of a computer- Based Instruction Management
System on Student communications in distance learning " , Educational Technology &
Society , 2(4) ,1999.( on-line )
166- Weiner, Bernard, Human Motivation, Rienchart & Winstom, New York, 1980.
167- Zeszotarski, Paula, " Computer Literacy for Community College Students ", ERIC Digest,
ERIC Identifier (ED 438010),U.S.A , 2000.
561
(ICoIT'17)
10thof April, 2017.
ABSTRACT
Distance education is one of the most common educational patterns in our world today,
especially in the field of university education, in order to provide higher education
opportunities to those who missed these opportunities due to economic and social conditions.
The aim of this research is to know the effect of using two methods of distance learning
(computer and internet) on achievement and motivation. The research community determines
the first grade students in the mathematics department at the Open Education Center (the
Open Study Center in Baghdad), which consists of (80) students divided into four study
groups. The researcher randomly selected 30 students randomly divided into two
experimental groups .a number of educational sites that teach calculus on the Internet has
been determined. Acomputer educational program has been prepared, designed and
implemented. An achievement test of subjective type was prepared and a measure of
motivation for mathematics has been used. The research results in the superiority of the first
experimental group that studied using the computer on the second experimental group that
studied using the Internet in achievement. There were no statistically significant differences
between the two research groups in the motivation measure of learning towards mathematics.
The research found that the use of the computer tutorial was better than the Internet in student
learning and had a positive impact on their achievement.
562

ICoIT17 Proceeding PDF

Uploaded by

Document Information

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

ICoIT17 Proceeding PDF

Uploaded by

Copyright:

Available Formats

The 1st International Conference

Lebanese French University

Once Steve Jobs said Lets go to invent tomorrow

The organizing committee is delighted to warmly welcome you to the

The conference committee received 112 research papers from 29 universities

The International Program Committee (peer reviewers) was made of 41

By the end of this process, the conference committee accepted 49 papers,

Thank you very much for your time and presence

ICoIT17 Conference Chair

Logistic Committee Chair : Dr. Yahya Tareq Hussein

Social Activities Committee Chair : Abdulsatar Abduljabbar

Publishing & Media Committee Chair : Sarah Y. Yousif

Financial Committee : Firas N. Mardan

Secretariat : Dina Khalid Khudher

Dr. Sarina Muhamad Noor Dr. Halgurd Sarhang

Dr. Huda K Aljobori Dr. Selcuk Cankurt

Dr. Akar Taher Dr. Yahya Tareq Hussein

Mr. Safwan Mawlood Mr. Mohammed Bajalan

Mrs. Mardin Anwar

Abduladhem Ali, PhD University of Basrah Iraq

Asst. Prof. Dr. Ali Qatan Dr. Amer S. Elameer

Dr. Kayhan Zrar Ghafoor Dr. Saman Mamyahia

A Comparative Analysis of Texture Methods for Visual Object Categorization

A Hybrid Neur-Fuzzy Optimization Method for PV Steady State Improvement in

A Multi Secret Sharing Approach for Vulnerability Identification in Social Media

A PAPER SURVEY ON THE IMPACT OF (IT) ON THE QUALITY OF FINANCIAL

A PAPER SURVEY ON THE IMPLEMENTATION OF THE PARALLEL FDTD ON

A Proposed Public Key Encryption Based on Hybrid Chaotic Maps

A Proposed Non Feistel Block Cipher Algorithm

A Survey On Unsupervised Evaluation Criteria For Image Clustering Validation

Alarm System using Microcontroller , GPS and GSM Devices

Design and Development of a MOOC for Academic Institution in Iraq

Design and Execute a Complete Vehicle System for the MOHEDSR

Design and Implementation of a Rehabilitation Wireless Networked System

Distance Varieties Wireless Signals Relations Connected Peer to Peer Computers

Distributed Data Aggregation protocol for improving lifetime of Wireless Sensor

DSP for Mobile Wireless LTE System: A Review

Effect of Doppler Shift frequency on the performance of 2x2 OSTBC-OFDM System

Energy-Efficient Sensor Activity Scheduling Protocol for Wireless Sensor Networks

E-Service Quality Factors Impacting Customers Purchase Retention in E-Retailing in

Establishing Educational Robotics Labs in Iraqi Universities

Image Encryption Based on Partitioning and Rearranged Pixels Position

Image Enhancement By Using Homomorphic Filtering Model

Implementation of an Economic Light Duty Three-Axis Computer Numerical Control

Intelligent Fault Recovery Controller for Power Generator at Perdawd CCGS in

Internet of Things (IoT): Readme

Internet of Things Based Cloud Smart Monitoring for Asthma Patient

MARKOV-BASED DEPLOYMENT APPROACH TO IMPROVE WSN COVERAGE

NEURAL FUZZY PETRI NET BASED ARABIC PHONEME CLASSIFIER WITH

New Weights in Laplacian Smoothing on Triangular Mesh

Proposed Secret Authentication Method for E-government Purposes

Security Issues in Social Media: Challenges and Solutions

Survey and New Security methodology of Routing Protocol in AD-Hoc Network

The Estimation of Wind Velocity Using Data Mining Techniques

Unbreakable Digital Watermarking Technique

Using RSA Digital Signature to Solve Pollution Problem in Single-Source Network

Virtualizing Network Services: Practical Perspectives

5G: MIMO BDMA SYSTEM TRENDS AND TECHNOLOGY

ARTICLE INFO ABSTRACT

Figure 1: Beam Division Multiple Access [4]

Figure 2: Feature of 5G Technology [6]

3. Analysis Methods of 5G technology.

3.1 5G millimeter wave technology.