Realizing 5G: Device-Centric Design

in a New Spectral Landscape

IEEE ComSoc Content Sponsored by

A spectrum perspective on 5G

Dr. Federico Boccardi

3rd December, 2015

Emerging services
Broadband++
high throughput, consistent QoE

M2M
low cost, low battery consumption
Source: modified from ITU

Critical
communications
low latency, high reliability
SPG
Technology
2

Enabling technologies
Broadband++
high throughput, consistent QoE

5G
/802.11xx

LTE /802.11ac
satellites
GSM/
LTE Cat 0/
GERAN/
5G/ Wi-Fi/Sigfox/
[]
Bluetooth/ ZigBee/

M2M
low cost, low battery consumption

5G
802.11p
LTE D2D / LTE-V

Critical
communications
low latency, high reliability
SPG
Technology
3

Broadband++

Licensed spectrum below 6 GHz

900
800
700

1800

2100

1.4GHz

2.6GHz
2.3GHz

3.4GHz

>3.6GHz

Total downlink (scenarios)

Today

Existing mobile bands

2016

Addition of 1452-1492 MHz, 2.3 GHz, 3.4 GHz

2017-2022

Addition of 700 MHz, potentially 1427-1452 MHz, 3.6-3.8 GHz , 1492-1518 MHz

Licence-exempt spectrum at 5 GHz

5150 MHz

5350 MHz

5470 MHz

5725 MHz

5850 MHz

5925 MHz

5010 MHz

Today: indoor and

power limited.
Different constraints
apply in different
parts of the world.
Studies on various
constraints on-going.

Today: not
available.
Studies on
interference
mitigation
methods ongoing.

Already available
indoor and
outdoor.
Now possible to
study and reconsider
constraints.

Already used for WLAN

Today:
available for
fixed
broadband in
UK.
Studies for
extension to
WLAN in
progress.

Today:
available for
ITS.
Studies for
compatibility
for WLAN
are ongoing.

Not available for WLAN, mitigation methods studies agreed for WRC 19
Not available for WLAN, not in scope of studies
6

The emergence of hybrid spectrum access schemes

LAA is a hybrid approach that builds

on both licensed and licence-exempt
spectrum

What is the impact on policy in the

long term?
What are the opportunities and the
risks for consumers?
What new business models will
emerge in the markets?
What will be the impact on bands
other than 5 GHz?

Source: Ericsson

Mm-Wave: technical assessment

6GHz

100GHz

Prop. loss in free space (downlink)

Prop. loss in free space (uplink)

Prop. loss in free space (sounding)
Real-world propagation effects
Intra-cell and inter-cell interference/
opportunities for sharing
Antenna dimensions
Beamforming and MIMO solutions
Transceiver design

Studies agreed in WRC 15

The full set of bands for study is

24.25-27.5 GHz
31.8-33.4 GHz
37-43.5 GHz
45.5-50.2 GHz
50.4-52.6 GHz
66-76 GHz
81-86 GHz

Will mm-Wave be beneficial for a few lucky ones?

10

M2M

11

Spectrum available for M2M connectivity

12

M2M: what can we learn from other markets?

The emergence of a dominant design
for cars

M2M: connectivity platform

?
GERAN

M2M: service enabling platform

?
13

M2M: what can we learn from other markets?

Technology adoption is getting faster and faster

Once a dominant design will emerge, the IoT market could grow very rapidly

Spectrum takes long time to be harmonised

14

M2M: whats next?

Technology moves very fast

NB-IoT on its way
Do we need more spectrum for IoT
in the long term?

Do we need different spectrum?

How much and how rapidly will the
M2M market grow?

15

Critical communications

16

Critical communications

Do we need new spectrum for critical

communications?
Where and how much?
With which access arrangement?

17

Conclusions

Its important to ensure that lack of spectrum will not inhibit the support
of current services and the deployment of new services
New studies for WRC 19 have been agreed
Its not fully clear how much and what type of spectrum are required to
support emerging services, like IoT and critical communications

18

Realizing 5G: Device-Centric Design in a New Spectral

Landscape
Dr. Amitava Ghosh
Technology and Innovation

Nokia Networks
3rd December, 2015

19

Nokia 2015

Heterogeneous use cases diverse requirements

100 Mbps
whenever needed

>10 Gbps
peak data rates

Extreme
Mobile
Broadband

10 000
x more traffic

10-100
x more devices

<1 ms
radio latency

M2M
ultra low cost

10 years
on battery

20

Nokia 2015

Massive
machine
communication

Critical
machine
communication

Ultra
reliability

Unlocking new spectrum assets | The Foundation for 5G

Leveraging all bands, ranging from ~400MHz - 100GHz

400 MHz

3 GHz 6 GHz

10 GHz

30 GHz

90 GHz

mmWave

cmWave

Lower
frequencies
translate into
continuous
coverage for high
mobility and
reliability cases

Higher
frequencies
translate into
higher capacity
and massive
throughput

Different characteristics, licensing, sharing and usage schemes

Carrier BW

n*20MHz

Leading channel
modeling know-how

Duplexing

Channel measurements
from 3-73GHz

Leading
METIS I & II
spectrum
work package

Cell size
Coverage

Capacity

Worlds 1st
Wide Area Single
Frequency Network
trial in UHF band

10,000 x

21

Nokia 2015

>10 Gbps

100 Mbps

<1 ms

10-100 x

ultra low

10 years

Worlds 1st
Licensed Shared
Access demos/trial

5G Scalable air interface design across frequency bands

Expanding the spectrum assets to deliver capacity and experience

Spectrum availability
LOS

Spectrum

Antenna size

Network layer

~Nx1GHz

Very small

Ultra high
capacity and
data rates

90 GHz

carrier bandwidth
Dynamic TDD

Low rank MIMO and

beamforming

~Nx100 MHz

Small

carrier bandwidth

High Rank MIMO and

Dynamic TDD

beamforming

~Nx10 MHz

Medium large

30 GHz

10 GHz

6GHz
3 GHz
10 cm

carrier bandwidth
FDD and TDD

High Rank MIMO and

beamforming

Cell size
LOS/NLOS

22

Nokia 2015

300 MHz
1m

Boosting
capacity and
data rates

Providing base
coverage and
capacity

Channel Measurements and Modeling

23

Nokia 2015

Propagation for 5G (less than 100 GHz)

Path loss increases with frequency
- However, wavelength decreases with frequency and hence larger antenna arrays are
possible at higher frequencies in the same area thus mitigating the larger path loss
Diffraction (e.g., the bending of rays around building corners/roofs) loss
increases with frequency
- No longer a dominant effect after around 10 GHz in outdoor channels
Oxygen/rain losses

- Are frequency dependent but are small (less than around 2.0 dB for worst-case rain)
for cells less than 200 m
Reflections appear relatively constant for all frequencies below 100 GHz
Scattering increases with frequency, but current measurements are not showing a
significant impact at least up to around 73 GHz
Penetration loss tends to increase with frequency
- However, it is heavily material dependent and certain materials allow even higher frequencies to
pass through without much attenuation (e.g., standard glass)

24

Nokia 2015

5G Outdoor and Outdoor-to-Indoor Priority Environments

Urban Micro (UMi): mix of type urban
scenarios
- Street canyon and open square both
separate UMi environments
- Cell radii typically less than 100 m
- Access points below rooftops (3-20 m high)
- Outdoor-to-outdoor and outdoor-to-indoor
(UEs from 1.5-22.5 m)
Urban Macro (UMa)
- Access points on or above rooftops (25-35
m high), cell radii>200 m
- Outdoor-to-outdoor and outdoor-to-indoor
(UEs from 1.5-22.5 m)
- Might be viable up to lower mmWave
frequencies

25

Nokia 2015

5G Indoor Priority Environments

High outdoor-to-indoor penetration losses
at higher frequencies may mean a
separate deployment indoors
- Penetration loss highly material
dependent and may still be reasonably
low (<~20 dB) even at 73 GHz for some
residential homes
Indoor office
- Desk/cube/offices
- Base 2-3 m high (wall or ceiling)
- Both open and closed offices separately
modeled
Shopping malls
- 3-5 stories, open area in middle
- Base (3 m high) on walls or ceilings of
corridors and shops

26

Nokia 2015

Channel Model Study Procedures : Summary

Variety of environments
- Urban Micro (Street Canyons, Open Squares, Stadiums), Urban Macro
- Indoor Environments (Open and closed offices, Shopping Malls, Trains Stations, Homes..)

Step 1:
Measurements /
simulations and data
analysis

Step 2:
Study on
propagation
phenomena

Step 3:
Establishment of
baseline model

Step 4: Initial
parameterization of
the baseline model

e.g.,
e.g.,
e.g.,
Where and how dominant Whether 3D channel
Parameter optimization
path travels
model can be reused
Group processed data
What impacts propagation What extension is
together and have single
property
necessary on 3D channel parameter set
How they can be modeled
model
Higher frequency specific
propagation property

27

Nokia 2015

5G mmWave Overview

28

Nokia 2015

5G mmWave Challenges & Proof Points

Unique difficulties that a mmWave system must overcome
- Increase path loss which is overcome by large arrays (e.g., 4x4 or 8x8)

- Narrow beamwidths, provided by these high dimension arrays

- High penetration loss and diminished diffraction

Two of the main difficulties are:

- Acquiring and tracking user devices within the coverage area of base station using a
narrow beam antenna

- Mitigating shadowing with base station diversity and rapidly rerouting around
obstacles when user device is shadowed by an opaque obstacle in its path

Other 5G aspects a mmWave system will need to address:

- High peak rates and cell edge rates (up to 10 Gbps peak, 100 Mbps cell edge)
- Low-latency (< 1ms)

29

Nokia 2015

Rapid Rerouting Architecture and Protocol

Stand-by
AP

Serving AP

X5
CSM

AP1

AP2

X5

Data Path

Stand-by
AP

UD

Alt. Data Path

AP3
Rapid Rerouting concept

Radio link to the serving AP may be frequently blocked due to obstacles,

e.g. moving trucks, pedestrians, trees, user rotation

AP1 (serving AP)

UD

On the average, a blockage may happen every 0.75 to ~10 sec,

Rapid rerouting of the radio link in the event of a blockage

A user device (UD) is served by a cluster of APs, its Cluster Set (CS)

Access Point Custer Set and Cluster Set Manager

APs in the CS are selected based on accessibility

Cluster Set Manager (CSM): A per-user logical entity that manages the APs in the
cluster set

DL alloc

X
X

DL
data

DTX detection

FastACK

Rerouting request
for UD

Rapid Rerouting: Fast radio link blockage detection followed by fast handoff

Fast radio link blockage detection during data transfer

Fast blockage detection at UD is challenging due to absence of any

efficient broadcast channel

Solution: FastACK channel carries an ACK for DL control messages

AP determines blockage on detection of DTX on FastACK

Fast handoff:

On detection of blockage, serving AP initiates handoff via CSM

UD detects blockage on receiving handoff command from stand-by AP in

its CS

DL ACK

Handoff command + DL alloc

DL data
FastACK
DL ACK

Rapid Rerouting protocol

30

Nokia 2015

CSM

AP2

Rerouting
command for UD

5G mmWave Frame Structure

20 ms superframe

Analog beamforming has implications for the

modulation format used on the mmWave link
-

Beamforming weights are wideband and, for OFDM, all

subcarriers within an OFDM symbol must share the same beam

Time division multiplexing (TDM) is favored over frequency

division multiplexing (FDM)

TDD Frame 500 ms

TDM suggests low PAPR modulation techniques can be

considered to reduce the PA backoff and maximize the
transmission power

PAPR is further reduced using /2 shifting of BPSK, /4 shifting

of QPSK

The QAM symbols are grouped into blocks of N=1024

symbols
-

150 blocks per 100 ms slot

10 blocks control (yellow portion), 2 pilot blocks, 138 data blocks

Control

Data

TDM Slot 100 ms

The modulation format is called Null Cyclic Prefix Single

Carrier (NCP-SC)
-

ND= 960 and Ncp = 64 provides 42 ns delay spread resilience

The null cyclic prefix can be increased or decreased on a per slot

basis without impacting the overall system numerology

RF beams can be switched during null period without destroying

the CP property

MCS
LEVEL

Data Rate* (2
stream)

1/5 BPSK

0.530 Gbps

QPSK

2.65 Gbps

1/2 16QAM

5.30 Gbps

5/6 16QAM

8.83 Gbps

3/4 64QAM

11.92 Gbps

The mmWave link utilizes single carrier modulation to

maintain a low. PAPR
-

37 38 39 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 0 1 2

ND=N-NCP Data Symbols

A commercial system is envisioned to use a 1-2 GHz

bandwidth
-

NCP null symbols

Achieves over 10 Gbps peak rate with 2x2 MIMO

Length N Symbol Block

NCP-SC Symbol Block

31

Nokia 2015

*Assumes dynamic TDD where entire slot can be UL or DL

mmWave Massive MIMO/Beamforming Solution

Power consumption is one critical aspect for

mmWave deployments
-

ADCs capable of sampling a 2 GHz BW signal will be

a major factor in power consumption

Full digital baseband transceiver behind each element

would consume an unacceptable amount of power

Analog (aka RF-radio frequency) beamforming

techniques will be employed to steer the array
elements on the panel

The antenna panel would host a highly integrated

mmWave circuit
-

Array of patch antenna elements bonded to an

antenna distribution layer with power amplifiers, low
noise amplifiers and phase shifters
Signal summed and down converted on the die and
mixed down to where it could be generated or
sampled by DAC and ADC
A separate antenna panel would be used for each
orthogonal polarization

RFIC Die signal distribution

Antenna
distribution layer

Baseband
Distribution layer
ADC

RF

DAC

xN

LO & PWR
distribution layer

32

Nokia 2015

LO

2x2 RFIC

mmWave 5G requirements can be met even in challenging environments

Performance in outdoor
environments
Enabled through
flexible backhaul
RFIC/antenna integration

75 AP/km2

150 AP/km2

187 AP/km2
AP density

2.1 Gbps

4.1 Gbps

5.1 Gbps

Average UE Throughput

Average UE Throughput

Average UE Throughput

<1 Mbps

222 Mbps

552 Mbps

Edge Throughput

Edge Throughput

Edge Throughput
Network capacity

16.4%

3.2%

1%

Outage Probability*

Outage Probability*

Outage Probability*

Multi-connectivity

*Outage defined as % of UEs with <100 Mbps

33

Nokia 2015

Massive MIMO:
An Essential Technology Component for 5G

34

Nokia 2015

What is Massive MIMO?

(M-1,N-1)

(M-1,0) (M-1,1)

Massive MIMO is the extension of

traditional MIMO technology to antenna
arrays having a large number of controllable
antennas

MIMO = Multiple Input Multiple Output = any

transmission scheme involving multiple transmit
and multiple receive antennas
- Encompasses all implementations:
RF/Baseband/Hybrid
- Encompasses all TX/RX processing methodologies:
Diversity, Beamforming, Spatial multiplexing,
SU & MU, joint/coordinated transmission/reception, etc.

Massive Large number: >> 8

Controllable antennas: antennas (whether physical or
otherwise) whose signals are adaptable by the PHY
layer (e.g., via gain/phase control)

35

Nokia 2015

(1,0)

(1,1)

(1,N-1)

(0,0)

(0,1)

(0,N-1)

Why Massive MIMO

Benefits:
- Enhance Coverage High gain adaptive
beamforming
Focus energy more towards the user, increase SINR

- Enhance Capacity High order spatial multiplexing

Multiple parallel spatial streams to a single user (SU) or to
multiple users (MU)

Relevance to 5G:
- Lower operating frequencies (e.g., <6GHz) are more
interference limited
LTE already designed for high spectral efficiency (<8 Antenna
ports)
Capacity-enhancing solutions become essential

- Higher operating frequencies (e.g., >>6GHz) have poor

path loss conditions
Coverage-enhancing solutions become essential
36

Nokia 2015

Signal Processing View: Fully Connected Arrays

RF

aQ
37

Nokia 2015

Stream B

TX-B

a1

Stream 1

a2

Stream 2
ooo

ooo

aQ

Stream K

a1
Stream 1

Multi-Beam
BF
K Beams
Q Antennas

Multi-Beam BF
B Beams
Q Antennas

TX-2

TX-B

Stream 1

Stream 1

v1

Stream 2

v2

Multi-Beam BF
K Beams
Q Antennas

v2,2
v2,K

Stream K

vQ,1

vQ

a2

v2,1
Stream K

aQ

v1,2
v1,K

Stream 2

a1

v1,1

a2

vQ,2

vQ,K

a2
aQ

a1

TX-1

ooo

BF
1 Beam
Q Ants

Multi-Beam BF
B Beams
Q Antennas

TX-2

a2
ooo

Legend:

Stream 2

TX-B

ooo

a1

TX-1

Multi-Beam BF
B Beams
Q Antennas

TX-2

ooo

aQ

BF
1 Beam
B Ants

ooo

TX-Q

a2

Stream 1

ooo

TX-2

a1

Stream 1

aQ

ooo

ooo

Stream K

Multi-Beam
BF
K Beams
Q Antennas

TX-1

aQ

TX-1

a2

BF
1 Beam
Q Ants

ooo

Multi-Stream

Stream 2

Stream 1

a1

TX-1

ooo

a2

TX-2

TX-Q

Stream 1

a1

ooo

BF
1 Stream
Q Ants

TX weights applied at both RF and baseband

ooo

a1

TX-1
Stream 1

Frequency non-selective weights applied at RF

(e.g., via analog phase shifters)

ooo

Single Stream

Frequency selective weights applied at baseband

(e.g., BF weights applied to OFDM subcarriers)

Hybrid

aQ

a2
ooo

Baseband

aQ

Phased Array Technology

Basic technologies vs. band of operation
3.5 GHz

15 GHz

28 GHz

38 GHz

60 GHz

73 GHz

83 GHz

94 GHz

Waveleng
th

m
m

86

20

11

7.9

5.0

4.1

3.6

3.2

Row/colu
mn

Total

64

64

64

64

64

64

64

64

Width/Hei
ght

m
m

342.9

80.0

42.9

31.6

20.0

16.4

14.5

12.8

T/R
Module
using
Mech
array
assembly

Monolith
ic T/R
Modules
on
Interpos
er

1 or more
MMIC on
Interposer
board

Multiple
MMICs ,
chipscale
antenna
or
interpose
r

Technolog
y

Migrate to
MMIC as
PA, LNA, phase shifter, frequency
VGA and T/R diplexing increases to
mechanically assemble reduce cost
into phased array.
and improve
MMIC solutions
manufacture
preferred

38

Nokia 2015

T/R
T/R
Modules Modules
or MMIC or MMIC
on
on
Transition region where
Interpos
either scalableInterpos
MMIC or
T/R
ermodule approach
er
may be viable

* MMIC = Monolithic Microwave Integrated Circuit

Silicon Image
60GHz MMIC on
LTCC interposer
board with
antenna array

Multiple
Multiple
MMICs ,
MMICs
chip-scale
using
antenna
chip
or
scale
Circuits
same
interposer
antenna

Transition region for

interposer board vs.
chip-scale
antennas?

size as antenna
array.(UCSD
94GHz Chip
Scale Ant array)

Trends for MIMO/BF in 4G & 5G as BW & Carrier Frequency

Increases
< 6 GHz/low BW

6-55 GHz/moderate BW

>55 GHz/high BW

Bandwidth Limited

Huge Bandwidths

Interference Limited

Noise Limited

Emphasis on Spectral
Efficiency

Emphasis on Gain

Per-antenna channel
knowledge

Per-beam channel
knowledge

Baseband Architectures

Hybrid / RF Architectures

Small Scale Arrays:

SU-MIMO sufficient
Large Scale Arrays:
high-order MU-MIMO

Large Scale Arrays are

required with an initial
emphasis on SU-MIMO

39

Nokia 2015

5G Standards

40

Nokia 2015

3GPP and ITU-R process geared for Full 5G meeting the needs of 2030
3GPP RAN WS on 5G agreed on Phasing

Phase 1

Phase 2

Initial 5G Features

Additional 5G Features

Catering for the immediate commercial

requirements

Catering for all the identified long-term

commercial requirements and meeting all
the ITU-R IMT-2020 requirements

Specification completed in 2H2018

Specification completed end of 2019

Meeting the needs of 2020+

Ready for the needs of 2030 and beyond

http://www.3gpp.org/ftp/workshop/2015-09-17_18_RAN_5G/Docs/RWS-150073.zip

41

Nokia 2015

3GPP release schedule and 5G work phasing plan in RAN

2015

2016

Rel-13

2017

Rel-14

2018

Rel-15

2019

2020

Rel-16

Requirements SI
3GPP Rel-12
Freezing
Content: 09/14
ASN.1: 03/15

Technology SI(s)
>6 GHz Chmodel SI

Phase 2 5G WI(s)
Phase 1
5G WI(s)

http://www.3gpp.org/ftp/workshop/2015-09-17_18_RAN_5G/Docs/RWS-150073.zip

42

Nokia 2015

5G PoC Systems

43

Nokia 2015

Key milestones on the road to 5G

5G PPP projects
started

3GPP 5G work in full

progress

Positive outcome at
WRC2015

IMT2020 evaluation
process in ITU-R

WRC19 outcome
clear with new
bands for IMT
5G phase 2 specs
ready

3GPP
standardization
kicked off for 5G

1st Brooklyn
5G Summit

2015

ITU-R process
nearing completion

5G System

2017

demos

Pre
standardized
mobile

2019

trials

demos

2014

2016

trials

2018

First 5G MWC
showcase

5G PPP projects in
full swing

5G phase 1 specs
ready

NTT DoCoMo
cooperation

5G standardization
on channel model,
requirements and
technology option
selection

WRC2019
preparation
underway

MoU with
CMCC/CMRI

Public
44 Nokia 2015

DOCOMO
Pre
commercial
mobile

Commercial
fixed
wireless
access

5G Radio

Commercial
network
opened

Technology trials
with key customers

2020
Research on
6G starts

Nokia 5G mmWave beam tracking demonstrator

Mobile device

First 5G demos
CEATEC 2014
Access
point

70 GHz band
1 GHz bandwidth

Lens antenna with

64-beam
switching

45

Nokia 2015

3 beam
width

5G mmWave Outdoor results @ AH campus

Parameters

Value

Operating
Frequency

73 GHz

Bandwidth

1 GHz

Modulation

Null CyclicPrefix Single

Carrier
16 QAM
Single Stream
(SISO)

Antenna
Beamwidth

3 degrees

Antenna
34 degrees
Outdoor
Experiments
GHz very promising
Steering
Range@ 73Azimuth
8 degrees
Maximum Range of 200meters
Elevation

46

Nokia 2015

mmWave PoC System @ 2GHz BW supporting 10 Gbps Peak rate

74 GHz
Receiver

IF
Downconverter

Baseband
Receiver
Processing

74 GHz
Receiver

IF
Downconverter

74 GHz
Transmitter

IF
Upconverter

Digital
Baseband
Baseband
Transmitter
Processing

74 GHz
Transmitter

IF
Upconverter

Parameters

Value

Operating
Frequency

~73.5 GHz

Bandwidth

2 GHz

Peak Rate

~10 Gbps

Modulation

Null CyclicPrefix Single

Carrier
R=0.9, 16
QAM
2x2 MIMO

Antenna

Horn Antenna

Baseband
Receiver
Analog
Baseband

IF

Data

Baseband
Transmitter

Data

10 Gbps peak rate using a prototype of NIs mmWave platform- demonstrated at 5G Brooklyn summit and GITEX in Dubai
47

Nokia 2015

Research on user-centric networks

Federico Boccardi
December 2015

User-centric networks

1G-4G cellular designs have

historically relied on the role of cells
as fundamentals units within the RAN

low power BS

sensors

uplink

Recent trends call for a redefinition of

the traditional cell-centric architecture

downlink

mmWave BS

User-centric architectures: the set of

network nodes providing connectivity
to a given device and the functions of
these nodes in a particular
communication session are tailored to
that specific device and session

macro
high power BS

control
data

mmWave
indoor

mobile
device

D2D

F. Boccardi, R. W. Heath Jr, A. Lozano, T. L. Marzetta, and P. Popovski. Five Disruptive Technology Directions for 5G."
Communications Magazine, IEEE 52, no. 2 (2014): 74-80.
F. Boccardi O. Aydin, U. Doetsch, T. Fahldieck, H. P. Mayer, User-centric architectures: Enabling CoMP via hardware
virtualization, PIMRC 2012

Downlink/ Uplink Decoupling (DUDe)

DUDe:
Downlink (uplink)
association: based on
downlink (uplink) reference
signals power

Benefits
Energy reduction
Increased uplink rate

Load balancing
No standardisation needed

51

DUDe: performance (rate)

Main assumptions:
2.6 GHz co-channel, 20 MHz band
5 Macro BS (46dBm) and 64 Small
Cells (30, 20 dBm)
560 UEs (20 dBm)
Prop. fair scheduling

Full buffer traffic

Antenna gains: Macro: 17.8 dBi,
Pico: 4dBi, UE: 0 dBi

52

DUDe: performance (cell loading)

H. Elshaer, F. Boccardi, M. Dohler and R. Irmer. Load & Backhaul Aware Decoupled Downlink/Uplink
Access in 5G Systems. IEEE International Conference on Communications(ICC), June 2015.
F. Boccardi, J. Andrews, H. Elshaer, M. Dohler, S. Parkvall and S.Singh. Why to Decouple the Uplink
and Downlink in Cellular Networks and How To Do It, to appear in IEEE Comm. Mag.
53

DUDe: three examples of implementation

Centralised processing, via
multiple reception in uplink

Possible in LTE-A

AS and NAS signalling

remote anchor

Shared Cell-ID

Dual Connectivity

Possible in Rel. 11, via

proprietary solutions

Possible in Rel. 12, for

inter-freq. deployments

AS and NAS signalling:

anchor at the master node

AS signalling: local anchor

NAS sign.: remote anchor

54

Extensions: Full Duplex via CoMP (CoMPflex)

Spatial displacement can
be exploited to allow full
duplex transmission
without analog cancellation

CoMP

CoMPflex

Full Duplex

Interference at the BS is
cancelled via digital
processing
Interference at the UE is
handled via scheduling and
power control
Preliminary results
available
H. Thomsen, P. Popovski, E. de Carvalho. N. K. Pratas, D. M. Kim and F. Boccardi, CoMPflex: CoMP
for In-Band Wireless Full Duplex, ArXiv, 2015.
55

User-centric networks: conclusions and next steps

Decoupled cell association is a simple idea that can provide major

gains in uplink
It can be implemented without the need of new standardisation
enablers, with proprietary solutions
Extensions based on centralised processing or shared-IDs provide the
possibility of fully digital full duplex
Current works are focused on cell association for hybrid cm/mm-Wave
networks
H. Elshaer, J. Andrews, F. Boccardi, M. Dohler, Downlink and Uplink Cell Association in Millimeter-Wave
Heterogeneous Networks, under submission

56

THANKS!!!

57

The Role of Spectrum Sharing

Thank
you for viewing
the IEEE ComSoc
Live Webinar.
in
Future
Wireless
Networks

Sponsored by