3/17/2015
IET
5G Radio Technology Seminar
17th March 2015
“Always Sufficient”
Professor Rahim Tafazolli
Director Institute for Communication Systems (ICS), 5GIC
University of Surrey, 5GIC
Tuesday, 17 March 2015
r.tafazolli@surrey.ac.uk
#5GIC
2
1
3/17/2015
Wireless Standards Evolution – 5G
Next generation Global standard around 2020
1983
1991
2001
2011
2021
Research & Std
Research & Std
Research & Std
Research & Std
2031
2041
5G
4G – LTE/LTE‐Advanced
3G – WCDMA/HSPA/HSPA +
2G – GSM/GPRS/EDGE
Systems tend to co‐exist rather then replace previous generations
1G ‐ TACS
Timescale getting shorter between Research/Standardisation and Commercialisation Tuesday, 17 March 2015
r.tafazolli@surrey.ac.uk
#5GIC
3
What is 5G?
2
3/17/2015
5G is SG
Special Generation
Past 30 Years
Next 30 Years
4G
Higher, Higher Speed
Mobile Communications
DATA, voice
3G
Higher Speed
Mobile Communications
Voice + DATA
2G
Speed
Mobile Communications
VOICE + data
Technologies that enable Digital Economy and Society





Mobile Broadband Internet
Transportation
Energy
Manufacturing
Health
Towards  Smart Homes, Cities and Countries
5
Tuesday, 17 March 2015
5G in one sentence
“Always Sufficient Rate” to give users the perception of Infinite Capacity”
3
3/17/2015
Killer Applications
The Internet!
Tuesday, 17 March 2015
r.tafazolli@surrey.ac.uk
#5GIC
7
Fundamental change in research approach
5GIC approach: Information T & Communication T & Control T
Start from end user QoE
 Unlike 2G….4G , designed for end device
Start from density cell network technologies
 Scalable to macro cells-----------unlike previous generations
 New waveform scalable from IoT to Broadband, spectrum aggregation ….
Capacity & Energy efficiencies, Latency , Reliability
 Speed is not the differentiator between 5G and previous generations
Spectrum and system agnostic
 Uniformity operation across licenced and licenced-exempt bands
 Broadcast, Cellular, WiFi technologies all support: Data, Video, Audio
Tuesday, 17 March 2015
8
4
3/17/2015
Connectivity for Communications and Control
Major Technical Requirements compared with 4G
<50x

Latencies

Reliability/Availability

Energy Efficiency >

Capacity (b/s/Hz/m2)

QoE

Security/privacy
>100x
100x
> 100x
9
Tuesday, 17 March 2015
Latency
Not only for mission critical application but
Making networking and advanced radio technologies work efficiently
Higher Speed
doesnotyield
But
Lower Latency
yields
Lower Latency
Higher Speed

Low latencies at all layers, new parallel Signal processing and Computing algorithms

Most delays in previous generations due to C-Plane signalling




New end-to-end signalling architectures
New RAN
New Network architecture (caching, Edge computing..)
New application protocols…
4G initial call setup total delay (from RRC IDL > CONNECTED mode) :
Tuesday, 17 March 2015
~ 80 ms
10
5
3/17/2015
Reliability, Resilience & Availability
>99.9999..99
Resilience by Duplication and adding Redundancy is not the answer

Equally important to Low latency

“always sufficient”… No restrictions to only one network

True Fixed-Mobile convergence

Spectrum Agility: All spectrum including licensed and licensed-exempt bands

Cell edge capacity problem

Network Agility: Transport/Networking/Routing protocols, software-based network architecture
and SDNetworking

Distributed control between device and network, for example new MAC
11
Tuesday, 17 March 2015
Energy Efficiency
Major Opex Element

Impacts: Radio, Network arch with distributed micro servers,

Signalling architecture, RF and baseband architectures and energy efficient electronics circuit designs

Small network nodes .. (cooling & operation energy)
New Signalling Architecture
Signalling
Standby
User Data
TODAY
Tuesday, 17 March 2015
New Waveforms
RF, Signal processing
algorithms
New RAN Architecture
>2020
12
6
3/17/2015
Capacity
Not only densification

Densifications + joint cell processing+ Massive MIMO + Multi-cell Resource
management

Cell-edge capacity

Indoor cells,

Waveforms with flexibility in spectrum aggregation New air Interface

Full duplex radio,

Must understand better characteristics of mm,THz,.. VL bands before developing
new techniques
Tuesday, 17 March 2015
Smaller Cells – Fundamental capacity limits
7
3/17/2015
Coverage and Capacity
 One size does not fit all!
 Low+ Medium+ High Dense cells


Capacity limited
Coverage limited
 Frequency options 


< 1GHz  Coverage
1‐6 GHz  Capacity and Cost
Millimetre Band Capacity
New Air-Interface?
 Ultra Dense Cells
 Scalable to

Medium and low density cells

From narrow band to broadband services
8
3/17/2015
New Air-Interface KPIs

Objectives ‐NOT link spectral efficiency

Low control signalling overhead for management, 
No stringent time‐frequency control as in OFDMA 
Light MAC protocol

Support fast and reliable spectrum sensing for opportunistic spectrum sharing with and without database support

Highly energy efficient

Flexible implementation of spectrum aggregation across highly fragmented spectrum including license‐exempt band

Allow full‐duplex operation

Sub‐millisecond Air‐Interface latency

Scalable for device to device and machine type communications

Candidates waveforms:

FBMC, GFDM, SCMA, NOMA, UFDM and their Hybrid with LDS … Bandwidth, How much?
Many factors influence

Important ones:

Latency of sub‐ms on air interface

Full diversity exploitation

Pilot pollution‐ Cell edged capacity
Initial study suggest: minimum of 100MHz
Tuesday, 17 March 2015
18
9
3/17/2015
Full Duplex Radio
Comparison of different academic full duplex state‐of‐the‐art
#
Parameters
Univ. of California Rice Univ.
1
Antenna
Configuration
1T1R
1T1R
Kumu
2014
2T2R
Stanford
2013
1T1R
Stndford
2014
3T3R
UESTC
2013
1T1R
UESTC
2014
2T2R
2
Operating
Frequency (GHz)
2.4
2.4
2.4
2.4
2.4
1.6‐4.0
2.5‐2.7
3
Signal Bandwidth (MHz)
30
10
20
80
20
20
20
4
Antenna Transmission
Power (dBm)
10
10
20
20
20
10
25
5
Antenna Isolation (dB)
0
20
35
15
15
20
45
6
RF Self‐Interference
Cancellation (dB)
47
35
45
63
65
55
45
7
Digital Self‐
0
Interference
Cancellation (dB)
Total Self‐Interference 47
Cancellation (dB)
26
50
35
35
31
40
55
120
105‐110
115
91
115
9
Experiment
Environment
Indoor
Multi‐path
Indoor
Multi‐path
Indoor
Multi‐path
Indoor
Multi‐path
Indoor
Multi‐path
Indoor
Multi‐path
10
Experiment
Equipment
WARP
Self‐made Self‐made Self‐made Self‐made uSDR
Self‐made uSDR
8
Anechoic chamber, Single‐
path Agilent
19
Tuesday, 17 March 2015
Impactful Research
10
3/17/2015
5GIC’s Mission
•
To be a model for collaborations with academic institutions, industry and the
community
•
To deliver innovative solutions in order to generate social and economic value
•
To influence future standards and regulation
•
•
Leadership
Research
•
Teaching
•
•
•
•
•
IPR
•
Revenue
Tuesday, 17 March 2015
Service providers
Infrastructure & device manufacturers
Test equipment vendors
Standardisation bodies
Application & content providers
Industry forums and alliances
Regulators and policy makers
Academia and research centres
SMEs & other solution providers
Leadership
r.tafazolli@surrey.ac.uk
#5GIC
Growth
5G Services
21
5GIC High Level Technical Activity Roadmap
June’15, (LTE‐A, IoT)
Tuesday, 17 March 2015
r.tafazolli@surrey.ac.uk
#5GIC
22
11
3/17/2015
VDF CORE Networks
BT METRO ETHERNET
Cluster‐1
Spectrum (OfCOM)
Cluster‐4
Cluster‐2
40MHz (2.6 & 3.8GHz, mmBand) 400G
400G
Cluster‐3
Cluster‐5
CLOUD
Infrastructure
EE RAN/Core
Testbed Baseline Configuration

All fibre backhaul 100G
 Duck Fibre Front-haul
 Metro-Ethernet for IP backhaul

400G Router

Gateway
r.tafazolli@surrey.ac.uk
#5GIC
12
3/17/2015
5G UDN Testbed Upgrade
5G UDN Overlay

5G Base-station Prototype
 RF: Massive MIMO (128x128)

Cloud Blade Server
 C-RAN Hyper-Transceiver

20 5G-test-terminals (Van Drive)
5G UDN Testbed Upgrade
Area Spectral Efficiency
 >>10Gbps/km2 @20MHz
TDD 2.6, 3.8 GHz, mmWave? Spectrum
End-User Performance
 1Gbps@mobile
13
3/17/2015
Achievement highlights so far…
 10 patents filed so far

MINT
100MHz Bandwidth 1024x1024 Multi‐User MIMO
Highlights of some of them are:
800 Gbps
600 Gbps
 Speed and spectrum efficiency



Highest-ever speed wirelessly
>800x highest speed in 4G
Interference is good
400 Gbps
200 Gbps
4QAM
16QAM
64QAM
256QAM
 Uniform user experience, all over cell coverage

15x capacity increase at cell edge compared with state-of-the-art technologies
Tuesday, 17 March 2015
r.tafazolli@surrey.ac.uk
#5GIC
27
Thank You
Landing
Tuesday, 17 March 2015
28
14