6/14/2014 1 1 Dr. Chih-Lin I CMCC Chief Scientist, Wireless Technologies CMRI, China Mobile Keynote ICC2014 13 June 2014, Sydney Defining the Wireless Future – Vision 2020: Perspectives of Mobile Operators (5G: Data Rate and More) 2 2 •How to increase network capacity dramatically while keep the TCO at a reasonable level. •What user needs must be met by 5G? •How will wireless technologies and mobile infrastructure be integrated then? •Are there any particular needs from mobile operator ’s point of view? •How will mobile operators stay on the center stage in the future? •What kind of technologies they are looking for? •What’s the likely roadmap towards next generation mobile network? Recap...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
6/14/2014
1
1
Dr. Chih-Lin I CMCC Chief Scientist, Wireless Technologies
CMRI, China Mobile
Keynote ICC2014 13 June 2014, Sydney
Defining the Wireless Future – Vision 2020: Perspectives of Mobile Operators
(5G: Data Rate and More)
2
2
•How to increase network capacity dramatically while keep the TCO at a
reasonable level.
•What user needs must be met by 5G?
•How will wireless technologies and mobile infrastructure be integrated then?
•Are there any particular needs from mobile operator’s point of view?
•How will mobile operators stay on the center stage in the future?
•What kind of technologies they are looking for?
•What’s the likely roadmap towards next generation mobile network?
Recap...
6/14/2014
2
3
3
Driving Force of Wireless Future
• Fast traffic growth from both
Mobile Internet and IoT
• 1000x within 10 years a
common consensus
4
4
Composition of the Growth
Data Composed from reports of Cisco, Informa, Strategy Analysis, etc.
M2M
Video
Others
Traffic
EB/Month
Service Partition
Smart Phone Func. Phone M2M Others
China Global
• CAGR of video service
will be 79% between
2010 and 2020, and will
be 70% of total traffic in
2020
• 70% of traffic will be
from smart phone
• CAGR of IoT will be
86% between 2010 and
2020, however, will be
7% of total traffic in
2020
• The major issue of IoT
is number of
connections rather than
traffic
6/14/2014
3
5
5
Challenges for MNOs
• Tencent, Google, Apple have built their open
platform one after another
• OTT service have challenged voice, SMS service
of operator.
• User growth of MNOs is slowing down, and it is
difficult to support the sustainable development
• Significantenergy efficiency improvement in the
network infrastructure
• Intelligent options for reducing operators’
CAPEX/OPEX
High CAPEX/OPEX of RAN result from BS equipment room
0
50
100
150
200
10^8KWH Power Consumption (CMCC)
通信机房及其他 基站耗电
93.3 111.4
119.3 129.4
169.1
143
178.6
2008 2009 2010 2011 2012 2013 2014
basestation Others
2020
~0.25E
B/M
@2010
500x
@2020
1000x
@2020
Revenue
Growth
Big
Gap
Internal
Cost and Energy Consumption
External
OTT service providers
Internal
Traffic v.s. Revenue
6
6
0.7Bilion users 0.83Million BTS 80GWH Power consumption
GSM/GPRS/EDGE TD-SCDMA
TD-LTE WLAN
75Million users; 0.45Million BTS 13GWH Power consumption
4.2 Million AP 2GWH power consumption
20,000 BTS by 2012 260,000 BTS by 2013 500,000 BTS by 2014
Special Challenge for CMCC
How to coordinate four networks to meet needs of 780M subs?
• On April 19,2013, IMT-2020(5G) is established by Ministry of Industry and Information, Development and Reform Commission, and Ministry of Science and Technology.
• FuTure Forum 5G SIG
6/14/2014
7
13
13
5G Vision (Whitepaper, IMT-2020 PG, 29 May 2014)
14
14
•be able to sustainably satisfy the requirement of 1000x mobile data
traffic growth.
•provide users with fiberlike access data rate and "zero" latency user
experience.
•be capable of connecting 100 billion devices.
•be able to deliver a consistent experience across a variety of scenarios
including the cases of ultra-high traffic volume density, ultra-high
connection density, and ultra-high mobility.
•be able to provide intelligent optimization based on services and users
Huawei ‘s E-band 115Gbps Prototype MiWEBA project under EU FP7
• mm-Wave for fonthaul/backhaul/access/direct link
22
22
• Peak data rate: 10Gbps
• Latency: < 1ms
• Bandwidth: 100MHz under 6G, 500MHz above 6G
• Subcarrier spacing: 60, 120, 240KHz
• Packetized mini-frame design within sub-frame
• backward compatible with LTE
• TTI: 0.1, 0.125, 0.2, 0.25ms
• Flexible UL/DL allocation
• Backward compatible with LTE
High Freq. Band: Numerology of New Air Interfaces
6/14/2014
12
23
23
… and More
24
24
Office Subway High-speed Train
Typical
Scenarios
UHD Video Streaming
Augmented Reality
Cloud Storage
Online Education
ITS Smart Home
Typical
Services
Freeway Stadium Residential Area
Mobile Internet services
High density High mobility High traffic
Internet of Thing
Remote
Medicine
Rich Requirements for 5G
Modern
Agriculture
Environment
Protection
24
6/14/2014
13
25
25
Expected Capabilities of 5G (IMT2020PG)
User Experienced Data Rate
Traffic Density
Peak Data Rate Mobility
Latency
Connection Density
• China input to ITU-R M. [IMT.Vision]
• IMT-2020 PG “IMT-2020(5G)PG-WHITE.PDF” released on May 29, 2014
Fiber-like access data rate and "zero" latency user experience
26
26
Identified Air Interface Technologies (IMT2020PG)
6/14/2014
14
27
27
Identified Networking Technologies (IMT2020PG)
28
28
Enabling Technologies for 5G Capabilities (IMT2020PG)
6/14/2014
15
29
29
NGMN 5G Use Cases (Interim draft, June 3-5)
29
The post-2020 outlook is vastly broad in terms of variety and variability. Sets of selected use cases
show both enriched service categories and also prospects for numerous new services.
Personal/Social
Interaction in
Connected
Society
Massive Internet
of Things Critical & Safety -
Lifeline
Extreme Real-
Time
Communication
Ultra-reliable
(M2M, M2H, H2M) Low Data/Power
Vehicle to Vehicle and to road communication
Set of use cases
Examples
Remote Operation
Automated Industries
Health & Assisted Living
Virtual collaboration
Out of coverage and proximity
30
30
MGMN Technology Components
R1 – Spectrum
access
Flexible use of
licensed
spectrum
Integrated
unlicensed
spectrum
Use of higher
frequency
bands
Duplex Mode
R2 – Radio link New waveforms
Advanced
multiple
access
technologies
Radio frame
design/
numerology
Massive MIMO
and enhanced
multi-antenna
schemes
Advanced
receivers
Interference
coordination
Technologies
for small
packet
transmission
R3 – Radio
access capacity
Densification:
Small cells/
Ultra-dense
networks
Dual connectivity –
capacity/ coverage
split system design
Enhanced
multi-RAT
coordination
Device-to-
device
communication
s
Wireless backhauling
(e.g., self-
backhauling and
relay)
N1- Network
flexibility
Software-
defined
networking
Virtualized
mobile core
network
Virtualized
RAN
Flexible split of
functions among
network nodes
State-
disintegrated
core node
Micro-servers
N2 – Efficient /
adaptive NW
resource usage
Traffic
optimization
Enhanced multi-
operator network
sharing
Scalable
service
architecture
Big data and
context
awareness
Content-
optimization
and adaptive
streaming
Intelligent
heterogeneous
management
N3 – Other
enablers
Technologies
for massive
connectivity
All optical
transport
network
6/14/2014
16
31
31
CMCC’s Timeline for 5G
Green Communication Research Center established
in Oct. 2011, initiated 5G Key Tech R&D.
• ITU-R M.[IMT.FUTURE TECHNOLOGY TRENDS] to be released on July, 2015
• ITU-R M. [IMT.Vision] to be released on Oct., 2014
32
32
CMCC 5G Themes
From UE to
Infrastructure
From CN to
RAN
Soft Green
IT Based core network
Anchor BS
Nano AP
Virtual BB pool Content
Pool
Anchor BS
Anchor BS
Massive RRU
Relay D2D
relay
D2D
Indoor Coverage
User Centric Access Network
Supporting exclusive usage of
available spectrum of each user
LSAS
6/14/2014
17
33
33
Rethink Shannon
(EE/SE Co-design)
Rethink Ring&Young
(No More “Cell”)
Rethink Signaling/Control
(Load/App Aware)
“Invisible”
BS
Rethink Spectrum
Refarming
SNR
C
“Towards Green & Soft: A 5G Perspective,” IEEE Comm. Magazine, Vol.52, Feb.2014
Five “Pearls”
34
34
CMCC Latest Publications
Rethink Spectrum Refarming
Invisible “BS”
Rethinking Signaling/
Control
EE/SE
No more “Cell”
• Book chapter: Towards Green and Soft, in book “ Towards 5G: Applications,
Requirements & Candidate Technologies”, John Wiley & Sons Ltd (in preparation)
• Book chapter: Rethink Ring and Young: Green and Soft RAN for 5G, “Fiber
Wireless Convergence in Next-Generation Communication Networks--- Systems,
Architectures, and Management” Springer Optical Networks Series (in preparation)
• “C-RAN: toward open, green and soft RAN” submitted to IEEE Network Magazine
• “Network Deployment and Operation Based on Spatial and Temporal Traffic Model”
submitted to Chinacom
• “Trillions of Nodes for 5G!?”, submitted to IEEE GC2014.
• “Full Duplex: Coming Into Reality in 2020?”, submitted to IEEE GC2014.
• “Optimal Antenna Configuration in Hybrid Digital and Analog Beamforming Structure”,
submitted to IEEE GC2014.
• “Large Scale Antenna System with Hybrid BFfor Millimeter Wave 5G”, submitted to
IEEE Communications magazine.
• "EE-SE Relationship for Large-Scale Antenna Systems“ in IEEE ICC 2014
• “Large Scale Antenna System with Hybrid Digital and Analog BF Structure” in IEEE
ICC 2014
• “Small Data Optimized Radio Access Network Signaling/Control Design” in IEEE ICC
2014
• “A Temporal Domain Based Method against Pilot Contamination for Multi-cell
Massive MIMO Systems” in IEEE VTC 2014
• "Fundamental Properties of the EE-SE Relationship“ in IEEE WCNC2014
• 5G Workshop in IEEE IWS2014
“Towards Green & Soft: A 5G Perspective”
IEEE Comm. Magazine, Vol.52, Feb.2014
6/14/2014
18
35
35
Rethink
Shannon
Rethink
Ring&Young
Rethinking
Signaling/Control
Invisible
“BS”
Rethink Spectrum
Refarming
Latest Developments: EE/SE Co-design
EE/SE
Framework Increasing
Speed when
the SE
approaches
zero
Decreasing
Speed when
the SE
approaches
infinity
EE-optimal
Points (Green
Point)
The impact of the circuit power and the CSI
on the EE-SE curves in the single-user
case,
The impact of the number of users on the
EE-SE curves in the multi-user case
No-CSI: the EE-optimal points always stays in a straight line regardless of circuit power, and the number of transmit antennas. CSI: the intercept of EE-optimal lines increases with M.
Through serving more users, the
speed of EE reduction will alleviated
with the increase of the SE.
0 1 2 3 4 5 60
1
2
3
4
5
6
7
8
9
10x 10
5
SE (bps/Hz)
EE
(b
it/J)
Pc= 0
Pc = 20 W
Pc = 10 W
Monotonic
tradeoff
Given EE,
two SE
values exist
36
36
EE/SE Co-design of Hybrid BF Structures ...
DAC
DAC
.
.
.
Antenna 0
Antenna (N*M-1)
S0(t)
SK-1(t)
0
0w
1
0
Mw
0
1Nw
1
1
M
Nw
PAPA
PAPA
Antenna (M-1)
Antenna ((N-1)*M)
.
.
.
.
.
.
DBF
D
UE 1
UE K
.
.
.
ABF
A1
ABF
AN
Transceiver 0
Transceiver N-1
For NM=L and independent N and M cases
– Given SE, there exists optimal N which yields highest EE
– Given SE, there exists optimal M which yields highest EE
– Antenna/transceiver On/Off • The BS can be designed with the maximum number of N and M under given SE requirement range,
and
• Enhance EE performance via antenna /transceiver On/Off based on the SE requirements
Green point EE optimization
For NM=L case • There exists optimal N which maximize the green point EE
For independent N and M case • The green point EE is monotonically increasing with N
10 20 30 40 50 600
0.5
1
1.5
2
2.5
3
3.5
4x 10
6
M
EE
SE=6
SE=12
SE=18
SE=24
SE=30
SE=36
SE=42
SE=48
SE=54
SE=60
6/14/2014
19
37
37
Rethink
Shannon
Rethink
Ring&Young
Rethinking
Signaling/Control
Invisible
“BS”
Rethink Spectrum
Refarming
Latest Developments:EE/SE Co-design
0 2 4 6 8 10
0
2
4
6
8
10
0 2 4 6 8 10
0
2
4
6
8
10
Two deployment schemes
Macro
Number
Micro
Number
Peak
Power
Low load
Power
Average
Power
Scheme 1 9 21 8016w 6252w 6546w
Scheme 2 7 47 9098w 5370w 5992w
Tab. Daily power consumptions of two deployment schemes
working area load
km
km
2 4 6 8 10
10
8
6
4
2
residential area load
km
km
2 4 6 8 10
10
8
6
4
2peak traffic load
km
km
2 4 6 8 10
10
8
6
4
2t = 11:00 am t = 9:00 pm Peak rate traffic
0 0
1 1 0
( ) ( )S S
T i k k i k k
i k i k
P t dt s P t s P t P tMin
Solution Space
Peak rate power Low load power
Energy efficient network deployment
• Peak data rate
• Traffic variation over time and space
0 2 4 6 8 10 12 14 16 1810
-1
100
101
102
103
104
105
106
SE (bps/Hz)
EE
(b
its
/J)
Current GSM Point
GSM Curve
Current LTE Macro-cell Point
LTE Macro-cell Curve
LTE Pico-cell Curve
Current LTE Pico-cell Point
38
38
Hetnet
From single-
layer coverage
to multi-layer
coverage
CoMP
From cell-level SP
to coordinatd SP
among CoMP set
Short-term sales-
up potential
BCG2
Signalling/data
decuoping
DAS
From SP to
centralizedcell-level SP
Joint processingCoMP processor
Coordinated multi-points Tx/Rx
DAS: Distributed (large-scale) antenna system
… RRU RRU RRU RRU
BBU
macro
micro
femto
Break in
coverage Break in
RRM
Break in Signal
Processing
Break in
Protocol
Rethink Ring&Young: No More “Cell”
C-RAN as a Starting Point
6/14/2014
20
39
39
Rethink
Shannon
Rethink
Ring&Young
Rethinking
Signaling/Control
Invisible
“BS” Rethink Spectrum
Refarming
Latest Developments: No More “Cell”
Prototype Profile: • TD-LTE BBU: HP ML350,
• Commercial UE (HiSilicon), RRH and vEPC
• Commercial BBU L1/L2/L3 protocol stack
• Demonstrate the feasibility of GPP’s capability of
wireless signal processing
DAS, CoMP, HetNet, BCG2 C-RAN
Soft BS Soft BBU Pool
• OpenSource Virtulization
• Accelerator-based C-RAN for world-first TD-LTE Ping
call based on commercial protocol
(Real time demo in MWC2014)
40
40
Successful demo in MWC 2014
Mobile Phone (Samsung)
Mobile Phone (Samsung)
Demo Booth
Equipment Room
R
R
H
R
R
H
Mobile
Phone (Samsung)
C-RAN (IBM/Agilent)
C-RAN (ALU)
Nanocell (Comba)
vSAEGW
(ZTE) Edge
Applications
vEPC
vEPC(Huawei)
vEPC (ZTE)
vMME
vIMS (ZTE)
vIMS (Huawei)
vMME(ALU)
vSAEGW (Cisco)
vHSS(Linker networks)
Future Networks: Softer and
Greener
Theme: Future network: Softer and Greener
6/14/2014
21
41
41
Made NFV E2E
04132014
NFV Approach
BRAS
Firewall DPI
CDN
Tester/QoE monitor
WAN Acceleration
Message Router
Radio/Fixed Access Network Nodes
Carrier Grade NAT
Session Border Controller
PE Router SGSN/GGSN
• 200 members from both IT and telecom industry since its foundation on Oct. of 2012
• Multiple WG/EGs, e.g INF, SWA., MANO, PER, REL and SEC. • Have released 5 ISG-level documents to the industry, including
architecture, use cases, terminologies, requirements and PoC promotion.
Independent Software
Vendors
High volume Ethernet switches
High volume standard servers
High volume standard storage
Orchestrated,
automatic & remote install.
Com
petitiv
e &
Innovativ
e
Ecosyste
m
Independent Software Vendors
Source: ETSI NFV ISG
Classical Network Appliance Approach
42
42
…
CPU I/O Accelerator
VM Of
TD-LTE
VM of
Edge APPs
VM of
FDD LTE
Standard Server
Video Optimization
TCP Optimization
Fashion
service Fitness service
KTV service
Food service
C-RAN Service Map Service pushing base on location
CDN/Cache on Edge
˙Passive cache and active pushing on the RAN side according to the popularity of the content of service
˙Reduce stress of network interaction, improve user experience
˙Reduce TCP signal feedback overhead through redundancy coding
˙Adjust the TCP congestion window dynamically based on wireless variation
˙ Real-time analysis towards the information of wireless channel realizes dynamic video coding and QoS guarantee on RAN side
•Advertising and details of buildings in the cell
•Accurate position
navigation
VM of
GSM
Virtual Switch
Hypervisor
C-RAN in MAE 2014: Open platform for diverse Edge Service
6/14/2014
22
43
43
Rethinking Signaling/Control
• VoIP: 65~375
• IM(Weixin): 2~6.6
• IM(QQ): <1
“Green” protocol for diverse QoE requirements of
future mobile applications • Adaptive protocol itself with slim signaling set/flow
• Application aware
• Connection oriented connectionless oriented
• …
0 10 20 30 40 50 60 70 80 90 1000
50
100
150
200
250
300
350
400
User Mobility by Cell Change Times in 10 Minutes
Data
to S
igna
ling
Ratio
(DSR
)
DSR vs. User Mobility
VoIP with RRCInactivityTimer=5s
VoIP with RRCInactivityTimer=60s
0 10 20 30 40 50 60 70 80 90 1000
1
2
3
4
5
6
7
User Mobility by Cell Change Times in 10 Minutes
Data
to S
igna
ling
Ratio
(DSR
)
DSR vs. User Mobility
Weixin with RRCInactivityTimer=5s
Weixin with RRCInactivityTimer=60s
QQ with RRCInactivityTimer=5s
QQ with RRCInactivityTimer=60s
Service
type
Ratio(
%)
Data
Rate(Kbps)
Packet
Length (s)
Packet
Size
(KB)
Packet
Arrival
Interval(s)
Text 60 1 20
Voice 35 1 10 10 20
Picture 4 150 2000
Video 1 25 60 1500 20000
Keep alive
Signaling
0.6 300
Data rate
(Kbps)
Packet
Length (s)
Arrival
Interval (s)
VoIP 1 60 2400
2014/6/14 43 Chih-Lin I
Data signaling ratio is extremely low for IM
44
44
Rethink
Shannon
Rethink
Ring&Young
Rethinking
Signaling/Control
Invisible
“BS”
Rethink Spectrum
Refarming
CMCC Latest Developments
Small Data optimized access signaling/control
•DSR performance of IM is improved
by 6-fold with the introduction of
optimized RRC state together with the
optimized signaling flows.
• Slim RRC state: new RRC
state for small data access
•Slim RRC procedures: no
RRC maintenance (H.O., CQI)
•Sparse
•Periodic
6/14/2014
23
45
45
Rethink
Shannon
Rethink
Ring&Young
Rethinking
Signaling/Control
Invisible
“BS”
Rethink Spectrum
Refarming
CMCC Latest Developments
Scheduling based access
Overhead:
• Control channel
Contention based access
Overhead:
• Guard interval
• Collision
Normalized load
Th/P
RB
Turn point
Connectionless
access region
Connection-oriented
access region
Load-aware control/ signaling reconfiguration
46
46
Rethink
Shannon
Rethink
Ring&Young
Rethinking
Signaling/Control
Invisible
“BS”
Rethink Spectrum
Refarming
CMCC Latest Developments
Service-aware RAN signaling/control optimization
• On demand long connection over the air
Avoid delay from connection setup and transition
Low overhead “weak” connection state: not fully
functional channel feedback and estimation
• Customized connection configuration for fast delivery
Shorter DRX periodicity
• Real time radio link feedback
Inform application timely in case of radio link failure
Better user experience
Real-time IP PUSH Delay sensitive
( <seconds), for example,
Shopping verification code
Application protocol optimization
• Adjust HTTP over TCP initial and congestion windows