Mobile Broadband: Vision

1
Mobile Broadband Vision Evolution

• Siavash M. Alamouti, Intel Fellow
• Chief Technology Officer, Mobile Wireless Group

2
Outline

• Intel Vision for Mobile Broadband
• Mobile Internet Requirements
• Mobile WiMAX Technical Overview
• Mobile WiMAX Performance and Comparative Analysis
• Evolution of Mobile WiMAX
• Areas of Future Research

3
Acknowledgements

• Hujun Yin, Intel
• Material for overview and Comparative Performance
• Pouya Taaghol, Intel
• Material for Systems and Network Architecture
• Jose Puthenkulam, Intel
• Material for Standards Timelines
• Shilpa Talwar, Intel
• Material for Future Research

4
What is the Killer Application?

• Anything the Internet can provide today
• possibly more

need a mobile broadband technology that can meet
the requirements of mobile internet
5
Mobile Internet Requirements

• Low cost of devices and subscription
• Transparency of service
• Ubiquity of service
• Ubiquity of user experience

6
Transparency Data Rate Comparison of Wireline
Wireless Technologies
Interesting rule of thumb the actual capacity
(Mbps per channel per sector) in a multi-cell
environment for most wireless technologies is
about 20 to 30 of the peak theoretical data
rate.
7
WiMAX Fixed, Nomadic, Mobile?

• Industry has evolved from a vision of Fixed
  Wireless to Mobile Wireless
• WiMAX industry is aligned on the vision that
  Fixed and Nomadic are special cases of Mobile and
  we need one technology for all to benefit from
  economies of scale.
• IEEE 802.16e is the basis of Mobile WiMAX.

8
Mobile Broadband Evolution to 4G
9
Primary Devices for Mobile Broadband?

• The gateway to the internet is the PC (desktops,
  laptops)
• The larger the screen, the larger the required
  bandwidth
• Primary devices for mobile internet will be
  smaller PCs (not larger handsets)
• PC-like application processing power (service
  transparency)
• Full Microsoft/MAC/Linux OS support (application
  transparency)
• Always on experience
• Leadership for a whole new class of devices
  (UMPCs)
• Small form factor
• Good battery life
• Low cost

10
Mobile WiMAX Technical Overview
11
Mobile WiMAX Salient Features
12
Key Radio Technology OFDMA
Backup?

• For large bandwidths TDMA and CDMA suffer from
  inter-symbol-interference in larger cells
• Large bandwidth small symbol duration
• Symbols gets smaller and channel does not change
• How to combat frequency selective fading?
• parallel orthogonal flat narrowband channels

• Orthogonal subcarriers, high spectral
  efficiency, efficient implementation
• Efficient MIMO implementation

OFDMA is a cost-effective technology for Mobile
Internet
13
Complexity of MIMO-OFDM vs. MIMO-CDMA

• Receiver complexity for MIMO-CDMA grows
  exponentially with bandwidth and linearly for
  MIMO-OFDM

14
Mobile WiMAX Media Access

• Fast dynamic scheduling
• Contention access for bandwidth requests only
• Resource allocation exclusively by BS retains
  tight policy control by network
• Efficient for both bursty, unpredictable traffic
  and voice

• Static Allocation (slot or code based)
• Efficient for voice traffic
• Inefficient for bursty traffic (email, http)

• CSMA/CA
• Efficient for unpredictable traffic in an
  unlicensed band
• Inefficient for predictable traffic (voice)
• Sharing model designed for unlicensed band
• No control of resource allocation policy

Optimal MAC for Mobile Internet
15
Smart Antennas and MIMO
16
Smart Antenna Support
2 Tx / 2 Rx

• Space Time Block Coding (STBC)
• Reduces fade margin by spatial diversity
• Open loop
• Peak rate is not increased
• Spatial Multiplexing (SM)
• Increases peak rate
• Open loop
• Requires good SINR and low spatial correlation
• Adaptive MIMO switch (AMS)
• Optimally select STBC or SM to adapt to channel
  condition
• Reduced feedback
• Significantly improves capacity
• AAS (beamforming)
• Improves link budget
• Reduce interference
• Minor change to client
• 4 or more antennas for significant impact

1 Tx / 2 Rx
1 Tx / 2 Rx

Minimum Configuration for Mobile WiMAX Wave II
17
MIMO-OFDMA Architecture

• MIMO operation in frequency domain
• Flat subcarriers - hij is scalar
• Simple frequency domain one-tap equalizer
• Scalable with bandwidth

h11 h21
H
h12 h22
Multi-Element Transmitter
Multi-Element Receiver
h11
h12
IFFT
h21
MIMO Encoder
MIMO Sub-ch Mapping
FFT
MIMO Decoder
h22
IFFT
y Hs n
18
DL Adaptive MIMO Switching (AMS)
STBC Encoder
AMS
Channel Encoder
Symbol Mapper
MCS
SM Encoder
MIMO mode
LA Decision Unit
CSI SNR
MCS Modulation and Coding Scheme LA Link
Adaptation CSI Channel State Information SNR
Signal-to-Noise Ratio
19
Performance of AMS
Spectral efficiency (SE)
AMS overcomes the deficiencies of STBC and SM and
leads to spectral efficiency very close to the
ideal one at both low and high SNR regions
20
Up Link Collaborative MIMO

• MSs spatially uncorrelated
• no 3dB power penalty

21
System Architecture
22
Mobile WiMAX Network Flat Very-Flat
Architectures
Flat Architecture
ASN
CSN
ASN GW
BS
R6
R3
R8
Policy Server
HLR HSS
MIP HA
R1
DHCP
AAA
R6
BS
R3
R5 (Roaming)
R4
R1
ASN
MS
Another Operators CSN
Policy Server
Very Flat Architecture
HLR HSS
MIP HA
DHCP
AAA
NSP (Network Service Provider)
NAP (Network Access Provider)
Mobile WiMAX networks offer co-existence
interoperability of Flat and Very-Flat solutions
23
Comparison of User Planes Data Flows
Backup?
3G User Plane and Data Flow
too many protocols, too many nodes
Mobile WiMAX User Plane and Data Flow
Simple. Few protocols. Easy-to-implement.
Mostly, IETF protocols. Few MS requirements
24
Mobile WiMAX- 3GPP SAE Interworking
All-IP Core Network
BS Radio Functions
PDN
IMS
Internet
SAE integrates WiMAX to operators core network
as other 3GPP access technologies are with
seamless vertical mobility
25
Mobile WiMAX Comparative Performance
26
Simulation methodology and assumptions

• Methodology based on CDMA2000 methodology
  specified by 3GPP2, now also adopted by 3GPP
• Channel models exhibit lesser delay spread than
  realistic broadband channels and will hence
  benefit the other technologies and soften the
  performance advantage of WiMAX
• Only standards-based features are included.
  Simulation results do not include
• Adaptive Antenna Systems (AAS)
• Interference cancellation or mitigation (at BS or
  MS)
• Novel receiver designs
• Novel frequency/space scheduling (very simple
  scheduler)
• Compared technologies performance are based on
  two receivers in MS and BS with known receiver
  combining techniques

27
WiMAX, EVDO and HSPA Features

1
28
System Configuration
29
Channel Models with Relatively Smaller Delay
Spread (according to EV-DV Methodology)
30
WiMAX, HSPA and EVDO Comparison

• 1Tx 2Rx Rake receiver for EVDO and HSPA
• 2x2 Adaptive MIMO for WiMAX DL
• 1x2 Collaborative MIMO for WiMAX UL

31
Mobile WiMAX, HSPA and EVDO Comparison
results do not include possible improvements with
AAS or interference canceling receivers
32
Beyond Access Opportunity

• Old Model Walled Garden
• Advantage complete control
• Disadvantage few applications, no leveraging of
  creative Internet application
• Broadband Model Open Internet (Dumb Pipe)
• Advantage access to all applications over the
  internet
• Disadvantage operator revenues limited to access
• Mobile WiMAX Model Smart Pipe
• Mobile operators partner with content and
  application providers to deliver enhanced mobile
  services
• Advantage user transparent quality access to
  Internet applications, opportunity for shared
  revenue on contents
• Win-Win

33
802.16 Evolution Vision
IEEE 802.16 Standards
WiMAX Forum Profiles
Networks Targeted
802.16e Corrigendum2
Mobile WiMAX System Profile Release 1 (R1)
(2006)
TDD Solution for Sprint, Clearwire, KT
FDD Solution (for FDD Spectrum) ?
802.16REV
Mobile WiMAX System Profile Release 1.x (R1.x)
(2007)
802.16m
Mobile WiMAX System Profile Release 2 (R2) (2008)
Global TDD FDD Europe, etc.

• IEEE 802.16m PAR Approved Dec 06 Work Starts
  Jan 07 Completion expected Q2 08
• IEEE 802.16 REV PAR starts Jan 07 and Work
  Starts May07 Completion Q4,07
• Mobile WiMAX R2 will be fully backward
  compatible with R1.
• Harmonized spectrum for IMT-2000 and
  IMT-Advanced available for Mobile WiMAX
• Mobile WiMAX R1 already competitive with 3GPP
  LTE and 3GPP2 AIE
• Mobile WiMAX R2 expected to deliver superior
  performance to 3GPP-LTE and 3GPP2 AIE
• Strong support of 802.16e/WiMAX community for
  802.16m

34
802.16m Candidate Requirements Competitive
Comparison
35
Areas of Future Research
36
How to Increase System Capacity?

• Capacity increases
• linearly with RF bandwidth
• as logarithm of signal to noise plus interference
  ratio
• linearly with number of independent spatial
  channels (cells, sectors, MIMO channels,)

• Can we go higher in frequency?
• As frequency increases, the line-of-sight range
• decreases with frequency if both antennas
  dimensions are scaled with wavelength
• remains constant if one of the antennas physical
  dimensions is held fixed
• increases with frequency if both antennas
  physical dimensions are held fixed

antenna gain
aperture size (m2)
Frequency (Hz)
smaller cells, larger bandwidths, higher order
MIMO, and interference management
Speed of light (m/sec)
37
System Capacity Metric

• A meaningful metric is needed to study capacity
  of scaleable systems
• Areal Capacity Sum information rate of users in
  a cell, normalized by cell area bandwidth
• The metric provides insight into means for
  increasing capacity
• Increase single-link capacity between client BS
  (Rk)
• Increase multi-link Cell capacity between clients
  in a cell BS
• Decrease Cell area, which is a standard cellular
  approach

Cell Capacity
38
Interference Limits Capacity Growth
In-Cell interference 2 devices attempt to
access BS simultaneously
Out-Cell interference interference from clients
in neighboring cells
39
Interference a theoretical perspective

• Cell capacity with N users in a cell is no more
  than a constant due to in-cell and out-cell
  interference
• Single cell medium-access methods (TDM, FDM)
  improve capacity, but still bounded by constant
• Multi-cell multi-user co-operative techniques
  promise to eliminate interference by treating it
  as useful information, and approach SNR limited
  capacity. Users add a new dimension to capacity
  improvement

40
Key Cooperative Technologies for 4G
NETWORK LAYER Interference co-ordination
Intelligent Relays
Channel-aware routing Network coding
PHYSICAL LAYER Interference cancellation
Multi-user MIMOSuperposition coding
MAC LAYER Fast link adaptationTraffic-aware
scheduling Multi-cell scheduling