Ariela Zeira- InterDigital- Next Generation Cellular Systems

May 2010

Next Generation Cellular SystemsLTE World Summit 2010

2010 InterDigital, Inc. All rights reserved.

Cellular Network Capacity will not meet Exponential Growth of Data DemandOver the past seven years, aggregate international capacity requirements have grown more than 22-fold. Increasing market penetration and use of Smartphones promise to increase mobile data demand by a factor of 16 or more within five years

Source: New America Foundation

2


Cellular 10x?Yes, but not by a single winning technologyShort Term(0-5 Years)

Mid Term(5-10 Years)

Long Term(>10 Years)

Spectral Efficiency SolutionsFuzzy Cells Coordinated MultiPoint (CoMP)

Spectrum Opportunities

Unlicensed/Lightly Licensed Spectrum High Frequency Solutions

New Topologies

Enhanced Relays Cellular Controlled Direct mobile-to-mobile communications

3


Growing Gap between Peak and Average EfficiencyAverage efficiency is not significantly impacted by increasing the peak efficiency due to its low coverage100

11x gap will further increase10

11x

bits/sec/Hz

1

4x0.1

0.01AMPS GSM GPRS EDGE W-CDMA HSDPA 802.16e LTE 802.16m LTE-A

Peak Efficiency

Average Efficiency

Source: Agilent presentation + IMT submissions

4


Adding Spectrum is Only Part of the SolutionProjected Demand for Spectrum1800 1600 1400 Cumulative Spectrum (MHz) 1200 1000 800 600 400 200 0227MHz BW in 2000(From 1997 ITU report)

ITU 2006 Forecast (M.2078)

1720MHz BW in 2020

No significant new mobile spectrum opportunities on agenda for debate at WRC-12 Further debates on allocations may not occur until WRC-2015

1300MHz BW in 2015

840MHz BW in 2010

Candidate New spectrum for Mobile identified by Final Acts of WRC-07

5


Does it make sense to further shrink the cell size? Growth in wireless capacity over the last 50 years has been dominated by shrinking the cell size An attractive longer-term alternative is cellular controlled direct communication between mobile devices Off loading local traffic Sub-nets with direct mobile-to-mobile communications scale more naturally as the mobile density grows Low deployment costSource: Agilent, 2008

6


Fuzzy Cells A Spectral Efficiency SolutionUE assigned to 2 Component Carriers (CCs) providing best SINR= High Power = Low Power

Low Power CCS have higher priority

BS Power

1

2

3

1

2

3

BS 17 2010 InterDigital, Inc. All rights reserved.

BS 2

Fuzzy Cell Demonstration on LTE Hardware Platform

8


For the same Cell Throughput Fuzzy Outperforms Reuse 1 by 60-100%.Cell-edge Throughput v. Total cell Throughput Proportional Fair Scheduling3

10%-tile cell-edge Throughput [Mbps]

2.5

40% Better Total cell TP at same Cell-edge TPFuzzy

2

1.5

60% Better Cell-edge TP at same Total Cell TP(Beta = 0.75 in PF Scheduler) Typical operating point

1

Reuse 1

0.5

0 70 75 80 85 90 95 100 105 110 115 120

Total Cell Throughput [Mbps]9 2010 InterDigital, Inc. All rights reserved.

For the same Cell Throughput Fuzzy Outperforms FFR by 40%1.8

Cell-edge Throughput v. Total cell Throughput Round Robin Scheduling

18% Better Total cell TP 1.6 at same Cell-edge TP10%-tile cell-edge Throughput [Mbps]1.4 1.2 1 0.8 0.6 0.4 0.2 0 60 65 70 75 80 85 90

40% Better Cell-edge TP at same Total Cell TPReuse 1 with RR offers no trade-off between cell-edge and total cell TP

Fuzzy

FFR

Total Cell Throughput [Mbps]10 2010 InterDigital, Inc. All rights reserved.

Coordinated Multipoint (CoMP) A spectral Efficiency Solution CoMP applies coordination among network nodes to reduce inter-cell interference and improve cell edge performance Gap between gains envisaged with perfect channel feedback (in academia and industry) and with current LTE Rel-8 feedback schemes remains highIntra-site CoMP gains vanish with feedback impairments4x 2, full buffer 10 users/cell 5 4

Gain vanishes due to sub-band frequency granularity1

Avg cell tp [bps/Hz/cell]

4.98 4.13 21%

0.8

3 2 1 0

0%

2.39 0%

0.6

2.02 -15%CoMP JT imp.

CDF

2XCell-edge user tp [bps/Hz]

MU-MIMO ideal CoMP JT ideal MU-MIMO imp.

0.4Single cell transmission CoMP (feedback case 1A) CoMP (feedback case 1B) CoMP (feedback case 2A) CoMP (feedback case 2B) CoMP (subband based ideal feedback) CoMP (subcarrier based ideal feedback)

0.2 0.15 0.1 0.05 0

0.189 0.136 39% 0% 0.073 0% 0.070 -4%CoMP JT imp.

0.2

0 0

2

6 4 UE throughput (Mbps)

8

10

MU-MIMO ideal CoMP JT ideal MU-MIMO imp.

Source: NTT- DoCoMo (R1-101220)

Source: Ericsson (R1-101646)

11


Way Forward for CoMP Address root causes of gaps between academia and current feedback schemes Need for improved Channel State Information (CSI) feedback resolution Need for improved frequency domain precoding granularity

Apply CoMP where most needed and/or theoretical gains can be approached Heterogeneous networks Interference problem is more severe than in macro-only deployment Especially for Femto Closed Subscriber Group and Pico Cells employing large cell extension

Lower delay spread and low mobility can be expected in Femto and Pico cells and reduce performance loss from feedback impairments Relay Backhaul Channel (RBC) More accurate CSI feedback from stationary relay station is possible enabling advanced non-linear precoding schemes. High rank MIMO transmission will not be effective due to higher probability of Line of Sight (LOS) channel from Macro to Relay

12


Enhanced Relays A new Topology Solution Relays promise increased coverage and/or cell-edge performance. Performance gains of simple relays are modest Relays in 3GPP Release 10 Work Item in 2010 Focus on Simple relays with limited coordination

13


Way Forward for Relays Realize the full potential of Relays by Extending the use of relays with enhanced Coordination and Cognition Enabling advanced interference mitigation techniques Forward all or part of the messages Open research area

Relay can forward interference To help subtract it out

Assist Interference neutralization and/or alignment Cognitively access underutilized spectrum Physical Layer Network Coding

14


Example - Coordinated Relays Relay enables orthogonalization of the desired signal and interference at receivers (UEs) via eNB-Relay coordination

Cooperation / Coordination between Relays in the same donor cell

15


Direct Communications Between Mobiles A New Topology Solution Ambitious approach to provide significant performance gains through Exploitation of the spatial diversity Low power mobile-to-mobile crosslinks reusing spectrum many times throughout the system

Cellular and non-cellular cooperative channels Crosslink communications over non-cellular, unlicensed or lightly licensed spectrum: Wi-Fi, White-space,

Initial results for cellular cooperative channels show ~2x gain in cell edge throughput and 40% gain in average cell throughput when compared to Reuse 1 macro deployment Similar gains when compared to FFR

16


Challenges in Direct Communication Between Mobile Devices Grouping of helper nodes Power Control mechanisms Basestation coordination Frame Structures HARQ Mechanisms Crosslink Synchronization Discovery of Neighbor terminals Power management Control signaling to enable crosslink communications Scheduling and resource management Relaying protocolsH-UE H-UE H-UE A subset of H-UEs selected as relays

Terminals

Regions of Interference

Relays

Terminal UE (T-UE)

Source eNB

Destination T-UE First phase Second phase

Backhaul Link Helper UE (H-UE) Cross Link UE without any helpers

17


Summary Cellular 10X requires innovations in multiple areas and technologies Spectral efficiency solutions New Spectrum New Topologies

Fuzzy Cells is a simple spectral efficiency solution, robust to impairments and does not require channel state information and accurate feedback CoMP is still probably the most promising spectral efficiency solution but need to focus on closing the gap between gains predicted by theory and those achievable with current LTE Release 8 Feedback Schemes Most gains will come from New Topologies Enhanced Relays Cellular Controlled Direct Mobile-to-Mobile Communications18 2010 InterDigital, Inc. All rights reserved.

Thank You!

19


Ariela Zeira- InterDigital- Next Generation Cellular Systems

Documents

total cell

total cell

edge tp

edge throughput

mimo imp

cell size

tile cell

rights reserved