Downlink MIMO Proposal for IEEE 802.16m C80216m-08/405 Downlink MIMO Proposal for IEEE 802.16m Document Number: C80216m-08/405 Date Submitted: May 12, 2008 Source: Guangjie Li (guangjie.li@intel.com) Intel Corporation Hongming Zheng (hongming.zheng@intel.com) Intel Corporation Yang-seok, Choi (yang-seok.choi@intel.com) Intel Corporation Shanshan Zheng (shanshan.zheng@intel.com) Intel Corporation Feng zhou (feng.zhou@intel.com) Intel Corporation Senjie Zhang (senjie.zhang@intel.com) Intel Corporation May Wu (May.wu@intel.com) Intel Corporation Sassan,Ahmadi (Ahmadi.Sassan@intel.com) Intel Corporation Venue: Macau, China Purpose: Discussion and Approval Notice: This document does not represent the agreed views of the IEEE 802.16 Working Group or any of its subgroups. It represents only the views of the participants listed in the “Source(s)” field above. It is offered as a basis for discussion. It is not binding on the contributor(s), who reserve(s) the right to add, amend or withdraw material contained herein. Release: The contributor grants a free, irrevocable license to the IEEE to incorporate material contained in this contribution, and any modifications thereof, in the creation of an IEEE Standards publication; to copyright in the IEEE’s name any IEEE Standards publication even though it may include portions of this contribution; and at the IEEE’s sole discretion to permit others to reproduce in whole or in part the resulting IEEE Standards publication. The contributor also acknowledges and accepts that this contribution may be made public by IEEE 802.16. Patent Policy: The contributor is familiar with the IEEE-SA Patent Policy and Procedures:<http://standards.ieee.org/guides/bylaws/sect6-7.html#6> and <http://standards.ieee.org/guides/opman/sect6.html#6.3>. Further information is located at <http://standards.ieee.org/board/pat/pat-material.html> and <http://standards.ieee.org/board/pat >.

Outline Objectives Design Criteria and Requirements MIMO Architecture Single-user MIMO Multi-user MIMO Feedback for MIMO Mode/Rank Adaptation Conclusions and Recommendations SDD Text

Objectives Meet 802.16m downlink Data requirement a) 5% CDF throughput: 900kbps (2x over 16e reference) b) Overall SE: 2.6bps/Hz/sector (2x over 16e reference) Single-user MIMO: to improve per link performance Multi-user MIMO: to improve system overall throughput especially when system load is high

Design Criteria and Requirements - SU-MIMO Evaluate and select schemes in 802.16e spec to make 802.16m solution concise and effective Improve 802.16e existing schemes if possible Low system complexity Support Single Codeword (SCW) Evaluate if there is obvious gain from MCW over SCW especially with MLD receiver

Design Criteria and Requirements - MU-MIMO Support FDD and TDD Support any antenna configuration e.g. 2x2 ,4x2, ½ lambda, 4 lambda No more than 2x CQI feedback overhead compared to SU MIMO Support the same precoding vectors or codebook design as those in SU Similar CSI feedback overhead compared to SU MIMO Support extra control signaling and specific protocol design for MU MIMO Comparable level of computing and protocol complexity compared to SU Evaluate the impact from frequency selective fading Evaluate the impact from feedback inaccuracy stems from delay and quantization; and CQI estimation/vector estimation error Evaluate the impact from interference uncertainty caused by CQI delay and the change of precoding matrix in interference cell/sectors

MIMO Architecture In SU-MIMO, only one user is scheduled to one RB CQI ( long - term ) , short BF Vector Index ACK / NACK ... User 1 data 2 User P User i Encoder Mode Rank Link Adaptation Retransmission Ind . MIMO Encoder STBC SM etc BEAMFORMER PRECODER IFFT OFDM SYMBOL CONSTRUCTION Preamble Permutation Distributed Localized Pilots Power boosting BEAMFORMING BF Vector Matrix MIMO Multiplexing OFDM Framing Scheduler Decision MCW & SCW control I N P U T SU MU MIMO RB0 RB1 RBn In SU-MIMO, only one user is scheduled to one RB In MU-MIMO, multiple users are scheduled to one RB SU-MIMO and MU-MIMO can co-existence in one frame occupying different RBs

Single-user MIMO Comparison and Summary Use Open-loop (OL) and Closed-loop (CL) for different scenarios - OL: high speed - CL: medium/low speed Example Antenna Configuration Open-loop Closed-Loop 2x2 STBC/SM Codebook precoding 4x2 Antenna Hopping CDD+STBC/SM Antenna selection Antenna Grouping Open-loop Support STBC/SM to backward compatible to 802.16e reference system CDD+STBC/SM is good selection for asymmetric antenna configuration e.g. 4x2 Low complexity in both Tx and Rx side Good performance Small pilot overhead for asymmetric antenna configuration Antenna Hopping is more complicated in receiver side than CDD, and need twice of the pilot For 4 Tx rate 4, Spatial Multiplex is used Closed-loop Codebook based solution is reliable and effective Balance of the feedback overhead and performance Antenna selection is not good in terms of power utilization issue

Multi-user MIMO Comparison and Summary Features Open loop Closed Loop Linear Non-linear Schemes Predefined MU Unitary MU Non-unitary (ZF MU) THP Desc. The precoding matrix is pre-defined and can be semi-statically changed Different precoding matrix for each band can be used User allocation and corresponding V matrix can be decided base on CQI feedback The precoding matrix is codebook unitary matrix Users are paired according to the preferred vector index and CQI The precoding matrix is from the zero forcing process on pair of user channel Successive and nonlinear precoding. Interference suppression for each layer is carried out sequentially with a nonlinear modulo operation Pros Simple for Tx/Rx and scheduling (stream independent scheduling) Good performance when user number is rather large No interference uncertainty Good performance Simple Best performance in ideal case Cons. Not as good as closed-loop when user number is small Interference uncertainty in deficient rank transmission e.g. 4x2 2 stream Interference uncertainty in any case sensitive to feedback inaccuracy Antenna power imbalance high complexity in Tx and user pairing Performance count on full knowledge of H in all subcarriers very sensitive to interference, feedback inaccuracy High complexity Power loss and module loss exist especial in low SNR range

Multi-user MIMO Comparison and Summary (Cont.) Open-loop The Predefined MU MIMO is a good solution with low complexity and overhead No CSI is required Same detection algorithms can be used as those in OL SU MIMO Simple precoding vectors Only dedicated pilot is required Good performance Stems from enlarged multi-user diversity and spatial diversity gain No interference uncertainty issue Easy extension to multi-user macro diversity and Distributed Antenna System Closed-loop Unitary MU-MIMO is a good practical candidate to be used Non-unitary MU-MIMO needs careful evaluation Non-linear MU-MIMO is not suggested to be used Very sensitive to feedback inaccuracy, delay, frequency selective fading High complexity

Feedback for MIMO CQI feedback CSI feedback for CL MIMO Wideband CQI Average CQI for PUSC or Voip Individual CQI For band-selection scheduling Frequency/time/spatial domain compression is required to reduce overhead CSI feedback for CL MIMO Codebook feedback Sounding Analog feedback

CSI Feedback for CL MIMO Feedback schemes Advantage s Disadvantages Codebook reliable design for TDD/FDD Approach perfect performance with limited overhead Feedback together with CQI through feedback channel (error protected) Quantization error ( with MLD receiver, the impact from the quantization error is small) Analog feedback Per tone CSIT processing Less overhead (??) Support TDD/FDD Difficult for whole band CSI feedback Sensitive to interference or not? CQI still need to be fed back Sounding Less overhead TDD only DL/UL Tx antenna number mismatch RF Calibration is required Difficult for wideband sounding Support Codebook feedback Sounding and analog feedback needs to be proved to show benefits if any

Mode/Rank Adaptation for MIMO Use Mode/Rank adaptation to keep the link performance even channel is changing Evaluate both flexible and semi-static solutions Flexible adaptation Feed back CQI of all adaptable rank/mode Can change mode/rank frame by frame Good performance with higher feedback overhead Semi-static adaptation MS requests adaptation when noticed channel change BS decides the adaptation Change mode/rank slowly Less feedback overhead

Conclusions and Recommendations SU MIMO can improve per link performance and Closed-loop SU MIMO can obtain better throughput MU MIMO can significantly improve system capacity A clearly defined design criteria and requirements is important to guide the evaluation of MIMO schemes Feedback, interference uncertainty etc non-ideal conditions and design complexity are all the important criteria to evaluate MU MIMO schemes Recommendations Define clear design criteria and requirement to guide evaluation of MIMO schemes Adopt the proposal in this contribution as the design criteria baseline Support both SU and MU MIMO for link performance and system capacity; For each of them support open-loop and closed-loop for different scenarios Support STBC/SM and CDD in SU open-loop Support codebook precoding in SU Closed-loop Support Predefined MU MIMO in open-loop MU Support Unitary MU MIMO in Closed-loop MU

Proposed Text for SDD In SDD Text DL MIMO.doc