中国邮电高校学报(英文) ›› 2009, Vol. 16 ›› Issue (5): 20-24.doi: 10.1016/S1005-8885(08)60263-0

• Wireless • 上一篇    下一篇

Performance evaluation of channel inversion precoding for
downlink multi-user MIMO system

陈智勇,王文博,彭木根, 李韧   

  1. Wireless Signal Processing and Network Laboratory, Key Laboratory of Universal Wireless Communication, Ministry of Education,
    Beijing University of Posts and Telecommunications, Beijing 100876, China
  • 收稿日期:2008-10-20 修回日期:1900-01-01 出版日期:2009-10-30
  • 通讯作者: 陈智勇

Performance evaluation of channel inversion precoding for
downlink multi-user MIMO system

CHEN Zhi-yong ,WANG Wen-bo, PENG Mu-gen, LI Ren   

  1. Wireless Signal Processing and Network Laboratory, Key Laboratory of Universal Wireless Communication, Ministry of Education,
    Beijing University of Posts and Telecommunications, Beijing 100876, China
  • Received:2008-10-20 Revised:1900-01-01 Online:2009-10-30
  • Contact: CHEN Zhi-yong

摘要:

In downlink multi-user multi-input multi-output (MU-MIMO) system, not every user (user equipment (UE)) can calculate accurately signal to interference and noise ratio (SINR) without prior knowledge of the other users’ precoding vector. To solve this problem, this article proposes a channel inversion precoding scheme by using the lower bound of SINR and zero-forcing (ZF) algorithm. However, the SINR mismatch between lower bound SINR and actual SINR causes the inaccurateness of adaptive modulation and coding (AMC). As a result, it causes degradation in performance. Simulation results show that channel inversion precoding provides lower throughput than that of single user multi-input multi-output (SU-MIMO) at high signal-to-noise ratio (SNR) (>14 dB), due to the SINR mismatch, although the sum-rate of channel inversion precoding is higher than that of SU-MIMO at full SNR regime.

关键词:

MU-MIMO,;channel;inversion;precoding,;CQI;feedback,;AMC,;lower;bound;SINR,;zero-forcing;

Abstract:

In downlink multi-user multi-input multi-output (MU-MIMO) system, not every user (user equipment (UE)) can calculate accurately signal to interference and noise ratio (SINR) without prior knowledge of the other users’ precoding vector. To solve this problem, this article proposes a channel inversion precoding scheme by using the lower bound of SINR and zero-forcing (ZF) algorithm. However, the SINR mismatch between lower bound SINR and actual SINR causes the inaccurateness of adaptive modulation and coding (AMC). As a result, it causes degradation in performance. Simulation results show that channel inversion precoding provides lower throughput than that of single user multi-input multi-output (SU-MIMO) at high signal-to-noise ratio (SNR) (>14 dB), due to the SINR mismatch, although the sum-rate of channel inversion precoding is higher than that of SU-MIMO at full SNR regime.

Key words:

MU-MIMO, channel inversion precoding, CQI feedback, AMC, lower bound SINR, zero-forcing