Reduced Complexity Max-Log-MAP Sphere Decoder Using Group Detection in MIMO-OFDM Systems

Tsuguhide AOKI
Yasuhiko TANABE
Hidehiro MATSUOKA
Hiroki SHOKI

Publication
IEICE TRANSACTIONS on Communications   Vol.E88-B    No.11    pp.4220-4228
Publication Date: 2005/11/01
Online ISSN: 
DOI: 10.1093/ietcom/e88-b.11.4220
Print ISSN: 0916-8516
Type of Manuscript: Special Section PAPER (Special Section on Software Defined Radio Technology and Its Applications)
Category: 
Keyword: 
MIMO,  OFDM,  ML,  MMSE,  sphere decoder,  IEEE802.11a,  IEEE802.11n,  

Full Text: PDF(1.2MB)>>
Buy this Article


Summary: 
This paper proposes a Group Detection (GD) algorithm with Max-Log-MAP Sphere Decoder (MLM-SD) in order to reduce the complexity of signal detection in a Multiple-Input Multiple-Output Orthogonal Frequency Division Multiplexing (MIMO-OFDM) system. The proposed algorithm divides spatial streams into multiple partial spatial streams by using Minimum Mean Square Error (MMSE) detector, followed by multiple MLM-SDs with reduced number of spatial streams. Although the spatial diversity gain in the MLM-SD degrades because of the lack of the degrees of freedom exploited by the MMSE detector, its diversity gain is recovered by combining the metrics obtained by the multiple MLM-SDs. In a MIMO wireless LAN multipath fading environment, the complexity of the proposed algorithm is 10% of that of the original MLM-SD and the performance degradation in terms of SNR is slightly less than that of the original MLM-SD in 4-by-4 MIMO architecture with 64 QAM achieving 216 Mbps. It is also found that the proposed algorithm is robust against the limitation of the number of searches in sphere decoder.