Description

We consider the problem of maximum likelihood (ML) signal detection in multiple-input multiple-output (MIMO) wireless communication systems. We propose a new preprocessing algorithm in the form of channel ordering for sphere decoders. Numerical results show that this new channel ordering leads to significantly lower complexity (in the form of the number of nodes visited by the search algorithm); for MPSK modulation where M ≥ 8 and a moderate SNR range of 15 – 24 dB, our channel ordering results in a two-fold to four-fold decrease in the number of nodes visited by the search algorithm. We also present a brief review of the SDR-ML detector, formulated using semidefinite programming and relaxation techniques. Finally, we propose a combined SDR-ML-sphere decoder and demonstrate that it further reduces the number of nodes visited by the search algorithm; for a 20 × 20 BPSK-modulated MIMO system and SNR of 8 dB, the SDR-ML-sphere decoder has an average complexity that is approximately 5 times less than the sphere decoder.