Evaluation of Near MLD Algorithms in MIMO SC-FDMA System

Multiple-input multiple-output (MIMO) single carrier frequency division multiple access (SC-FDMA), which combines the advantage of spatial and frequency diversities with frequency domain equalizers (FDE), has drawn great attention recently. Among all possible detection techniques, maximum-likelihood detection (MLD) is optimal, but its exponentially computational complexity precludes its use in practical MIMO SC-FDMA system. In this paper two classes of near MLD algorithms including sphere decoding algorithm (SDA) and Chase decoding are considered. In particular, Chase decoding is first introduced for a MLD in SC-FDMA systems and then it is compared with SDA. Simulation results show that the use of SDA and Chase decoder both provide superior BER performance compared to linear MIMO FDE. In addition, SDA achieves the exact maximum-likelihood (ML) solution with approximately a factor of 103 reduction in complexity when compared to brute-force ML decoding, and in low SNR (SNR<;8dB), Chase decoding offers the same performance with significantly lower complexity as compared to SDA.