LOW-COMPLEXITY DETECTION AND TRANSMIT ANTENNA SELECTION FOR SPATIAL MODULATION

In this paper, we propose a novel low-complexity detection technique for conventional Spatial Modulation (SM). The proposed scheme is compared to SM with optimal detection (SMOD), SM with signal vector based detection (SM-SVD) and another reduced complexity detection technique, presented in the literature. The numerical results show that the proposed scheme can match the error performance of SM-OD very closely, even at low bit-error rate (BER). The computational complexity at the receiver, is shown to be independent of the symbol constellation size, and hence, offers a much lower complexity compared to existing schemes. To further improve the error performance, we then propose two closed-loop transmit antenna selection (TAS) schemes for conventional SM. We assume the receiver has knowledge of the channel and a perfect low-bandwidth feedback path to the transmitter exists. From evaluation of the BER versus normalized signal-to-noise ratio (SNR) and the complexity analysis, the proposed schemes exhibit a significant improvement over SM-OD and other improved SM schemes in terms of error performance and/or complexity.