A Two-Stage Clustering Detector for SM-MIMO Communications

Abstract

In this letter, we propose a novel two-stage clustering detector for spatial modulation multiple-input multiple-output (SM-MIMO) systems. Firstly, the quality of the clustering result (i.e., cluster centroids) of the conventional KMC detector will be checked through the proposed rule in the first stage. Then, with the aid of the good cluster centroids of the bad clustering result and the symmetry property of the good cluster centroids, the bad cluster centroids are obtained from the first stage will be corrected to good cluster centroids in the second stage. Furthermore, a new two-stage clustering detector is proposed correspondingly. Our simulation results show that the proposed two-stage clustering detector outperforms other clustering detectors and is capable of achieving near maximum-likelihood (ML) performance with considerably low complexity.