Improved multiple feedback successive interference cancellation algorithms for near-optimal MIMO detection

Abstract

In this study, the authors propose an improved multiple feedback successive interference cancellation (IMF-SIC) algorithm and an ordered IMF-SIC (OIMF- SIC) algorithm for near-optimal multiple-input multiple-output (MIMO) detection. In particular, the multiple feedback (MF) strategy in successive interference cancellation (SIC) detector is based on the concept of shadow region, where, if a decision falls in the shadow region, then multiple neighbouring constellation points are used in the decision feedback loop followed by the SIC technique, and the best candidate symbol is selected by using maximum likelihood cost. However, while deciding the best symbol, the shadow condition is not checked in the subsequent layers which may result in an unreliable decision. Thus, to improve the accuracy of a decision, the authors propose an improved MF strategy where the shadow region condition is checked recursively. Further, the authors also propose an OIMF-SIC algorithm where the log likelihood ratio based dynamic ordering is utilised for ordering the detection sequence. Simulation results validate superiority of the proposed algorithms over the other SIC based detection techniques. In addition, to validate robustness of the proposed algorithms, BER performance is computed and compared under channel state information mismatch. © The Institution of Engineering and Technology 2016.