Truncation for Low-Complexity MIMO Signal Detection

Joint maximum-likelihood (JML) detector may be used in memoryless multiple-input multiple-output (MIMO) systems to obtain optimal detection performance. However, JML detector performs an exhaustive search and has prohibitively large decoding complexity. To reduce the complexity of MIMO signal detection, minimum mean-square-error (mmse) linear detector (LD), decision-feedback detector (DFD), group detector, and sphere detector (SD) may be used. In this correspondence, we propose a truncation based detector for low-complexity MIMO signal detection, and give theoretical insight into the design and performance of such a detector. We study bitruncation in detail and present two bitruncation approaches. These approaches have low-complexity, and computer simulation results show that they outperform mmse-LD and mmse-DFD