Complexity-reduced maximum-likelihood hybrid detection for MIMO systems

For wireless communications, the multiple-input multiple-output (MIMO) system efficiently makes use of the spectrum and enhances the transmission throughput. In this work, the maximum-likelihood (ML) detection for the MIMO system is studied, and two ML detection algorithms are first considered for the MIMO system, including the sphere decoding (SD) algorithm and an algorithm based on differential metrics (DMs). Each of the two algorithms has its advantages and disadvantages. The two algorithms are first modified such that they are based on the same signal model. Then, a new ML detection algorithm is proposed for the MIMO system based on the hybrid operation of the two modified algorithms on the tree search process, in which both the branch-and-bound principle and indicative functions are applied to remove unnecessary searches. The proposed algorithm can attain the ML detection with lower average complexity over low and high ranges of signal-to-noise ratio (SNR), as the authors verify by simulations. The proposed ML detection can also generate soft output, and anti-phase sequences are exploited to further reduce the complexity.