Low-complexity MIMO detection for approaching near-ML performance

The research of finding better ways for multiple-input multiple-output (MIMO) signal detection is important and still goes on. In this paper, we propose a low-complexity MIMO detection method mainly with basic and hardware-friendly operations, i.e., zero-forcing (ZF), successive interference cancelation (SIC), double-symbol maximum likelihood (ML) detection, and M-algorithm. We show that it can approach the optimal error rate employing quadrature phase-shift keying (QPSK) signals with affordable complexity in many cases. Simulation results show that in terms of giving a good trade-off between the error rate and implementation complexity, the proposed detection schemes can significantly outperform some contemporary detectors, including the ordered SIC-ZF method and the conventional ML detector employing QR-decomposition and M-algorithm (QRM-MLD), and own an error rate quite close to that of the optimal ML detector.