An Efficient Parallel Algorithm with Partial Decision Feedback for Near-Optimal MIMO Detection
Cong Xiong,Xin Zhang,Kai Wu,Dacheng Yang
DOI: https://doi.org/10.1109/GLOCOM.2009.5425716
2009-01-01
Abstract:Maximum-likelihood detection (MLD) is the optimal scheme for multiple-input multiple-output (MIMO) channels. However, due to its exponentially high complexity, many alternative algorithms, including some parallel detection (PD) ones with low complexity and high stability, have been proposed for practical applications. Nevertheless, the existing PD algorithms are unable to exploit sufficiently the diversity order increment for low-complexity algorithms via MLD for partial layers, consequently, the complexity of the sub-detectors is still undesirably high. In this paper, a novel PD algorithm with relative low-complexity sub-detectors, i.e., the partial decision feedback sub-detectors, has been developed. Numerical analysis indicates that the proposed parallel algorithm can achieve the near-optimal performance with much lower complexity in comparison with the existing PD algorithms. Thus, this algorithm makes the parallel detection more feasible in real-life systems with limited parallel processing elements.