空间调制信号的QRD-M检测算法及其改进

针对空间调制(spatial modulation,SM)系统中最优检测算法,即最大似然(maximum likelihood,ML)算法存在的高复杂度问题,提出了基于QRD-M(QR-decomposition with M-algorithm,QRD-M)算法的空间调制信号检测算法.该算法运用M算法树搜索策略,每层只计算最优的M个分支,其性能近似最优且运算量较低,有利于硬件实现.但随着发收天线数增多,传统QRD-M算法的检测性能会下降并需要较长的算法执行时间.因此,采用并行检测的思想,提出了PQRD-M(parallel QRD-M,PQRD-M)检测算法.该算法在各个分支上分别独立地进行搜索,提高了执行效率.对所提出的算法进行了复杂度分析,并在不同天线数目和不同保留节点数下对其误码性能进行了仿真,结果表明,相比于QRD-M算法,PQRD-M算法以增加一定的计算量为代价,能显著地改善空间调制信号检测性能,同时还能节约硬件资源.

QRD-M detection algorithm and its improvement for spatial modulation signal

Aiming at the problem of high computational complexity in the optimal maximum likelihood ( ML) detection al-gorithm for spatial modulation ( SM) system, the SM signal detection algorithm based on QRD-M algorithm is proposed. By using tree searching stage of M algorithm which only computes the optimal M branches at each layer, the QRD-M detection algorithm is close to the optimal performance but has low computational complexity, which is suitable for the hardware im-plementation. However, with the increasing number of transmitting and receiving antennas, the performance of traditional QRD-M algorithm will decline and its executing time will be extended. Hence, by using the idea of parallel detection, a parallel QRD-M ( PQRD-M) detection algorithm is proposed. The algorithm searches each branch respectively and inde-pendently in order to improve the executing efficiency. The complexity of proposed algorithms is analyzed and the perform-ance of BER with different antenna numbers and different reserved node numbers is simulated. Results show that compared with QRD-M algorithm, the PQRD-M algorithm can significantly improve the detection performance of SM signal and reduce the hardware requirements at the cost of certain increase of computational complexity.