一种新的QR-SIC-MSD联合MIMO信号检测算法

New QR-SIC-MSD Joint MIMO Signal Detection Algorithm

针对MIMO信号检测中球形译码(Sphere decoding,SD)算法在低信噪比时接近最大似然(Maximum likelihood,ML)算法复杂度的缺点,提出了一种基于QR分解的串行干扰抵消(Successive interference cancellation,SIC)和修改的球形译码(Modified sphere decoding,MSD)联合MIMO信号检测算法,称之为QR-SIC-MSD算法。该算法在低信噪比时采用SIC算法,在高信噪比时,采用MSD算法,根据噪声方差仅选择一次初始搜索半径,若搜索失败则用SIC解代替。通过SIC算法与MSD算法的结合,大大地降低了球形译码算法的复杂性,同时保证了在中高信噪比时逼近ML算法性能,在低信噪比时接近ML算法性能。文中同时给出了完整的算法流程图及计算机仿真结果。

SD algorithm for MIMO signal detection approaches the computational complexity of ML algorithm. To overcome the shortcoming, a successive interference cancellation and sphere decoding joint detection algorithm, called the QR-SIC-MSD algorithm, is proposed based on QR decomposition. The algorithm adopts SIC algorithm when the SNR is low and MSD algorithm when the SNR is high. MSD algorithm selects the initial searching radius only once, and SIC solution replaces MSD algorithm if the searching is failed. The combination of SIC algorithm and MSD algorithm reduces the computational complexity. Meanwhile, the performance approaches ML algorithm in middle and high SNR and is close to ML algorithm in low SNR. This paper also provides the whole algorithm flow chart and the computer simulation results.