联合SIC和QRD-M树搜索的低复杂度VBLAST检测算法

Joint SIC and QRD-M Tree-Search Detection Algorithms with Low Complexity in VBLAST Systems

联合SIC和QRD-M树搜索,提出一种低复杂度的VBLAST检测算法:SQRM-SIC算法。该算法基于信道矩阵的排序QR分解(SQRD),首先对搜索树前几层采用QRD-M检测,然后对后续层进行SIC检测,获得了检测信号列表。在所提算法基础上,通过修改SQRD算法中的排序规则,得到MSQRM-SIC算法。复杂度分析和性能仿真表明,通过调整参数,SQRM-SIC算法和MSQRM-SIC算法都可获得较好的复杂度和性能折衷。其中,后者性能明显优于前者;且后者与QRDM算法相比,可以有效降低复杂度,而基本不损失性能。

Joint SIC and QRD-M （SQRM-SIC） algorithm is proposed for VBLAST system as a low complexity detection algo- rithm. The proposed algorithm firstly decomposes channel matrix using sorted QR decomposition （ SQRD）, then detects the first several layers by QRD-M and the remainder by SIC ,finally obtains detected signal list. By changing the sorting rule of SQRD ,a modified SQRM- SIC （MSQRD-SIC） algorithm is presented. The analysis of complexity and computer simulation demonstrates that two algorithms achieve good compromise between complexity and detection performance by adjusting some specific parameters. Furthermore ,the MSQRM-SIC al- gorithm significantly improves detection performance compared with the SQRM-SIC algorithm, and efficiently reduces detection complexity without obvious performance loss compared with QRD-M algorithm.