研究了V—BLAST系统中基于QR分解的几种信号检测算法,分析了各种算法的优缺点,指出信号检测顺序是降低误差传播概率的关键。针对传统检测算法计算量大的缺点,采用基于改进的Gram—schmidt正交化的排序QR算法,该算法利用迭代运算代替矩阵求逆运算,从而有效地降低了计算量,并把该算法和MMSE—SIC算法结合进行了比较。仿真表明,该算法能够使系统在复杂度和性能之间取得很好的折中。
In this paper, several signal detection algorithms based on QR decomposition as applied to V-BLAST system are studied and their characteristics are analyzed, and it is a key to reduce errror probability by sorting the detection sequence, sorted QR decomposition of the channel matrix base on improved Gram-sehmidt is used. This algorithm uses iterative operation instead of contrary matrix operation, so the computational complexity of the conventional detection algorithms is effectively reduced, and the performance between this algorithm and MMSE algorithm by using successive interference cancellation (SIC) structure is compared. Simulation shows that this algorithm can gain the good tradeoffs between the performance and the complexity.