一种基于Kaczmarz算法的大规模MIMO信号检测方法

A Massive MIMO Signal Detection Method Based on Kaczmarz Algorithm

大规模MIMO系统的上行信号检测算法中,Kaczmarz检测算法可简化求逆过程,但计算量大,复杂度高。为此,利用Kaczmarz算法的强几何解释性质,把部分迭代过程用递推关系代替,进一步降低Kaczmarz算法的计算量。仿真结果表明,在维持其检测性能基本不变的情况下,和Kaczmarz算法相比,所提出的算法至多可降低17%的计算量。

Among the large-scale uplink signal detection algorithms of MIMO system,Kaczmarz detection algorithm can simplify the inverse process,but it needs large computations and is of high complexity.Therefore,it is necessary to use the strong geometric explanatory property of Kaczmarz algorithm and replace some iteration processes with recurrence relations,which will further reduce the calculated quantities of Kaczmarz algorithms.The simulation results show that if the detection performances are maintained to be stable,compared with traditional Kaczmarz algorithm,the proposed algorithm can at most reduce 17%of the calculated quantities.