空间调制系统中改进的MML检测算法

An improved MML detection algorithm

在空间调制(Spatial Modulation,SM)系统中,基于M算子的ML(M-algorithm to ML,MML)信号检测算法采用逐层减少搜索分支的方法降低最大似然(Maximum Likelihood,ML)检测算法的计算复杂度.针对MML算法进行了研究,提出了i-MML算法,该算法在保证检测性能不变的情况下降低了计算复杂度.i-MML算法与MML算法的不同之处在于对最后一层保留节点累积度量的处理,i-MML算法将最后一层所有保留节点的累积度量与上一层的累积度量进行比较,判断是否需要对其进行计算.因此,给出的i-MML算法在最后一层不需要计算所有保留节点的累积度量就可以得到检测结果.通过仿真分析可知,i-MML算法与MML算法的检测性能一致,但i-MML算法一定程度上降低了计算复杂度.

In a spatial modulation(SM)system,M-algorithm to ML(MML)signal detection algorithm reduces the computational complexity of the Maximum Likelihood(ML)detection algorithm by decreasing the search branches layer by layer.In order to further reduce the computational complexity as well as ensuring the detection performance,this paper proposes the i-MML algorithm via analyzing MML algorithm.The difference between these two algorithms lies in the processing of the accumulated metrics of the reserved nodes in the last layer.The i-MML algorithm will compare the accumulated metrics of all the reserved nodes in the last layer with those in the previous layer to determine whether they need to be calculated.Therefore,the i-MML algorithm proposed in this paper does not need to calculate the cumulative metrics of all the reserved nodes in the last layer to get the test results.Through the simulation results,it is clear that the detection performance of the novel i-MML algorithm is the same as that of the MML algorithm;however,its computational complexity is reduced to some extent.