基于贪心策略的大规模MIMO系统信号检测算法

针对传统球形译码性能和计算复杂度受到初始半径及搜索策略制约的问题,提出了一种新的基于M算法的贪心策略球形译码检测算法,对树搜索的方法进行了改进,先将该层信号集合中的距离增量进行排序,然后选择距离增量最小的M个点为信号点,这样每一次选取的信号点相对该层都是局部最优的.仿真结果表明,相比于传统球形译码检测算法,当M为1时,该算法可以降低约30％的计算复杂度.使球形译码算法的效率得到了很大的提高,可以运用于大规模MIMO系统中.

Signal Detection Algorithm Based on Greedy Strategy in the Large-scale MIMO System

In view of the problem that traditional sphere decoding performance and the computational complexity are restricted by the initial radius and the search strategy, a new greedy strategy sphere decoding detection algorithm based on M algorithm is presented in this paper and the tree search method is improved. Firstly, the distance increment of the signal set in the layer is ordered, and then M points with incremental distance minimum are selected as signal points, so that each selected signal point is locally optimal in the layer. Simulation results show that, compared to the traditional sphere decoding algorithm, the computational complexity can be reduced about 30% by the algorithm when the M is 1. The efficiency of the sphere decoding algorithm is greatly improved and can be also used to the massive MIMO.