Studies On Signal Detection Algorithms For MIMO System

MIMO technology is the key technology for future wireless communications, which has been used in 3G, WiMAX, HSPA and LTE and other wireless communications systems. In MIMO system, signal detection is an important part, and is also the focus of this study. This paper firstly studies the channel transmission characteristics, channel capacity, and some classical detection algorithm of MIMO communication system. Second, after the in-depth analysis of a classic Sphere Decoding (SD) algorithm it shows:in the traditional SD, it is the loose conditions of the first several layer radiuses that led to large computational complexity. In this paper, it proposes three SD algorithms to aim to improve the loose of layer radiuses:(1) It proposes a statistical pruning SD based on the expectations of chi-square distribution (EPSD). In the classical SD algorithm, the root cause of the loose layer radiuses is:the enlarged treatment in solving the layer radiuses. This article uses the expectations chi-square distribution as an estimation of the enlargement part, and the improved algorithm is simple, but also greatly reduces the computational complexity, at the same time the performance loss is small.(2) It proposes a statistical pruning SD algorithm based on the maximum value of chi-square distribution (MPSD). The algorithm has the same detection thinking like EPSD algorithm; just the estimation used the maximum of chi-square distribution. MPSD is a compromising algorithm between detection performance and computational complexity, which has a very small performance loss.(3) It proposes a layered statistical pruning algorithm based on variance of the noise samples (LPSD). The main idea of the LPSD is to average the noise variance samples by layer, and then treat the results as a basis for further pruning. The LPSD algorithm has low computational complexity and good detection performance; at the same time it has a stronger ability to adapt.In the article, a large number of simulation experiments for different detection algorithms were carried out, and the results were analyzed. Experimental results show that the improved methods are effective.