大规模天线系统中基于软判决的MIMO信号检测算法

An algorithm for MIMO signal detection based on soft decision in large-scale antenna system

在大规模多输入多输出(MIMO)系统下,提出了一种基于软判决的改进MMSE(IMMSE)信号检测算法。在IMMSE算法中,把MMSE算法检测值作为算法的初始值并采用迭代干扰消除技术。进一步使用对数最大似然比(LLR)将检测序列进行排序,提出一种有序的IMMSE(OIMMSE),并使用软判决技术来提高算法的检测性能。在不同天线数的MIMO系统下,对IMMSE算法和OIMMSE算法进行误码率性能仿真。仿真结果表明,OIMMSE算法和IMMSE算法性能明显优于MMSE。而且提出的新算法随着天线数的增加,越来越接近单输入单输出(SISO)在加性高斯白噪声下的性能。由此可见,新算法对大规模MIMO系统是有效的。

In this paper, we propose an improved minimum mean squared error （IMMSE） for detecting the symbol vector in massive MIMO sys- tems. In the IMMSE algorithm, the minimum mean squared error （MMSE） estimate of the received symbol vector are used as an initial solu- tion and the iterative successive interference cancellation technology is aolopted. Then we propose an ordered IMMSE （OIMMSE） algorithm which uses the log likelihood ratio （LLR） based ordering in the detection sequence. And soft decision technique is used to improve the detec- tion performance of the algorithm. The bit error rate （BER） performance of IMMSE and OIMMSE is simulated for different antenna configura- tions in MIMO system. Simulation results show that IMMSE and OIMMSE algorithm performance is better than MMSE algorithm. Furthermore, the performance of the proposed algorithm is improved with increase of antennas and approaches towards single-input single-output （SISO） ad- ditive white Gaussian noise （AWGN） , which proves the effectiveness of OIMMSE algorithm for massive MIMO systems.