多入多出系统中一种新的部分极大似然算法

A New Partial Maximum Likelihood Algorithm for MIMO Systems

提出了一种新的应用于多入多出(MIMO)系统的结合最小均方误差串行干扰消除(MMSE-SIC)的部分极大似然(ML)算法.该算法对一部分符号进行ML估计,并结合MMSE-SIC算法对其他符号进行检测,最后选择与接收信号欧式距离最小的一组符号向量作为解调输出.仿真结果表明,在4发4收V-BLAST系统中BER=10-2处,所提出算法比传统的MMSE-SIC算法有大约4 dB的性能增益,甚至接近于ML算法的性能,而且对于空间相关也具有较强的鲁棒性.

This paper proposed a new partial maximum likelihood algorithm for multiple-input multiple-output （MIMO） systems. The algorithm makes maximum likelihood （ML） estimation for part of the signals, meanwhile it combines MMSE-SIC to detect other signals. The simulation shows that in 4 by 4 V-BLAST systems, the new scheme can achieve about 4dB performance gain over the conventional MMSE-SIC algorithm at BER=10^- 2, and it even approaches the ML performance. Furthermore, the proposed algorithm has good robustness against spatial correlation.