低复杂度的格约减辅助MIMO检测算法  

Low-complexity lattice reduction aided MIMO detection algorithm

作　　者：刘金铸[1] 翁秀娟[1] 

机构地区：[1]南京信息工程大学电子与信息工程学院

出　　处：《计算机工程与设计》2016年第7期1709-1713,共5页Computer Engineering and Design

基　　金：国家自然科学基金面上基金项目(61372128)

摘　　要：为降低MIMO检测算法的计算复杂度并保证系统性能,在格约减辅助MIMO检测算法的基础上,引入噪声投影按序逐次消去(OSNPC)算法,采用最小均方误差(MMSE)准则,构成改进的LLL-MMSE-OSNPC算法。OSNPC算法从判决变量向量中逐次消去已知的噪声投影,避免排序干扰,逐步消去(OSIC)算法中对矩阵的求逆运算,降低复杂度。以QPSK和16QAM调制方式为例进行理论分析和仿真,其结果表明,LLL-MMSE-OSNPC算法在性能上等效于LLLMMSE-OSIC算法,但复杂度大为降低。To reduce complexity of the detection algorithm in MIMO system and guarantee the performance of the system,an im-proved algorithm called LLL-MMSE-OSNPC was proposed.Ordered successive noise proj ection cancellation （OSNPC）algorithm was introduced on the based of LR-aided minimum mean square error （MMSE）detection.The OSNPC algorithm successively cancelled the known noise proj ections from the decision statistic vector.The inverse of the matrices in the ordered successive in-terference cancellation （OSIC）algorithm was avoided and the complexity was reduced.Taking QPSK and 1 6 QAM modulation means as examples,results of theoretical analysis and simulation show that the performance of LLL-MMSE-OSNPC algorithm is equivalent to that of the LLL-MMSE-OSNPC algorithm,but it has significantly less complex computation.

关 键 词：格约减 最小均方差 噪声投影按序逐次消去 排序干扰逐步消去 最大似然 

分 类 号：TP911.23[自动化与计算机技术]