基于复数格基缩减的V-BLAST系统检测算法及性能分析

为了提高系统性能,从实数格基缩减出发,推广得到了一种通用的复数格基缩减算法,给出了其在V-BLAST系统中的应用。通过与MMSE扩展相结合,分析缩减与扩展的先后顺序,提出了MMSE-Ⅰ和MMSE-Ⅱ两类基于复数格基缩减算法的V-BLAST检测方案。分析和仿真结果表明,检测中采用复数格基缩减算法与实数算法相比,可以大幅降低复杂度而不影响系统性能;基于复数格基缩减的MMSE-Ⅱ逐层干扰抵消检测算法,可以解决格基缩减引起的有效发送符号之间的相关性问题,且性能接近于ML检测。为了进一步降低复杂度,给出了一种性能损失较小的次优排序方案。

Detection algorithm and per formance analysis of V-BLAST system based on complex lattice basis reduction

To Dramatical ly improve the performance o f MIMO system, a universal complex Lat t ice basis reduct ion( LBR) algo rithm w as pr opo sed based o n real lat t ice basis reduction( LBR) and its applicat io n to V-BLAST sy stem w as also provided. Combining w ith MMSE extension, two types of V-BLAST detection algo rithm based o n complex LBR, namely, MMSE-I and MMSE-II, were proposed by analy zing the o rder of the reduct ion and the ex tensio n. Analy sis and simulat ion result s show that compared w ith real LBR algo rithm, the complex one can great ly reduce the complexity w hile keep the same per formance, and that MMSE-II successive interference cancellatio n ( SIC ) detect io n alg orithm based on complex LBR can deal w ith the dependency o f ef fect ive t ransmit t ing symbols caused by LBR and achiev e near-ML detection performance. To further low er the complex ity, a sub-o pt imum ordering scheme w ith lit tle performance degr adat io n w as prov ided