联合ML和OSIC的格规约辅助检测算法  

Lattice Reduction Aided Detection Combined ML and OSIC

作　　者：王振朝[1,2] 朱春燕[1] 

机构地区：[1]河北大学电子信息工程学院,河北保定071002 [2]河北省数字医疗工程重点实验室,河北保定071002

出　　处：《小型微型计算机系统》2015年第4期764-767,共4页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61074175;61203160)资助;河北省自然基金项目(F2014201168)资助

摘　　要：为了提升高信噪比情况下次优检测算法的检测性能,在一层遍历搜索的排序串行干扰消除算法(ML-OSIC)的基础上引入格规约技术,提出一种新的格规约辅助检测算法(ML-LLL-OSIC).该算法首先对噪声干扰最小的子信道信号进行遍历检测,然后利用格规约技术处理剩余子信道矩阵以改善信道矩阵的正交性,最后利用改进的信道矩阵进行排序串行干扰抵消检测.上述算法既减少了遍历检测搜索次数,又改善了信道矩阵的正交性,可更加有效地避免层间干扰和误码扩散.仿真结果表明,在多用户多入多出系统(MU-MIMO)中,QPSK调制方式下信噪比大于10d B时,或16QAM调制方式下信噪比大于14d B时,本算法可以获得接近ML算法的检测性能.In order to improve the detection performance of suboptimal detection algorithm in the high SNR,This paper introduces the lattice technology based on ML-OSIC, and presents a new lattice reduction aided detection algorithm （ ML-LLL-OSIC ）. The algorithm first traverses the best noise detection sub-channel signals, then the lattice processing the remaining channel to improves the orthogonal- ity of the channel matrix,finally the OSIC algorithm on the basis of new channel matrix. This algorithm reduces search times, while improves the orthogonality of the channel matrix and more effectively avoids the interference between layers,and error diffusion. The simulation results show that the algorithm can get the same performance as ML detection algorithm in MU-MIMO system when the SNR is greater than 10dB in QPSK modulation mode or the SNR is greater than 14dB in 16QAM modulation mode.

关 键 词：多用户MIMO系统 遍历检测 排序串行干扰消除检测 格规约 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]