鲁棒的基于几何均值分解的THP预编码算法  

Research on Robust GMD-Based Tomlinso-Harashima Precoding

作　　者：袁赛[1,2] 朱瑛[1] 王德辉 张胜磊 刘景泉 YUAN Sai ZHU Ying WANG De-hui ZHANG Sheng-lei LIU Jing-quan(School of Electronic Engineering, Xidian Univ. , Xi＇an 710071, China Department of Teachingand Research, Army Communications Training Base, Zhangjiakou 075100, China The Information Network Management Center of Aerospace System department,Beijing 100720,China Zhangjiakou intermediate peopleg court,Hebei 075000,China)

机构地区：[1]西安电子科技大学电子工程学院,陕西西安710071 [2]陆军通信训练基地教研部,河北张家口075100 [3]航天系统部信息网络管理中心,北京100720 [4]张家口市中级人民法院,河北张家口075000

出　　处：《中国电子科学研究院学报》2017年第1期96-99,共4页Journal of China Academy of Electronics and Information Technology

摘　　要：在无线多输入多输出(MIMO)系统中,由于接收端不能完全准确的估计信道矩阵,因此会产生估计误差,本文在非理想信道模型下,提出了一种鲁棒的基于几何均值分解方法模代数预编码算法。该算法在MMSE准则下,利用信道误差矩阵的二阶统计量及误差与信号正交的方法,推导出需要几何均值分解的矩阵,进而可求出前馈矩阵、接收矩阵、加权矩阵及反馈矩阵。通过实验仿真表明该鲁棒算法优于传统算法性能,在高信噪比时,误码率可提高4 dB。In the wireless multiple-input multiple-output（MIMO） system,the receiver can not estimate the channel matrix completely and accurately.A robust Tomlinso-Harashima precoding（THP） algorithm based on geometric mean decomposition（GMD） method was proposed with imperfect channel state in-formation（CSI）.The geometric mean decomposition matrix is received by exploiting statistical information of CSI and minimizing the mean square error of sig-nal vectors transmitted.The feedforward matrix,the receiving matrix,the weighting matrix and the feedback matrix are obtained.The simulation results show that the robust algorithm is superior to the traditional algorithm,and the BER can be improved at high SNR.

关 键 词：多输入多输出系统(MIMO) THP预编码 非理想信道状态信息 GMD预编码 

分 类 号：TN911.72[电子电信—通信与信息系统]