基于叠加训练序列的MIMO信道估计及其性能分析  

MIMO channel estimation and performance analysis using superimpose training sequences

作　　者：肖海林[1] 欧阳缮[1] 聂在平[2] 陈紫强[1] 

机构地区：[1]桂林电子科技大学信息与通信学院,广西桂林541004 [2]电子科技大学电子工程学院,四川成都610054

出　　处：《系统工程与电子技术》2009年第1期18-20,82,共4页Systems Engineering and Electronics

基　　金：国家“973”高技术计划基金(2008CB317109);国家“863”高技术计划基金(2002AA123032);国家自然科学基金(60572054);桂林电子科技大学博士启动基金(Z20706);广西信息与通讯技术重点实验室(10903)资助课题

摘　　要：提出一种基于叠加训练序列的MIMO信道估计模型,该模型信道估计方法不占用额外的信号带宽并具有较高的估计精度。通过最小二乘(LS)方法,推导出模型信道估计误差的均方差、信道容量的下限以及更加紧凑的Cramer-Rao界。数值模拟结果表明,相同条件下,采用叠加的训练序列要比直接采用训练序列对系统容量的改善5 dB左右。此外,模型所采用的算法结构简单,计算量小,有很大的实际应用前景。An MIMO channel estimation model using superimposed training sequences is proposed, which doesn＇t take any extra signal spectral and has high accuracy. Through exploiting the least square method, a closed-form solution for the estimation of variance and the lower bound of channel capacity, as well as a rather compact Cramer-Rao bound are derived. Simulation results show that the system has higher performance by using superimposed training sequences than that using training sequences, which is improved about 5 dB. Moreover, the algorithm has a simple structure and low computation complexity, which is very promising in the reality wireless communication.

关 键 词：Cramer—Rao界 信道估计 MIMO 信道容量 叠加训练序列 

分 类 号：TN911.3[电子电信—通信与信息系统]