MIMO-OFDMA系统中基于奇异值分解的信道估计  

Channel Estimation Based on SVD for MIMO-OFDMA Systems

作　　者：卢鑫[1] 王建华[1] 

机构地区：[1]深圳信息职业技术学院信息技术研究所,广东深圳518029

出　　处：《无线电工程》2008年第10期17-19,共3页Radio Engineering

摘　　要：基于最小二乘方(LS)信道估计技术,并结合MIMO-OFDMA系统中导频的特点,提出一种基于时域LS的奇异值分解(SVD)信道估计法,该方法是对T矩阵进行奇异值,而T矩阵要比信道自相关矩阵R容易获得得多。理论研究与仿真结果显示,该算法的性能略劣于Ove Edfors等人提出的基于线性最小均方误差估计(LMMSE)的SVD信道估计方法,但计算复杂度要低,并且对导频的要求比较宽松,具有广泛的应用性。This paper presents a channel estimation algorithm by singular value decomposition （SVD） for MIMO-OFDMA systems. According to the defined downlink transmission and symbol pilot structure, the proposed channel estimation method based on least square （L S） is intended as a improved algorithm for Ove Edfors＇ , which will more easily get singular value based on matrix T than that on autocorrelation matrix R. Simulations and analysis show that the performance of the novel method is slightly inferior to those for Ove Edfors＇ LMMSE-SVD algorithm, however it provides less complexity and more universal application without strict pilot structure conditions.

关 键 词：MIMO OFDM LS LMMSE 奇异值分解 

分 类 号：TN941.4[电子电信—信号与信息处理]