基于总体最小二乘准则的OFDM系统时变信道估计

Time-varying Channel Estimation Based on Total Least Squares Criterion in OFDM Systems

信道估计的准确程度直接影响正交频分复用系统的性能。为了提高时变信道估计算法的精度,基于总体最小二乘准则(TLS)提出了一种时变信道的估计方法。该方法用线性模型对时变信道进行建模,不仅考虑了信道噪声,同时也兼顾了模型误差。该方法能较好地跟踪信道的变化,显著消除模型误差。仿真结果表明所提算法的均方误差介于最小二乘算法与最小均方误差算法之间,在不同归一化多普勒频移下,该算法具有较好的稳健性。

The performance of orthogonal frequency division multiplexing（ OFDM） systems depends on theaccuracy of channel estimation. In order to improve the estimation accuracy,a time-varying channel estimationmethod applying the total least squares （ TLS） criterion is proposed. By modeling the time - varyingchannel for piece-wise linear model,the errors will inevitably bring. A novel method for estimating channelparameters varying within an OFDM symbol is presented. The proposed method takes both channel noiseand the error of piece-wise 1 inear model into account. The method can better track channel change and eliminatethe error of channel model. Simulation results show that mean squared error（ MSE） of the proposedmethod is between that of least square algorithm and least mean square algorithm,and the new method ismore robust with the increasing of the normalized Doppler frequency.