基于正则总体最小二乘准则的OFDM系统信道估计算法

基于导频的最小二乘(LS)估计方法的优点是结构简单,容易实现,但对噪声和干扰敏感。通过对OFDM系统中由于信道的时变性产生的载波间干扰进行了分析,在此基础上提出正则总体最小二乘(TLS)信道估计算法。根据Tikhonov正则化原理,导出了总体最小二乘正则化解法的计算公式。与一般的LS信道估计相比,TLS方法同时考虑了信道噪声和信道时变特性,所以该算法不仅在一般的信道条件下具有良好的性能而且也适合快变信道条件下的OFDM系统。仿真结果表明,该算法具有很好的均方误差和误比特率性能。

Algorithm of Channel Estimation Based on Regularized Total Leas-squares in OFDM System

The greatest advantage of least-squares （LS） estimation algorithm based on pilot is its simple structure, complexity is low, but the accuracy of LS algorithm is limited due to ignoring the effect of noise and interference. The inter-carrier interference caused by time-variant fading channels in OFDM systems was analyzed and a regularized total least-squares （TLS） channel estimation algorithm was proposed. The formulas of regularized total least squares solution are derived based on Tikhonov regularization criterion. Compared with the common LS channel estimation,TLS channel estimator takes into account channel noise and the character of time-varying of channel. The proposed algorithm could provide more accurate channel information not only in ,slowly time-variant environments, but more important also in fast time-variant environments. The simulation results confirm good MSE and BEIK performance of the algorithm.