基于矢量平均的信道估计算法

提出一种适用于DS-CDMA无线通信系统的低计算量而性能良好的信道估计算法。在DS-CDMA系统中，当接收信号经过解扩或多用户检测等时域预处理后，可以认为干扰大大减弱、期望信号在处理后的信号中占主要地位。这样，用预处理后信号相关阵的最大特征值对应的特征矢量可以很好地近似期望信号的信道矢量。但是，直接特征分解需要很大的计算量，特征跟踪计算量较低，但瞬态性能较差。本文提出一种基于相关矩阵列矢量平均的信道估计算法，该算法不需要特征分解或跟踪。仿真结果表明：新算法在降低计算量的情况下可以获得同直接特征分解方法几乎相同的性能。

Channel estimation algorithm based on vector average

A low computational complexity and efficient channel estimation method is proposed for DS-CDMA wireless communication systems. After the received signal of the DS-CDMA systems is pre-processed with despread or multiuser detection in the time domain,the interference signal can be significantly reduced and the desired signal can be regarded as the dominating component in the despreaded signal,thus the channel vector of the desired signal can be well approximated with the maximum eigenvector of the autocorrelation matrix of the processed signal. However,direct eigendecomposition needs heavy computational burden and eigenvector tracking with 1 ow complexity has worse transient state performance. A new method based on column vector average of the autocorrelation matrix without eigendecomposition is presented to estimate the channel vector. Simulation results indicate that the performance of the new method with low complexity is very close to that of the method with direct eigendecomposition.