无线测试系统中CDMA时变色散信道的盲辨识算法

Blind Identification of CDMA Time-varying Dispersive Channels in Wireless Test System

本文针对无线移动测试系统中信道的传输问题,在CDMA时变色散信道的离散正则模型的基础上,给出了一种新的基于二阶统计量的算法来完成对CDMA时变色散信道的盲辨识。利用正则模型,可以将时变的SIMO系统等效的转化为是不变的MIMO系统,使用时频解相关接收机从时变的特征波形中分离出时不变的信道系数后,可由解相关输出后信号的自相关矩阵的主特征向量直接完成对信道的估计。与基展开模型方法相比,该算法具有不需要利用观测数据的高阶统计量来估计信道的展开基频率,也不必假设相邻符号间的时延的变化为线性关系,同时其运算复杂度小于通常的子空间算法。仿真结果证明了算法的有效性。

A new blind identification algorithm for code-division multiple-access （CDMA） time-varying dispersive channels is proposed, which based on the discrete-time canonical model （DTCM） and exploiting the second-order statistics of output signals. With DTCM, a time-varying single-input, multi-output system （SIMO） can be transfered into a time-invariant multi-input, multi-output （MIMO） system. So the time-varying channel coefficients can be abstracted from the time-invariant signature waveform through a time-frequency decorrelating receiver and estimated with the principle eigenvector of the output autocorrelation matrix. Compared with the methods that based on the basis expansion model （BEM） , the new algorithm doesnt require the estimate of the basis frequencies with higher-order statistics of output signals, nor need the assumption of linearly varying delays across symbols, and is more computationally efficient. Simulations demonstrate the effectiveness of the algorithm.