基于离散正则模型的时变信道盲辨识算法

以CDMA时变信道离散正则模型为基础,提出了基于改进的矩阵外积分解的时变信道盲辨识算法。算法对传统的矩阵外积分解算法进行修正,使其适应离散正则模型两级盲辨识,并引入了精确的时延阶数估计,克服了现有算法需要预先知道信道时延实际阶数的局限,使其在仅知道信道阶数上界的条件下完成盲辨识。仿真结果表明了该算法的有效性。

Blind Identification of Time-varying Channels Based on Discrete-time Canonical Model

An improved matrix outer-product decomposition method for blind identification of time-varying channels was proposed based on the discrete-time canonical model of CDMA time-varying channels. This algorithm modified the conventional matrix outer-product decomposition method to adapt to the two-stage blind identification of the discrete-time canonical model. Furthermore, a more accurate order determination of time delay spread was used to overcome the disadvantage of the existing methods which required the prior knowledge of the actual order. The new method can do well if the upper bound of the channel time delay spread is known. Simulation results prove the efficiency of the algorithm.