Abstract: Based on matrix eigenvalue decompostion, an effective blind multichannel identification algorithm is proposed in this paper. Different from the Prediction Error Method, the new algorithm does not require the input signal to be independent and identical distribution, and even the input signal can be non-stationary. Compared with Least-Square Approach, the new algorithm is more robust to the overestimation of channel order. Finally,the experiments prove the good performance of the proposed algorithm.

Key words: Blind channel identification, SIMO, Eigenvalue decomposition

摘要： 本文提出了一种有效的基于矩阵特征值分解的多路信道盲辨识算法。与基于预测误差盲辨识方法不同，新的算法不要求输入信号独立同分布，甚至输入信号可以是非平稳的。与最小均方盲辨识算法相比，这种新算法对信道时延阶数估计有更好的鲁棒性。本文最后实验演示了新算法的有效性。

关键词: 盲信道识别, 单输入多输出, 特征值分解