将过采样技术和线性独立分量分析（ＩＣＡ）神经网络结合在一起，得出一种新的盲均衡方法。单输出信号通过采样转换为多输出观测信号，从而可以直接利用ＩＣＡ方法得到信道均衡输出。

In this paper, a new blind equalization algorithm is proposed, which combines over-sampling technique with linear ICA neural network. The algorithm converts single-output system into multi-channel system by over-sampling and uses ICA method directly, and then obtain the equalized output of channel.

TN911.5

国家自然科学基金