中文说明：基于互功率谱的时延估计，关于神经网络控制，用于特征降维，特征融合，相关分析等,对HARQ系统的吞吐量分析，D-S证据理论数据融合，实现了对10个数字音的识别程。
English Description：
Based on the cross power spectrum time delay estimation, a neural network control for feature dimensionality reduction, feature fusion, correlation analysis, analysis of the throughput of the HARQ system, D-S evidence theory data fusion, the realization of the 10 digital audio recognition.