多层线性神经网络自适应滤波器性能仿真

Performance Simulation of Adaptive Filters Based on Multi-layer Linear Neural Networks

介绍了一种基于多层线性神经网络的自适应滤波器。仿真结果表明，将这种基于ＢＰ算法训练的滤波器应用于扩频接收机的干扰抑制，比常规的采用ＬＭＳ算法训练的滤波器具有更优的性能。文中还讨论了滤波器层数对于ＢＰ算法收敛过程的影响。

An allptive filter on the layer neund network is Ptdeed in this.Simulation results that this filter trained with back Pron has better perfonce than theC adaptive filter teaing with the LMS algorithm When it is applied tO intetheence suPPt in a PNsprat spetrum receiver. The influence of the number of on BP convergemp behaviour is alsothe in ac Paper.