基于人工神经网络的数字调制信号的快速识别方法

Method of Fast Recognition of Digital Modulation Signals Based on ANNs

基于人工神经网络的理论,提出了快速识别数字调制信号的方法:1)BP神经网络量化共轭梯度算法;2)径向基(RBF)神经网络法;3)小波系数的BP和RBF神经网络法。这些方法收敛速度快,性能好,仿真结果说明了这些方法的有效性。

The methods of fast recognition about digital modulation signals based on ANNs are presented in this paper. The first is using the scaled conjugate gradient algoritbm in the BP neural network; the second is using RBF neural network; the third is using wavelet coefficients in the BP and RBF neural network. These methods own fast convergency and good performance. The result of simulations proves the effectiveness of the methods.