Study of Chaotic Spread-Spectrum Sequences Based on Neural Networks
基于神经网络混沌扩频序列的研究

The chaos generation neural network based on the excellent learning ability and synaptic weight database are built to generate many chaotic spread-spectrum sequences trained by the modified back-propagation algorithm with various discrete chaotic time series. The chaotic sequences are very easily generated by changing weights of neural network model, and their number is large. The computer simulation results show that, the output chaotic sequences have good correlation property, balance property and linear complexity, therefore they are good candidates for the optimal encrypting code and the spread spectrum code.