基于径向基神经网络的混沌加密方案

Chaotic Encryption Based on RBF Neural Networks

提出了一种基于径向基(RBF)网络的混沌序列产生方法,并基于这种模型提出了一种新的混沌加密方案。计算机仿真证明利用RBF网络良好的逼近任意非线性映射和处理系统内在的难以解析表达的规律性的能力,及快速的收敛速度,在统一的系统结构下通过权值的切换方式(即用不同的混沌映射)可产生比单一混沌映射更多的、性能更接近理论值的混沌序列,同时基于该模型的混沌加密方案具有高度的保密性和灵敏性。

A new chaotic sequence generation method,based on the RBF neural network,which has the strong learning ability and nonlinear function approximation capacity and a new chaotic encryption with this method are proposed in this paqer.Experimental results show that this scheme can very easily generate much more chaotic sequences with desirable statistical properties than single chaotic map by changing weights of this RBF neural network and this new chaotic encryption has higher security and sensitirity.