混沌加密技术应用于信息加密是近年信息安全研究的热点问题;首先概括了GRNN神经网络的主要理论,然后将GRNN神经网络和混沌系统相结合建立了一种混沌序列的生成模型,并给出基于GRNN的混沌图像加密算法,该算法只需改变混沌系统的初值,便可获得不同的混沌序列,进而实现对图像数据进行置换加密;最后对算法进行计算机仿真实验;结果表明GRNN神经网络具有更快的收敛速度和更准确的逼近能力,而基于GRNN神经网络的混沌加密算法具有很高的安全性。
Chaotic encryption presently becomes the focus of information security area.Firstly,this paper gives the main theory of GRNN,and proposes a model of chaotic series based on GRNN and chaotic system.Then it gives chaotic image encryption algorithm based on GRNN that can get different chaotic series by varying the system initial value.According to the sort permutation and inverse permutation from the chaotic series,it can achieve image encryption and decryption.Finally,through computer simulation,the experiment results show that GRNN has faster convergent and more exact approximation.This new chaotic encryption algorithm has higher security.