一种基于神经网络混沌序列的对称分组加密方法  

A Symmetric Block Encryption Scheme Based on Chaotic Series of Neural Networks

作　　者：刘年生[1] 郭东辉[1] 吴伯僖[1] Gerard Parr 

机构地区：[1]厦门大学技术物理研究所,厦门361005 [2]英国Ulster大学

出　　处：《计算机工程与应用》2002年第16期49-54,共6页Computer Engineering and Applications

基　　金：国家自然科学基金(编号:69886002;60076015);福建省自然科学基金(编号:A0010019)

摘　　要：根据混沌神经网络的特性如混沌行为特性、对初始条件的敏感性和并行处理特性等等,提出了一种新的对称加密方案;利用混沌神经元的联结矩阵和初始条件作为密钥,使神经网络产生的难以预测的混沌序列来实现非线性数字序列加密运算。从理论分析和仿真结果表明,该系统具有较好的实际保密性,加密效率高,加密速度快,比较适用于数字信号加密通信。A new symmetric block encryption scheme based on the properties of chaotic neural networks,which mainly include chaotic behavior,sensitivity to initial conditions,and parallel feature is described in this paper.The synaptic ma-trix and initial conditions of chaotic neural networks are regarded as symmetric key.Every block of plain text is en-crypted by chaotic sequence generated by neural networks.The results from theoretical analysis and simulation show that the scheme has a better practical security,the efficiency of its encryption is high,and the rate of its encryption is fast.It is well to encrypted communication of digital signal.

关 键 词：神经网络 混沌序列 对称分组加密方法 对称加密 加密通信 

分 类 号：TN918[电子电信—通信与信息系统] TP183[电子电信—信息与通信工程]