摘要
针对目前图像加密方法中密钥值选择单一、混沌系统效率低等问题;提出了一种基于忆阻器神经网络和改进Logistic映射的图像加密算法.该算法引入指数函数对一维Logistic映射进行改进,选择忆阻器与Chebyshev混沌多项式结合作为激活函数的神经网络,将神经网络更新的权值均衡化后作为混沌系统的初始值;采用混沌系统获得替换索引矩阵,完成对图像像素级,以及bit级的置乱操作;使用两组随机序列对置乱后的密文进行两轮方向相反扩散操作,完成图像加密.神经网络中的权值作为混沌系统的初始值选择与更新的密钥源,生成的混沌序列经NIST等检验证明了其具有较好的随机性;安全性分析表明该算法密钥空间大,并且可以抵抗统计分析攻击,具有较高的安全性.
Aiming at the problems of single key value selection and low efficiency of chaotic systems in current image encryption methods,an image encryption algorithm based on memristor neural network and improved Logistic map is proposed.The algorithm introduces an exponential function to improve the one dimensional Logistic map,selects the neural network combining the memristor and the Chebyshev chaotic polynomial as the activation function,and equalizes the updated weight of the neural network as the initial value of the chaotic system;The chaotic system is used to obtain the substitution index matrix,and the im⁃age pixel-level and bit-level scrambling operations are completed;two sets of random sequences are used to perform two rounds of reverse diffusion operations on the scrambled ciphertext to complete image encryption.The weight in the neural network is used as the key source for the selection and update of the initial value of the chaotic system.The generated chaotic sequence has been verified by NIST and other tests to prove that it has good randomness;Security analysis shows that the algorithm has a large key space and can resist statistical analysis attacks,and has high security.
作者
李涵
葛斌
LI Han;GE Bin(School of Computer Science and Engineering,Anhui University of science and Technology,Huainan 232001,China)
出处
《通化师范学院学报》
2021年第12期108-115,共8页
Journal of Tonghua Normal University
基金
国家自然科学基金项目(51874003,61703005)
安徽省自然科学基金项目(1808085MG221).
关键词
密码学
图像加密
混沌映射
忆阻器神经网络
cryptography
image encryption
chaotic map
memristor neural network
