基于CNN和CML的时空混沌序列图像加密算法*

基于细胞神经网络(CNN)和耦合映像格子模型(CML),提出了一种密钥长度为128 bit的对称图像加密算法.该算法用具有复杂动力学行为的CNN网络和分段线性混沌映射去驱动CML模型,以快速产生时空混沌序列,并与原始图像异或完成加密过程.算法分析表明具有密钥空间大、对密钥敏感、容易快速实现等特点,并能抵抗穷举攻击、已知明文攻击和唯密文攻击等.数值仿真结果验证了该算法的正确性和有效性.

Image encryption algorithm using spatio temporal chaotic sequences based on CNN and CML

A symmetric image encryption algorithm that has a 128 bit key was proposed based on cellular neural network (CNN) and coupled map lattice (CML) model. Within the algorithm, a CNN that has a complex dynamical behavior and a piecewise linear chaotic map (PWLCM) were used to drive CML model as to quickly generate spatio temporal chaotic sequences, which would be bitwise XORed with original image later. Algorithm analyses indicated that it has some excellent features such as a large key space, the sensitivity to the key, easily and quickly implementation, etc. Furthermore, the algorithm could resist the brute force attacks, the know plaintext attacks and ciphertext only attacks. The computer simulations are carried out and the results show the correctness and effectiveness of the algorithm.