基于CNN和CML的时空混沌序列图像加密算法

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

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。

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.