Image Encryption Algorithm Based on Memristive BAM Neural Networks

Jiuhong Xiao, Weiping Wang, Meiqi Wang

Abstract: This paper concerns with the application of memristive bidirectional associative memory neural networks (MBAMNNs) for image encryption. Firstly, we propose a new model of MBAMNNs with leakage delays and probabilistic time-varying delays. Furthermore, we design a color image encryption algorithm based on this model. The algorithm shows high efficiency, and it has large key space. It can also resist strong atttacks including differential attack, exhaustive key search, etc. Finally, the effectiveness of the proposed algorithm is validated with numerical experiments.