Generating grid double-scroll attractors from magnetized SC-CNN and its application in image encryption

Abstract

Cellular neural network has attracted widespread attention due to its easy hardware implementation and general application in information processing. In this paper, a new memristive State-Controlled-CNN (MSC-CNN) model is established by introducing two novel flux-controlled memristors into a three-cell network, where one of which simulates the electromagnetic radiation and the other characterizes the coupling weight. The present MSC-CNN can generate multi-double-scroll attractors (MDSAs) in one direction or grid multi-double-scroll attractors (GMDSAs). Stability analysis shows that each asymptotically stable equilibrium point of the memristors serves as index-1 saddle-focus in the MSC-CNN, which induces the corresponding connecting bond orbits between two scrolls. Thus, the number of GMDSAs can be flexibly adjusted by varying the number of the asymptotically stable equilibrium points of the memristors. Moreover, the memristive coupling strength also controls the number of MDSAs. Through numerical simulations, complex dynamic behaviors such as extreme multistability and amplitude control are also observed. Then, an image encryption scheme based on the proposed MSC-CNN is designed from the application perspective. Finally, a microcontroller-based hardware system is implemented to verify the effectiveness of numerical simulations.