FN-HNN Coupled With Tunable Multistable Memristors and Encryption by Arnold Mapping and Diagonal Diffusion Algorithm

With the rapid development of intelligent information technology, it is significant to construct neural network models that conform to biological characteristics. In this paper, a memoristor model with tunable multistable properties is proposed. By changing the memory parameters, the number of multistable states can be adjusted. Based on the memoristor, an asymmetric memristive FN-HNN neural network (MFNHNN) containing five neurons is constructed. The fundamental dynamical theories, such as equilibrium points, bifurcation diagrams and Lyapunov exponents, are used to reveal the complex dynamic behaviours of MFNHNN. The different dynamic behaviors with coupling intensity control, the tunable coexistence of infinite chaotic attractors and the coexistence of initially controlled chaos and periodic attractors are observed. Furthermore, the equivalent circuit of MFNHNN is implemented. On the basis of random chaotic sequences, an image encryption scheme combining Arnold mapping and diagonal diffusion algorithm is proposed. The findings indicate that the proposed scheme exhibits superior encryption performance, rendering it suitable for application in remote sensing information security.