Dynamical analysis and DSP implementation of discrete memristor coupled discrete bi-neuron Hopfield neural networks

Abstract Among the studies on neural modeling of the human brain, Hopfield neural network (HNN) is more famous and has been widely analyzed and studied. However, discrete HNN has been studied relatively less. The study of coupling into a memristor in a discrete HNN is even rarer. In this paper, a memristor model is proposed to introduce discrete HNN to simulate synapses. And there is crosstalk between the memristors. In turn, a discrete memristor coupled discrete bi-neuron HNN (DMCBHNN) model is constructed. Using various methods such as bifurcation diagrams (BD), Lyapunov exponential spectra (LEs), phase diagrams, and chaotic sequences, the extensive and complex dynamical behavior of the model is highlighted. It includes hyperchaos, chaos, cycles, attractor coexistence, complexity, firing patterns, and state transfer. In addition, the hardware implementation of DMCBHNN is completed by using DSP platform. The feasibility of the system is verified. This provides a basis for studying the complex workings of the human brain.