Multi-ringlike volumes and offset of Hopfield neural networks based on a discrete memristive self-synapse: Multi-ringlike volumes and offset of Hopfield neural networks: Y. Wei et al.

Recently, memristive Hopfield neural networks (MHNNs) have received considerable attention in designing chaotic systems with complex dynamics because of their distinctive connectivity structures and memory properties. However, the literature rarely reports on the integration of discrete memristors with HNNs and the subsequent observation of their dynamical behavior. Based on the synaptic plasticity of memristors, this paper constructs a novel three-dimensional discrete memristive Hopfield neural network (3DDMHNN) model based on discrete memristor simulating self-synapse. Such a new neural network can exhibit chaotic attractors with different topological structures. Analysis reveals that the phase diagrams produced by the 3DDMHNN display several phenomena, including the formation of multi-ringlike volumes, attractor growth, hyperchaotic attractors, periodic offset behavior, and infinite coexistence of initial value offset behavior. To confirm the physical existence and feasibility of the 3DDMHNN, a simulation circuit incorporating a discrete sinusoidal memristor and the 3DDMHNN was designed. Meanwhile, the 3DDMHNN is realized based on a DSP hardware platform. With the pseudo-randomness of the 3DDMHNN confirmed by relevant tests, the potential directions for its future applications have been significantly broadened. [ABSTRACT FROM AUTHOR]