Synchronization coexistence in a Rulkov neural network based on locally active discrete memristor
Yang Yang,Ma MingLin,Xie XiaoHua,Li ZhiJun,Sun YiChuang
DOI: https://doi.org/10.1088/1674-1056/acb9f7
2023-02-08
Chinese Physics B
Abstract:Abstract At present, many neuron models have been proposed, which can be divided into discrete neuron models and continuous neuron models. Discrete neuron models have the advantage of faster simulation speed and the ease of understanding complex dynamic phenomena. Due to the properties of memorability, nonvolatility and local activity, locally active discrete memristors (LADMs) are also suitable for simulating synapses. In this paper, we use a LADM to mimic synapses and establish a Rulkov neural network model. It is found that the change of coupling strength and the initial state of the LADM leads to multiple firing patterns of the neural network. In addition, considering the influence of neural network parameters and the initial state of the LADM, numerical analysis methods such as phase diagram and timing diagram are used to study the phase synchronization. As the system parameters and the initial states of the LADM change, the LADM coupled Rulkov neural network exhibits synchronization transition and synchronization coexistence.
physics, multidisciplinary