A memristors-based FHN neuron model design

Hongmei Xu,Hemeng Gao,Jingrui Wang,Di Liu
DOI: https://doi.org/10.1109/ICETCI61221.2024.10594529
2024-05-24
Abstract:A memristor is a nonlinear electronic device that utilizes changes in electrical resistance to achieve information storage and processing. Its nonlinear properties can lead to more complex dynamic behaviors in chaotic systems and have potential in constructing artificial neural synapses. The introduction of memristor as synapses bring artificial neural networks closer to biological neural networks, and the combination of neurons and neural networks allows the system to display richer neuronal firing phenomena. Efficient computation and storage functions of neural networks can be realized, which is possible for research in the fields of computation, brain computing, and neuroscience. Therefore, in this paper, the traditional modified-Nagumo (FHN) neuron system is introduced by combining the memory resistor with neurons. Using the theoretical knowledge of chaotic circuits, the modified-memory resistor neuron model with rich dynamic properties and simple circuitry is proposed.
Engineering,Computer Science
What problem does this paper attempt to address?