An enhanced FitzHugh–Nagumo neuron circuit, microcontroller-based hardware implementation: Light illumination and magnetic field effects on information patterns

Abstract

This contribution introduced and investigated an improved photosensitive memristive FitzHugh–Nagumo (FHN) neural circuit. The mathematical equations of the model have been derived from Kirchhoff’s electrical circuit law. Based on Helmholtz’s theorem, the Hamilton statistical function that enables us to obtain the physical energy of the neuron necessary to sustain electrical activity has been established. Therefore, the energy dependency on the light intensity and the magnetic field has been provided. Using two-parameter diagrams, bifurcation diagrams, Lyapunov exponents, and time series, the window of parameters in which regular and irregular patterns of the neuron occur have been characterized. Besides, the proposed FHN neuron model is implemented using a simple microcontroller platform and various firing patterns are acquired experimentally to validate the numerical results. Furthermore, light and field-induced pattern formation in a chain regular lattice network made of 100 photosensitive memristive neurons is being studied. Numerical experiments are carried out to identify the effects of photocurrent and magnetic flux parameters on information coding patterns translated by the membrane potential from voltage output. The spatiotemporal patterns revealed that the network presents localized wave patterns with some features of synchronization, sensitive to parameter change. This confirms that information coding and other collective behaviors of the network could be efficiently controlled by taming the intensity of light illumination and magnetic field exposure.