Abstract:Most external stimuli, including sound, temperature, and illumination, exhibit spatially heterogeneous, and different amplitudes of the same signal are received by neurons at different positions in the neural network. To address this issue, we constructed a grid-like neural network using memristive FitzHugh-Nagumo neurons. The neuronal responses depend on the spatially distributed stimuli, with the stimulus amplitudes being determined by the distance from the central area. Consequently, complete synchronization occurs in the network comprising periodic neurons, chaotic neurons, and their hybrid forms. Periodic patterns maintain the highest Hamilton energy whereas the lowest Hamilton energy appears in chaotic neurons. In a network consisting of chaotic neurons, the synchronization threshold is larger compared to the other types. In particular, the periodic neurons with the highest energy oscillations can regulate the low-energy chaotic neurons into periodic patterns. Similar conclusions are drawn in a chain-like network. The results advance the understanding of the synchronization mechanisms in the presence of spatial heterogeneity.

Collective Behaviors of Neural Network Regulated by the Spatially Distributed Stimuli