Non-Gaussian noise optimized spiking activity of Hodgkin–Huxley neurons on random complex networks

In this paper, we numerically study how the NGN's deviation q from Gaussian noise (q = 1) affects the spike coherence and synchronization of 60 coupled Hodgkin–Huxley (HH) neurons driven by a periodic sinusoidal stimulus on random complex networks. It is found that the effect of the deviation depends on the network randomness p (the fraction of random shortcuts): for larger p (p > 0.15), the spiking regularity keeps being improved with increasing q; while, for smaller p (p < 0.15), the spiking regularity can reach the best performance at an optimal intermediate q value, indicating the occurrence of “deviation-optimized spike coherence”. The synchronization becomes enhanced with decreasing q, and the enhancing extent for a random HH neuron network is stronger than for a regular one. These behaviors show that the spike coherence and synchronization of the present HH neurons on random networks can be more strongly enhanced by various other types of external noise than by Gaussian noise, whereby the neuron firings may behave more periodically in time and more synchronously in space. Our results provide the constructive roles of the NGN on the spiking activity of the present system of HH neuron networks.