Coherence resonance modulated by hybrid synapses and time delay in modular small-world neuronal networks with E–I balanced state.

Neurons communicate primarily through synapses. A neuron is usually affected by multiple synapses, which could be chemical/electrical or excitatory/inhibitory ones at the same time. Here, we make the realistic assumption that a excitatory and inhibitory balanced modular small-world network is established and focuses on the effects of hybrid chemical and electrical synapses, noise and time delay on coherence resonance of the constructed network. It is found that when the ratio f of chemical synapses to electrical synapses approaches odd ratios, coherence resonance is better than those f close to even ratios for appropriate noise intensities. Furthermore, with f increasing, it is observed that effects of chemical and electrical synapses on coherence resonance are nearly opposite. It indicates that electrical synapses are more efficient than chemical ones. Meanwhile, multiple coherence resonances are observed when time delay is introduced into the network, and it is independent of f. Finally, we demonstrate that coherence resonance decreases as the number of subnetworks increases, and when the number of subnetworks is larger, the resonance behaviour weakens or vanishes with increasing f. [ABSTRACT FROM AUTHOR]