Coherence resonance induced by non-Gaussian noise in a deterministic Hodgkin-Huxley neuron

In this paper, we report a novel effect of a particular kind of non-Gaussian noise (NGN), especially its deviation from Gaussian noise, on the spiking activity in a deterministic Hodgkin-Huxley (HH) neuron. It is found via numerical simulations that the coefficient of variation (CV), characterizing the spiking regularity, nonlinearly changes with varying deviation and passes through a minimum at an optimal deviation value, representing the presence of the coherence resonance (CR). This phenomenon shows that, when the NGN is optimal, the neuron can exhibit the most regular firing behavior. This result provides a constructive role of the NGN in the timing precision of information processing and signal transduction in neural systems. (C) 2009 Elsevier B.V. All rights reserved.