The coherence resonance (CR) of globally coupled Hodgkin-Huxley neurons is studied. When the neurons are set in the subthreshold regime near the firing threshold, the additive noise induces limit cycles. The coherence of the system is optimized by the noise. The coupling of the network can enhance CR in two different ways. In particular, when the coupling is strong enough, the synchronization of the system is induced and optimized by the noise. This synchronization leads to a high and wide plateau in the local CR curve. A bell-shaped curve is found for the peak height of power spectra of the spike train, being significantly different from a monotonic behavior for the single neuron. The local-noise-induced limit cycle can evolve to a refined spatiotemporal order through the dynamical optimization among the autonomous oscillation of an individual neuron, the coupling of the network, and the local noise.

In the presence/absence of external noise, dynamical behaviors of periodically forced neural systems and firing modes of interspike interval (ISI) are investigated by employing the Hindmarsh-Rose model. In the biologically relevant range of the forcing frequency, the interplay among the intrinsic oscillation, the forcing oscillation, and the noise leads to three kinds of firing modes: multi-modal firing, bi-modal firing, and intrinsic oscillation, in terms of which we can roughly classify the relevant experimental observations on the periodically forced sensory neural systems through their dynamical status. The resonant feature of subthreshold intrinsic oscillations shown in the ISI, the output signal-to-noise ratio, and the mean firing rate, appears to be an indication of stochastic resonance (SR) without external noise, or the 'intrinsic' SR. In the multi-modal firing region where SR leads to the skipping phenomenon, based on the 'intrinsic' SR, a possible explanation of a specific ISIH observed in experiments is given. Moreover, a neural system can tune itself to be chaotic to encode weak signal, rather than relying only on the external noise.