HOW FAST CAN A NEURON TRANSFER INFORMATION: BANDWIDTH IS THE REAL ISSUE

The information channel capacity of neurons is calculated in the stochastic resonance region by using the Shannon formula. This quantity is an effective measure of the quality of signal transfer, unlike signal-to-noise ratio or information entropy measures, which characterize only the resolution of the output and not the rate of information transfer. This fact has also been realized by some authors,. however, until now, no published results exist for real stochastic resonators or neuron models. The most probable reason of missing results is the problem with defining the maximal bandwidth of the system. The trick we use is to applying Wiener's sampling theorem to define the maximal bandwidth. The channel capacity exhibits a well expressed maximum versus the input noise intensity, and the location of the maximum is at a higher input noise level than it was observed for classical measures, such as signal to noise ratio or entropy. In conclusion, more noise is needed for the optimal transfer than it has earlier been assumed.