Stochastic resonance based on correlation coefficient in parallel array of threshold devices

摘要

The phenomenon of stochastic resonance (SR) based on the correlation coefficient in a parallel array of threshold devices is discussed. For four representative noises: the Gaussian noise, the uniform noise, the Laplace noise and the Cauchy noise, when the signal is subthreshold, noise can improve the correlation coefficient and SR exists. The efficacy of SR can be significantly enhanced and the maximum of the correlation coefficient can dramatically approach to one as the number of the threshold devices in the parallel array increases. Two theorems are presented to prove that SR has some robustness to noises in the parallel array. These results further extend the applicability of SR in signal processing.