Improvement of Sensitivity of Signal Detection in Stochastic Resonance Using Oversampling

Abstract

Stochastic resonance have been known as a phenomenon which can enhance weak input signals added nonlinear systems with the presence of noise. In this paper, we considered the summing network as a multi-input system configuration, which occur the optimal stochastic resonance without tuning of system parameters. This configuration can be expected to improve the detection sensitivity of the input signals measured simultaneously by integrating multiple sensors, but the number of sensors cannot be increased indefinitely depending on the problem. Then we propose a method to reduce the number of elements required for the configuration by using oversampling. It was shown that the proposed configuration could improve the input-output correlation while reducing the number of nonlinear elements.