Signal-to-noise ratio in stochastic resonance

We solve the Fokker-Planck equation for an overdamped Brownian particle in a periodically forced bistable potential by means of a path integral method, obtaining the propagators in the steepest-descent (small-noise) approximation. We compute the long-times asymptotic probability distribution, the asymptotic correlation functions, and the time-averaged spectral density, which allows us the immediate calculation of the signal to noise ratio, a directly measurable quantity useful to characterize the phenomenon of stochastic resonance. Our numerical algorithm is fast and runs on a desktop computer, and the results agree with experiments and with former theoretical calculations of the amplification factor, in addition it allows us to calculate the experimentally more accessible signal to noise ratio.