A NEW METHOD TO RECOVER THE SIGNALS OBTAINED BY STOCHASTIC RESONANCE
随机共振系统输出的一种新的反演方法

Stochastic resonance (SR) is a counterintuitive phenomenon wherein the noise plays a constructive role in nonlinear dynamic systems, and it becomes a promising way of signal processing recently. Since most of the previous works concentrated on distorted simple-harmonic input and digital signals, we are tending to observe more signal processing features of SR by use of multi-frequency analog signals in this paper. It is difficult to recover the information in the output obtained by a SR system. But for simple input, i.e. the simple-harmonic analog signal and digital signal, there are some typical methods. When treating a multi-frequency analog signal, the recovery procedure becomes more complex. In this paper, we firstly discuss the reason why the output of a nonlinear system is distorted in detail. Based on this work, a new signal recovery method is suggested. The method includes a simple formula to restore the signal data, a linear interpolation in transient timeslice and a least square polynomial fitting to deal with the data obtained above. According to the simulation, this method gives good results when the system parameters are well selected.