Blind Adaptive Stochastic Resonance in Base Band Binary Signal Processing

Stochastic resonance can increase signal to noise ratio greatly which remains a hot subject in signal processing over last two decades. There are two ways for nonlinear system to achieve stochastic resonance: tuning the system parameter and tuning the intensity of noise. The method of tuning system parameter is much flexible than the other. However, the best value of system parameter is difficult to choose, which confines its application in practice. Stochastic resonance is used as a nonlinear receiver of the base band binary digital signal processing system in this paper. The relationship between the system parameter and the kurtosis of the output signal from stochastic resonance is found out, and a new method of blind adaptive stochastic resonance based on the criteria of kurtosis is proposed in this paper. The system parameter of stochastic resonance is tuned automatically in the gradient direction of the kurtosis. The method is much robust and flexible than traditional method of tuning noise. The validity of the proposed method is proved by the simulation results at last.