Study On Non-Dynamic Stochastic Resonance Model For Optimum Signal Detection

Noise in communication system is usually considered a nuisance. However, in certain nonlinear systems, including electronic circuits and biological sensory systems, the presence of noise can enhance the diction of weak signal. This implies that optimal system noise level need not be zero and such nonlinear phenomenon is termed Stochastic Resonance. The fact may seem at odds with almost a century of effort in signal processing to filter noise or to mask or cancel it. For this reason such paradoxical mechanisms are appealing candidate for use in a variety of signal processing context.In this thesis, Poisson regime is introduced in order to characterize Level-Crossing Detector (LCD) Stochastic Resonators, and an optimum signal detection model for Non-Dynamic Stochastic Resonance is put forward. The main works are included as follows:1 The output signal of LCD stochastic resonator is modeled as stochastic Poisson process. The characteristic of this point process is analyzed and the time-varying expectation of output Poisson process is discussed.2 The flow of information in such system is analyzed by using statistical method. Then an optimum signal detection algorithm based on LCD stochastic resonator is obtained.3 The performance of LCD-based maximum-likelihood receiver is explored and Stochastic Resonance characteristic of detection is demonstrated.