The Research Of Stochastic Resonance In Threshold Network Based On Mutual Information

In signal processing and other fields, the traditional wisdom is that the noise are the main causes of deterioration of performance, so the suppression of noise has been the traditional signal processing method to improve the quality of communication. With the development of Nonlinear Science, it was found that Stochastic Resonance(SR) makes the output signal of the system significantly increased and the transmission performance improved by making the signal and the noise achieve certain synergy in nonlinear systems. The noise are considered as favorable factors based on stochastic resonance, which makes groundbreaking significance of the signal transmission research in nonlinear system.This paper studies the realization of the stochastic resonance in signal transmission, the main work is as follows:This paper firstly introduces additive noise and multiplicative noise to study the supra-threshold stochastic resonance phenomenon in threshold network in the effect of Gauss noise and the noise improving the information transmission based on mutual information. Especially, we study the effect of the threshold and the noise intensity on the supra-threshold stochastic resonance, and find that the additive noise are more likely to induce the stochastic resonance.Then, this paper studies the supra-threshold stochastic resonance phenomenon in threshold network in the effect of three typical noises and studies the noise improving the information transmission based on different application environments. By comparing the experimental results of different noises, it is concluded that the effect of noise on stochastic resonance is different. It is found that different noise distributions can induce the supra-threshold stochastic resonance, which confirms the robustness of the stochastic resonance to the noise.Lastly, the paper builds a text encoding, transmitting and decoding of simulation experiment, studies the effects of different noises on the ratio of error characters which may arise in the process of symbols transmission, and demonstrates the realization of the stochastic resonance in the specific application. For different noise distributions, there are stochastic resonance, which proves the robustness of the stochastic resonance to the noise once more.