Parameter-induced Stochastic Resonance And Binary Baseband Digital Signal Transmission

Stochastic resonance (SR) was a synergistic phenomenon happened among nonlinear dynamic system, stochastic force (noise) and input signals. SR had generated a counter-conventional image upon the conception that noise played inherently destructively in signal processing. Studies showed noise was a constructive but elementary part in SR phenomenon. However, as a ubiquitous nonlinear phenomenon, SR laid in almost every subject of natural science. Thus, SR had attracted broad attention in the past 20 years. The main subject of this thesis was Theory of Parameter-induced Stochastic Resonance (PSR) in A Class of Bistable System with Tunable Exponent and its application in binary baseband digital signal transmission.Normally, the classic researches about SR were concentrated on investigations of the insufficiency of noise density by which SR was going to take place. However, it had been proved that SR via tuning noise could be induced equivalent to tune system parameters. Additionally, PSR proved to be a more powerful methodology than SR via adding noise. PSR had lower limitation and higher applicability, and it more practically supposed that the noise existed in system kept unchanged while processing. In the case of white noise, we made detailed research on the output characters of a class of bistable system with tunable exponent on the basis of the system stochastic differential equation and Fokker-Planck (FPK) equation. From the perspective of parameter tuning, the mechanism of SR was explained by the conception of non-stationary probability distribution and system response speed.The application of PSR of a class of bistable system with tunable exponent in binary baseband digital signal, mainly the pulse amplitude modulated (PAM) signal, was explored in this article. The previous researches in SR were mainly focused on quartic bistable system and used its stationary solution as output probability distribution. In the case of equivalent amplitude between serial input signals, the stationary solution was appropriate for output evaluation. But, combining a newly introduced tunable exponent , we applied an approximation but with higher accuracy expression as output probability distribution on the consideration of possible amplitude alteration in binary input signals. Over an additive white Gaussian noise (AWGN) channel, the methods to calculate communication system's performance indexes, Probability of Error and Channel Capacity, and how to tune system parameters were put forward via theoretical analysis and simulation experiments. Finally, the output result demonstrated that the system performance was enhancedwith an appropriately selected exponent But, it should be noticed that the introduced approximation expression for output probability distribution in this paper was not a precise formula. Especially when exponent was set in a large value, because big produced strong nonlinearity to system, it caused the low accuracy of theoretical calculation if parameter had comparative greater value.And, only a short range of theoretical analysis and simulation experiments about applications of PSR in binary baseband digital signal transmission had been done in this thesis. In order to consummate this subject, further work needs to be done in the fields of adaptive PSR, SR under colored noise and more accurate methodology for the evaluation of communication system's performance.