| In the view of stochastic resonance (SR) and signal processing, parameter-turning SR (PSR) theory and method is proposed and the characteristics of bistable systems are studied. The applications of PSR in multi-frequency analog signal processing, digital baseband pulse-amplitude modulation signal transmission and digital carrier modulation signal transmission are investigated. This thesis has also introduced the concepts of parameter-adaptive SR and array-induced SR. It is importance for the development of SR theory and its application in nonlinear signal processing.In the case of white noise, we research the system stochastic differential equation and Fokker-Planck (FPK) equation detailedly. The mechanism of SR is re-explained by system response speed and a new method is introduced, i.e., PSR theory and method. It is shown that the output signal-to-noise ratio (SNR) obtained by adjusting systems parameters can exceed that by turning noise intensity, especially when the input noise intensity has already been beyond the resonance region or point. It is demonstrated that the theory and the method presented here can markedly improve the output SNR, and minimize phase lag as well as the distortion of the system output signal with multi-frequency. A whole framework for multi-frequency signal processing is formed, including the methods of selecting system parameters and signal recovery.In the point of-information theory;-the- baseband binary pulse amplitude modulated (PAM) signal transmission, via turning the nonlinear receiver's parameters, is studied over an additive white Gaussian noise (AWGN) channel. It is demonstrated that the channel capacity of binary communication systems, for a given signal added noise, can be maximized by optimal designed receivers. This new form of SR is referred to as PSR in a broad sense. The communication system's performance is briefly analyzed by channel capacity, which is a function of bit error rate (BER). The decision rule is based on the most likelihood method. For reproducing the original binary signals more correctly, time scale transformation method and the approach of ensemble average probability of error bits are introduced. We observe a marked enhancement of the channel capacity for binary PAM signals transmission by PSR in numerical simulations. With this theory and method, the practical applications are considered.We study the effect of colored noise upon the information transmission via PSR over an additive noise channel. This work extends the results of previous studies on binary information transmission. The one-dimension effective FPK equation of colored noise serves as our starting point to evaluate the effect of noise color on communication system design. The theory of BER versus input SNR per bit is developed and a new procedure to determine the system parameters that influence the minimum BER is proposed. It is shown that the communication system's performance is related to the turnable system parameters as well as the noise correlation time.It is demonstrate that SR realized by turning system parameters can be explained as SR via adding noise after a transformation. It is found that SR via turning noise could be made equivalent to turning system parameters and adjusting signal amplitude accordingly. But the PSR method can appear under the more worse environment than SR via adding noise does.With the development of SR, there are two active investigation directions, i.e., adaptive SR and array-induced SR. The adaptive SR, studied in this thesis, means that the system automatically selecting parameter with a recursive algorithm, not adaptively adding noise. A parallel connection SR system is introduced and its application in digital frequency shift modulation signal transmission is discussed.
|
PDF Full Text Request