Conference Paper

Enhancement of stochastic resonance by tuning system parameters and adding noise simultaneously

Dept. of Electr. and Comput. Eng., Polytech. Univ. of Brooklyn, NY
DOI: 10.1109/ACC.2006.1657196 Conference: American Control Conference, 2006
Source: IEEE Xplore

ABSTRACT The stochastic resonance effect can be realized by tuning system parameters or by adding noise. This paper investigates the possibility to enhance the stochastic resonance effect by tuning system parameters and adding noise simultaneously. First, we use examples to demonstrate the situation where only the system parameters or noise can be adjusted to maximize the stochastic resonance effect. Then, it is shown, using standard optimization theory, that the normalized power norm of the bistable double-well system with aperiodic input signal can reach a larger maximal value by tuning the system parameter and adding noise simultaneously. Finally, for the purpose of practical implementation, searching for the optimal system parameter and noise intensity is realized by an on-line fast-converging optimization algorithm

1 Bookmark
 · 
76 Views
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Nonlinear bistable double-well stochastic resonance systems have been successfully used for one-dimensional signal processing, based on the concept of parameter-tuning stochastic resonance. This paper investigates the applications of parameter-tuning stochastic resonance in image processing. First, a two-dimensional stochastic resonance system is introduced as a nonlinear filter for image processing. The equation satisfied by the dynamic probability density function of the images processed by this stochastic resonance filter and its solutions are then discussed. Finally, this nonlinear filter is used to process a black-white image corrupted by additive white Gaussian noise to reveal the possibility to extend the concept of parameter-tuning stochastic resonance to two-dimensional cases. This provides an innovative approach for image processing
    Acoustics, Speech and Signal Processing, 2007. ICASSP 2007. IEEE International Conference on; 05/2007
  • [Show abstract] [Hide abstract]
    ABSTRACT: Parameter-tuning stochastic resonance has been successfully applied to one-dimensional signal processing. This paper explores the feasibility to extend this technique for image processing. Based on the two-dimensional nonlinear bistable dynamic system, the equation satisfied by the system output probability density function is derived for the first time. The corresponding equation for the one-dimensional system is the famous Fokker-Planck-Kolmogorov (FPK) equation. The stationary solution, eigenvalues and eigenfunctions of this equation are then investigated. The upper bound of the system response speed and the related calculation algorithm which are necessary for the applications of this technique to image processing are also proposed in this paper. Finally, the potential applications of this approach in image processing and some future research are suggested.
    American Control Conference, 2007. ACC '07; 08/2007
  • [Show abstract] [Hide abstract]
    ABSTRACT: Stochastic resonance has been increasingly used for signal estimation, signal transmission, signal detection and image processing. The stochastic resonance effect can be realized by tuning system parameters or by adding noise. In our recent paper, we have investigated the possibility to enhance the aperiodic stochastic resonance (ASR) effect by tuning system parameters and adding noise simultaneously for the Gaussian-distribution weak input signal. This paper extends the result to a more general case using standard optimization theory. It is shown that the normalized power norm of the bistable double-well system with a small input signal can reach a larger maximal value by this scheme. An on-line fast-converging optimization algorithm is also proposed for searching the optimal values of system parameters and noise intensity

Full-text

Download
3 Downloads
Available from