Stochastic resonance in threshold systems
Abstract
We consider signal processing in simple threshold systems with nonstationary additive and/or multiplicative noise. A discrete-time process with a small periodical signal masked by noise represents an input. The systems convert sampled input data to a nonstationary random point event flow carrying some information on an input signal. As it is shown in our previous study [M.M. Alibegov, Astron. Lett. 22, 564 (1996)], the Rayleigh spectral function of a random point event train estimates a signal-to-noise ratio (SNR) at selected frequencies. Based on these results, we compute a system response at signal frequencies as a function of threshold and input noise intensity. The threshold systems are shown to reveal stochastic resonance (SR), i.e., output SNR exhibits a maximum at resonant noise intensity (intensities) and threshold(s) at rather common conditions. We show that SR and Rayleigh spectral technique allow us to carry out numerical signal detection in data sets with noise.
- Publication:
-
Physical Review E
- Pub Date:
- May 1999
- DOI:
- 10.1103/PhysRevE.59.4841
- Bibcode:
- 1999PhRvE..59.4841A
- Keywords:
-
- 05.40.-a;
- 02.50.-r;
- 07.05.Kf;
- Fluctuation phenomena random processes noise and Brownian motion;
- Probability theory stochastic processes and statistics;
- Data analysis: algorithms and implementation;
- data management