Effects of multiscale noise tuning on stochastic resonance for weak signal detection

Noise enhanced signal detection via stochastic resonance (SR) is generally realized by white noise tuning with an optimal noise intensity. This paper explores a new mechanism of SR that is induced by the noise at multiple scales for enhanced detection of weak signals under heavy background noise. A strategy is proposed to realize the SR via multiscale noise tuning according to the property of 1/f noise. The presented new method combines the benefits of colored noise and parameter tuning to the SR phenomenon. Under the strategy, effects of noise intensity, analysis scale, and driving frequency on the SR are analyzed through numerical simulations. Three merits are displayed for the proposed multiscale noise-induced SR model: insensitivity to noise intensity, activity of multiple scale noise, and capability of detecting high frequency. A practical application to structural defect identification has confirmed the effectiveness of the proposed method in comparison with traditional methods.