Physical Acoustics Research Institute, Princeton, NJ
(Received 22 January 1996; accepted 15 October 1999)
Stochastic resonance allows below-threshold signals to be detected because of the presence of favorable background noise. In this paper the author describes stochastic resonance and anti-resonance and presents some examples to illustrate how these affect the detection rates of acoustic emission signals or ultrasonic pulses in low signal-to-noise environments. The author introduces the signal fraction, a measure for quantifying the fraction of detected events that contain a signal that appears to have advantages over the signal-to-noise ratio for certain kinds of data, and a formalism for calculating the probability of signal detection when the noise is a time-dependent sinusoid.
Paper ID: JTE12074J