Detection of Underwater Weak Signals Based on Stochastic Resonance Multi-Measures Fusion

摘要

Stochastic resonance (SR) has been proved to be an effective approach for weak signal detection, which may probably to be used for a variety of oceanic fields such as underwater acoustic concealed communication, passive sonar detection, seismic wave detection, and etc. This paper presents a new multi-measurements fusion index, named entropy kurtosis variation product, of Stochastic Resonance for weak signal detection. This new measurement index is built which to combine two novel indexes of the approximate entropy and kurtosis variation, and a weak signal detection strategy with the residual signal that reconstructed by remove the peak spectrum spike in frequency domain, which is anticipate to be used for unknown weak signal detection (unknown prior information). Simulation comparations and an tank experiment have verified the proposed method can be convenient and effective for detecting underwater weak signals.