Target detection in shallow-water reverberation based on parameter-induced stochastic resonance

A method based on parameter-induced stochastic resonance (PSR) is proposed for the target echo detection in the presence of shallow-water reverberation. The signals received by the horizontal sensor array are first preprocessed to form the input to a bistable stochastic resonance (SR) system. The detection decision is then made by a linear correlation detector based on the output of this SR system. The performance of our proposed detection system, in terms of deflections, is derived based on the solutions to the corresponding Fokker-Planck equations (FPE). Using PSR-based techniques, the detection performance can be optimized by tuning the system parameters properly. In addition, numerical simulations are carried out to demonstrate the efficiency and detection performance of our proposed detection system with receiver operating characteristic (ROC) curves, and the results show that it has good performance under the condition of the weak signal-to-reverberation-noise ratio (SRR).