Detection of Weak Audio Signal in Ocean Based on Stochastic Resonance

Abstract

There are plenty of audio information in ocean, whose audio feature often brings disturbance on the detection of ocean object, and some new methods are requried to realize the object classification on audio frequency. Stochastic resonance (SR) can realize the detection of weak signal, however the most researches concentrate on single weak periodic signal, and less research on aperiodic audio signal. Starting from the bistable SR system, this paper adopts the combina- tion method of theory analysis and numerical simulation. and conducts a theory analysis on the detection of weak aperiod- ic audio signal under strong noise background. By using time scale transformation method, a complete detection system for weak digital signal is established, and it is applied in underwater audio signal. The analysis and simulation results pre- sent that, the audio signal is successfully recovered in the condition that the signal-to-noise ratio (SNR) is -13dB. The re- search results possess a vast engineering significance on the further development and application of SR in the detection of weak audio signal.