Optimal structure of the passive sonar for detecting very weak broad-band signals

摘要

OPTIMAL STRUCTURE OF THE PASSIVE SONAR FOR DETECTING VERY WEAK BROAD-BAND SIGNALS - The article deals with the application of the optimal statistical detection for the passive sonar. The Neyman-Pearson statistical criterium of Bayes family probability tests is considered. It is supposed that the underwater vessel noise on the receiving location (useful signal) is a very weak broadband, stationary and ergodic colored stochastic process with the Gaussian distribution. The deep-sea ambient noise (interference) is supposed to be a broadband, stationary and ergodic colored Gaussian stochastic process. Its power spectral density is a known and deterministic function versus frequency. The article covers the frequency band from 500 Hz to 10 kHz.