SOURCE ESTIMATION ALGORITHM IN NON COOPERATIVE BI-STATIC SONAR SYSTEM

Research and surveillance at sea may benefit from the covertness of some of the involved platforms. Within this context, the paper focuses on a non-cooperative Bi-Static sonar configuration: a specific active sonar system in which the surface transmitter and the underwater receiver do not communicate with each other and the covert receiver has no information about transmitter source depth, course and speed, waveform and frequency spectrum, etc. Environmental conditions (bathymetry, sound speed) in the whole surveillance area are assumed known. Within this framework, several works have been proposed and this paper delves into the preliminary phase of one of these: a target detection algorithm based on the joint source-target localization at the receiver through matching the multipath arrivals with model based predictions. Such algorithm requires the application of an unknown source deconvolution processing to extract the multipath arrivals, plus the inversion of the arrivals times. Unknown source deconvolution is performed through a novel algorithm in the time-frequency domain, which aims to estimate the characteristics of the unknown transmitted signal. The algorithm is described and finally validated with synthetic acoustic signals and applied to real data. Tests on real data, though at a preliminary stage, confirm the theoretical and simulated analysis.