Range, velocity and immersion estimation of a moving target in a water-filled tank with an active sonar system

The knowledge of the impulse response of the propagation channel is required for many underwater applications as communication, sonar detection and localization, marine mammals monitoring, etc. In the context of target classification, this impulse response informs about the relative motion between the source/receiver and the target through the Doppler effect. Knowing the emitted signal, the Doppler compression coefficient of each acoustic path can be estimated with the wideband ambiguity plan [1]-[3]. Warping operator based-filter and wideband ambiguity plan can be used to separate acoustic paths in order to remove interferences and estimate properly the target distance and speed [4]. This paper presents a complementary approach to jointly estimate the distance and speed of the target at a small speed with reasonable resolutions. The investigated sources are Binary Phase Shift Keying (BPSK), 22-Welch-Costas, and Pulse Train Frequency Modulation (PTFM) signal. Sources have a large Time-Bandwidth product (high TB) and provide high resolutions. For this reason, an echo model that takes into account a signal temporal compression (Doppler effect) can be used. A reduced-scale laboratory experiment was conducted to estimate the speed vector and depth of a moving target. Results for speed vectors are compared for the three different sources. A ray back propagation algorithm was used and results show correct estimation of the target depth.