Bayesian passive acoustic tracking of a cooperative moving source in shallow water

This study addresses the problem of tracking acoustically a moving source in shallow water, which is challenged by multi-path sound propagation. A specific source of interest is an autonomous underwater vehicle with periodic sound transmissions. Given a simple horizontal linear receiving array, the authors develop a Bayesian tracking method based on measurements of signal arriving angle and difference of respective sonar transmission and receiving time intervals. To counter the multi-path effect and thus improve estimation accuracy of time differences, passive time reversal processing is exploited. Further, the unscented Kalman filter is adopted in tracking to handle nonlinearity in measurements. The proposed method is tested in simulations as well as an at-sea experiment, and the results validate the effectiveness of the developed approach.