A Bayesian Approach to Passive Sonar Detection and Tracking in the Presence of Interferers

In this paper, a Bayesian approach to tracking a single target of interest (TOI) using passive sonar is presented. The TOI is assumed to be in the presence of other loud interfering targets, or interferers. To account for the interferers, a single-signal likelihood function (SSLF) is proposed which uses maximum-likelihood estimates (MLEs) in place of nuisance parameters. Since there is uncertainty in signal origin, we propose a computationally efficient method for computing association probabilities. The final proposed SSLF accounts for sidelobe interference from other signals, reflects the uncertainty caused by the array beampattern, is signal-to-noise ratio (SNR) dependent, and reflects uncertainty caused by unknown signal origin. Various examples are considered, which include moving and stationary targets. For the examples, the sensors are assumed to be uniformly spaced linear arrays. The arrays may be stationary or moving and there may be one or more present.