Active sonar target tracking for anti-submarine warfare applications

Tracking potential submarine targets using active sonar is a challenging task, due to the high level of clutter that results in a large number of false alarms. This can potentially overwhelm or critically limit the capabilities of an active sonar tracker.The focus of this paper is to investigate the performance of conventional tracking algorithms for both a single target and multiple targets in cluttered environments. The tracking algorithms considered in this study are the Extended Kalman Filter (EKF) and the Converted Measurements Kalman Filter (CMKF). For single target in clutter the Probabilistic Data Association (PDA) is used, while for multiple targets in clutter, the Global Nearest Neighbour (GNN) data association algorithm is implemented. A new previously unpublished version of the PDA is developed for use with the CMKF tracker to take into account that the measurement covariance matrix, innovation covariance matrix and filter gain are all dependant on the individual measurements in the gate at a given time, making the updated state estimate more complicated than that used for the EKF. The performance of the tracking and data association algorithms is tested using Monte Carlo simulations.