Self-tuning active sonar signal processing for a towed hydrophone array during a turn

This paper addresses the problem of hydrophone array shape estimation of a towed array system especially during maneuvers of the tow ship. Uncertainty in the position and velocity of each sensor of the array degrades the performance of a beamformer and, hence, leads to a loss of performance for the entire data processing chain, including target tracking and (if applicable) data fusion. A self-tuning approach is developed which uses information generated by automated target tracking during the period prior to the tow ship maneuver. This information is then used to perform a sequential tracking of the antenna system during the maneuver. The antenna tracking includes the estimation of the sensor positions which is implemented as a Gaussian sum filter. This approach allows to keep the handling of multi-target situations also during the turn. First results of validating the new approach in two simulated scenarios are presented, which indicate that the approach is capable of maintaining good tracking performance during the periods of tow ship maneuvers, whilst assuming only rough knowledge about the antenna in the beamforming leads to the expected performance loss. Tracking performance is expressed in terms of probability of detection and mean squared error of the target positions. For validation purposes, the sensitivity analysis of the new approach contains Monte Carlo (MC) selections of parameters describing both the tow ship maneuver and the multi-target setup.