Comparing passive localization methods for ocean vehicles from their underwater sound received on a coherent hydrophone array

Abstract

Simultaneous localization and tracking of multiple ocean vehicles over instantaneous continental-shelf scale regions using the passive ocean acoustic waveguide remote sensing (POAWRS) technique employs a large-aperture densely-sampled coherent hydrophone array system to monitor the underwater sounds radiated by ocean vehicles. Here, particle filtering is implemented for bearings-only localization and tracking of surface ships where the estimated bearing-time trajectories of ship-radiated underwater sound detections are used as inputs to provide two-dimensional horizontal position estimates. Coherent beamforming of the acoustic data received on the hydrophone array not only provides estimates of ship bearing, but also significantly enhances the signal-to-noise ratio. Results of passive acoustic surface ship localization and tracking from recordings in the Gulf of Maine and the Norwegian Sea are presented. The particle filtering approach for passive source localization are compared with three other approaches: moving array triangulation (MAT), array invariant (AI), and modified-polar-coordinates extended Kalman filter (MPC-EKF). The passive source localization accuracies, determined by comparison with the GPS-measured position of the surface ships, are dependent on numerous factors such as source-receiver geometry, range, and relative speed. By combining several of these approaches, which have respective pros and cons under different circumstances, the surface ships can be localized with improved accuracy.