Comparative analysis of EKF and Particle Filter performance for an acoustic tracking system for AUVs exploiting bearing-only measurements

This paper aims to compare different navigation state estimation approaches for the tracking of an Autonomous Underwater Vehicle (AUV) based on Direction of Arrival (DoA) measurements. The use of passive acoustic devices to compute the DoA information allows for a high rate (up to 1Hz in the proposed tests) of the measurements while the AUV moves underwater. The informative content of these data integrated with the synchronised data of the on-board sensors allows for a fine tracking of the vehicle under test for navigation performance evaluation. The paper addresses the issue of the selection of the proper data filtering strategy for the estimation of the navigation state of the AUV. In particular, it investigates whether the adoption of a Particle Filter (PF) has significant advantages with respect to a classical approach based on an Extended Kalman Filter (EKF). The two approaches exploit the same models and parameters to limit the comparison to the filtering strategy. The ground truth is the GPS signal provided by the receiver integrated within the antenna on the AUV under test. The comparative analysis considers both the estimation performance and the computation load. Results are extensively provided for a subset of all the processed dataset, all the remaining part of data confirm the reported analysis.