MS Plan B Defense: Direction of arrival estimation of acoustic signals from a 2-element hydrophone array

Jacob Foster
Master’s Student
Department of Ocean and Resources Engineering
University of Hawai’i at Mānoa

Zoom Meeting Link

Meeting ID: 838 9499 8063
passcode: ORE
https://hawaii.zoom.us/j/83894998063

This project explores the performance of the Maximum Eigengap Estimator in the omnidirectional hydrophone setting. The estimator is an unsupervised signal subspace steering vector estimator that we use to conduct direction-of-arrival estimation of ocean acoustic signals. It uses narrow-band aggregation of a broadband signal with an optimal weighting to maximize the signal-to-noise ration and identify the eigenvector that is closely oriented to the source signal that is computed from the sensor array output. The algorithm is designed to estimate a single broadband signal direction-of-arrival from two-channels of acoustic data. The maximum eigengap estimator has only been used to conduct direction of arrival estimation from vector sensor data, and in this project we extend its application to an omni-directional hydrophone array. We explore the estimator’s performance through simulation and experimental results on NOAA Drifting Acoustic Spar Buoy (DASBR) data. We show the estimator performs well in simulation, and we see reasonable results in the DASBR estimates. The estimator performs exceptionally strong well when we introduce an interfering non-broadband signal in the acoustic record.