Abstracts Submitted: 510
Number of Users: 658
View Abstracts Submitted
Back to home Page
Acoustic Source Localization in underwater is the process of detecting and determining its bearing and range using single or an array of hydrophones. The Vector Sensor which measures both pressure and particle velocity has many advantages over Omnidirectional Hydrophones and is very useful for Source Localization. The objective of this work is to develop an algorithm for the estimation of direction of arrival (bearing) of the source using the measurements of vector sensor array. A Vector Sensor Array consisting of three elements is used for this study. Each element has a hydrophone and a tri axial accelerometer to measure the particle velocity along with acoustic pressure at any point in three orthogonal directions. The Bearing estimation in Source Localization includes finding the Azimuth (horizontal) and Elevation (vertical) of the source. In order to estimate these parameters, two techniques namely conventional beam forming technique and MUSIC are used. It is observed that the MUltiple SIgnal Classification (MUSIC) method gives good results. It is based on the Eigen structure of the input covariance matrix that decomposes into signal and noise subspaces. The bearing is calculated from the MUSIC spectrum which peaks for steering vectors that are orthogonal to the noise subspace. Compared with the other bearing estimation technique, the MUSIC method outperforms with high spatial resolution, i.e., few numbers of sensors result in high accuracy. The algorithm was validated using the data obtained from experiment data conducted at Acoustic Test Facility (ATF), NIOT, Chennai.
© Copyright 2017 All Rights Reserved