Implementation and Comparison of Multiple Source Location Algorithms: Results of a Simulation.
Abstract
This report compares the resolution and the detection threshold of various multiple source location algorithms. Such algorithms are digital signal processing techniques that use signals received by an array of sensors and provide estimates of the bearings of the signal sources. The report covers only linear arrays consisting of omni-directional acoustic sensors. The signals are assumed to be narrowband. The methods under consideration are: (1) Classical method; (2) Adaptive method; (3) Maximum entropy method (approximate); and (4) Eigenvector method. The first three methods involve estimating the signal power incident on the array from various bearings. The second and third method involve computation of the maximum-likelihood and maximum-entropy estimates of the power respectively. The eigenvector method is based on certain properties of the eigenvectors of the correlation matrix of an array illuminated by a field of discrete sources. It represents the acoustic field as a finite sum of plane waves. The algorithms have been implemented by computer programs written in FORTRAN and have been tested with simulated data on the VAX-11/780 system at the Naval Surface Weapons Center, White Oak, Maryland. The specific details of the computer programs and the results of the simulation are described in Chapters 3 and 4 respectively. Chapter 4 compares the resolution, detection threshold, and computational effort of the four methods.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1983
- Accession Number
- ADA146241
Entities
People
- L. T. Barkakati
- R. S. Hebbert
Organizations
- Naval Ordnance Laboratory