Localization of a target in clutter by multistatic active sonar

PDF

Authors
  1. Peters, D.J.
Corporate Authors
Defence Research and Development Canada, Atlantic Research Centre, Halifax NS (CAN)
Abstract
A method is presented for localizing a single target with multistatic active sonar, from a single ping, by means of a numerical search for the highest peak in a probability distribution function. That function is constructed using Bayesian principles. The method takes into account errors in the locations of the transmitter and the receivers, errors in the estimate of the local speed of sound, errors in time and bearing measurements, an imperfect probability of detection, and the appearance of false contacts due to noise. The efficacy of the method is demonstrated by means of simulated multistatic data. Because of the false contacts and the imperfect probability of detection, it is inevitable that the resulting localization will sometimes be very inaccurate. But there is a tendency for multiple receivers to corroborate each other, leading to frequent successful localization in spite of these difficulties. The distribution of the error in localization over many runs is similar in form to the corresponding distribution in the case of perfect detection and no false contacts, except for the appearance of a long tail in the distribution. However, the computational load is very heavy. A streamlined version of the probability distribution function, in which some of the uncertainties are neglected, speeds up the computation by several orders of magnitude with the localization performance being degraded only slightly. The single-ping localization method presented here appears to be a p

Il y a un résumé en français ici.

Keywords
localization;multistatic active sonar;sensor fusion
Report Number
DRDC-RDDC-2017-R054 — Scientific Report
Date of publication
01 May 2017
Number of Pages
38
DSTKIM No
CA045037
CANDIS No
805417
Format(s):
Electronic Document(PDF)

Permanent link

Document 1 of 1

Date modified: