A Bayesian Method for Localization by Multistatic Active Sonar


  1. Peters, D.J.
Corporate Authors
Defence Research and Development Canada, Atlantic Research Centre, Halifax NS (CAN)
The question of localizing a target with multistatic active sonar is re-examined from the perspective of finding a peak in a probability distribution function. The probability distribution function is constructed using straightforward Bayesian principles. Both a position estimate and a covariance matrix can be found, provided that an implementation of a numerical algorithm for finding a local maximum is available. The localization method developed herein can account for transmitter and receiver location errors, sound speed errors, time errors, and bearing errors. A Monte Carlo test is conducted in order to compare the accuracy of the proposed method to that of a more conventional method used as a baseline. In each iteration, a transmitter, several receivers, and a target are positioned randomly within a square region, and the target is localized by both methods. The proposed method is generally more accurate than the baseline method, within the range of parameters considered here. The degree of improvement over the baseline is greater with a larger region area, with a larger bearing measurement error, and with a smaller time-of-arrival measurement error, and slightly greater with a larger number of receivers.
localization;multistatic active sonar;sensor fusion
Report Number
DRDC-RDDC-2016-P030 — External Literature
Date of publication
12 Apr 2016
Number of Pages
Electronic Document(PDF)

Permanent link

Document 1 of 1

Date modified: