Passive localization using propagation models

Abstract

Passive ranging using propagation models can be effected in two different ways: forward modeling

and direct inversion. The forward-modeling method performs an exhaustive recomputation of the field

over a set of assumed source locations and seeks the best match to the measured data. The direct

inversion method inverts a set of linear equations constituted by the normal mode model. An overview

of these two approaches is given along with a summary of studies performed to date. This is followed

by the presentation of a maximum-likelihood algorithm for the direct inversion method. A discussion of

the merits and weaknesses of each method is given. Also, an appendix which discusses some general aspects of the inverse problem is included.