Ray Path Identification in
|Josť M. F. Moura
Carnegie Mellon University
Department of Electrical and Computer Engineering
5000 Forbes Avenue
Pittsburg, PA 15213-3890
Institut Fur Meereskunde
Duusternbrooker Weg 20
D-24105 Kiel Germany
Abstract Ocean acoustic tomography (OAT) is a promising tool for the study of large-scale averaged oceanographical properties (from decades to hundreds of kilometers). This is a result of the ability of acoustic waves to propagate and sample vertically the SOFAR channel. We distinguish two major signal processing tasks. The first (localization) estimates the multipath delays for the multiple arrivals at the receiver. The second (identification) assigns these delays to the different paths as predicted by an acoustical model. Although, in principle, these two tasks can be dealt with in a single signal processing step, we will consider in this talk only the problem of identification of multiple echoes in the context of OAT. Ray identification (RI) in OAT has become a major critical problem at the signal processing level. It is related to the data association problem in radar tracking. This presentation introduces a new approach for ray identification, illustrates its application with synthetic data, and then applies it to real data collected in an acoustic tomographic experiment run in the Gulf of Gascogne.
Direct comments and questions to: firstname.lastname@example.org
© MIT Lincoln Laboratory. All rights reserved.