NORMAL-MODE MATCHING LOCALIZATION IN SHALLOW-WATER - ENVIRONMENTAL AND SYSTEM EFFECTS

Matched-field processing is a passive range and depth source localization technique that has been extensively used in shallow-water environments. A vertical array of sensors is used to spatially sample the acoustic waveguide where the source signal embedded in additive ambient noise propagates. The array output is then matched with the signal replica field generated by a normal-mode model based on the environmental parameters that characterize the waveguide. Recent results obtained from real data show the feasibility of the technique and give evidence of its strong dependence both on the array aperture and on the knowledge of the environmental parameters used in the model. This paper describes a modified matched-field technique, called normal-mode matching, that is applied to real shallow-water data. Its performance is compared to that obtained by conventional matched-field processing using the same data set. Unlike conventional matched-field processing, the results indicate that unambiguous localizations can be obtained even for "short" arrays spanning only half of the water column.