Simulations of matched-field processing in a deep-water Pacific environment

Abstract

Matched-field processing is a signal processing technique for arrays in which field vectors for assumed source positions (range and depth) are substituted for plane-wave steering vectors in conventional linear and nonlinear beamformers. The field vectors are computed by standard acoustic field models (FFP, normal mode, etc.) which take into account propagation effects in an oceanic waveguide. The output is an ambiguity surface over possible source positions in which a peak is expected at the true source position. Accuracy of the computed fields is limited in large part by our knowledge of the environment. This environmental mismatch causes degradation in localization performance, sometimes leading to large errors in estimation of source position. In order to assess the significance of this effect, simulations were performed in which a measured field is synthesized using a slightly different environmental model from that used for the steering vectors. The differences were introduced to simulate expected errors in sound-speed profile, sediment thickness, and elastic wave speed. Calculations were made for a cw source operating at 10 Hz and depths of 25 and 250 m in a 500 m-deep ocean. The receiver was a 16-element vertical array at ranges of 25 (shadow zone) and 100 km (second convergence zone). A typical Pacific sound velocity profile was assumed. The bottom was modeled by a thin (50-m) sediment layer overlying an elastic subbottom. Degradation in localization performance due to environmental mismatch will be discussed both quantitatively and qualitatively.

A multilevel sampling method for detecting sources in a stratified ocean waveguide

“…Locating acoustic sources in waveguides has been widely studied, say, [5,6,30,33,34]. One of the significant methods is the "matched-field processing" method which was proposed in 1976 in [6].…”

Source localization in the presence of internal waves

“…Both types of parameters were considered in the original work on focalization, 4 which we refer to as deterministic focalization. Before the concept of a parameter hierarchy was introduced, the problem of environmental uncertainty appeared to severely restrict the applicability of matched-field processing, and research on this topic was limited to illustrating the failure of matched-field processing under different types of uncertainties [14][15][16][17] and attempting to reduce the problem by designing ambiguity functions that are relatively insensitive to uncertainties. [18][19][20] The inverse problem involving both types of parameters had never been considered because it seemed impossible to simultaneously determine both types of parameters.…”

Matched Field Processing in Ocean Acoustics

“…Environmental mismatch has been studied early in great detail, [79,94,29,40,80). Environmental mismatch refers to uncertainty in the propagation model, e.g.…”