Effectiveness of mode filtering: A comparison of matched-field and matched-mode processing

Publisher Website
Google Scholar
Abstract
In a previous paper [T. C. Yang, J. Acoust. Soc. Am. 8 2, 1736–1745 (1987)] an eigenvector (EIG) mode decomposition method was used as the basis for modal beamforming (matched‐mode processing). In this article, it is pointed out that matched‐field processing is equivalent to a modal beamforming method with a different mode decomposition algorithm (denoted as the MFP‐mode decomposition method). The difference in performance between the two processing schemes can be understood in terms of the effectiveness of mode filtering of the mode decomposition method used. It is found that the difference between the MFP‐ and EIG‐mode decomposition is small if the modes associated with the signal and noise are adequately sampled by a (filled) vertical array, and potentially large if they are not. The difference is analyzed analytically and illustrated numerically using vertical arrays in an Arctic and Pacific environment. Array gain and peak‐to‐sidelobe ratio are evaluated for the MFP and EIG processors under identical environmental conditions, and compared with the matched‐mode processing assuming perfect mode decomposition. Examples are given to illustrate the use of mode filtering to suppress the sidelobes caused by sound‐speed mismatch in the upper (<1000 m) water column in the Pacific. It is found that higher‐order modes are essential in keeping the sidelobe level down, despite the fact that they are not as well resolved as the lower‐order modes. The (lower) modes that are responsible for the sidelobes (due to the sound‐speed mismatch) are identified. It is concluded that source localization is possible in a mismatch environment, if proper mode filtering is applied to the data.

This publication has 0 references indexed in Scilit: