A MATCHED-MODE PROCESSING TECHNIQUE FOR LOCALIZING A TRANSIENT SOURCE IN THE TIME DOMAIN

A matched-mode processing method separates the range and depth estimation tasks, thereby obtaining a sequential procedure that circumvents the need for a two-dimensional search over all hypothesized source locations. Therefore, when applicable, it is more efficient than a matched-field method. Here, the conventional frequency-domain, matched-mode method is extended to deal with more realistic nonharmonic, narrow-band signals directly in the time domain. The modes are separated by the conventional modal filtering approach. Then, the source range and depth information can be estimated by the relative arrival times and energies of a relatively well-behaved subset of the modal arrivals. In this way, the problem of modal phase ambiguity is removed, and the entire frequency content of the modal arrivals is coherently combined by operating directly on the time waveforms. Moreover, this approach does not require a prior knowledge of the source waveform. In the numerical simulation, the source and receiver arrangement and the ocean parameters are taken from an experimental setup in the Mediterranean Sea [S. M. Jesus, J. Acoust. Soc. Am. 90, 2034-2041 (1991)].