Passively localizing a quiet moving source with less prior information

Bill Kuperman had a restless energy and insatiable appetite for new research topics, which led him to contribute in many areas of acoustics including rough elastic surface scattering, sea surface noise, matched field processing (MFP), time reversal, and even the acoustic properties of neurons. When I first met Bill, he summarized his research interest as finding ways of “doing something with less.” In the context of passive acoustic localization, this amounted to finding a source with lower signal-to-noise ratio (SNR) or with a smaller array or with less knowledge of the acoustic environment. This led us to develop my thesis work on “range-coherent” matched field processing which coherently combines data collected from a moving source traversing multiple range cells. The method successfully localized low SNR tonal sources in the SWellEx-96 experiment. Another topic of interest to Bill was the use of ambient noise as a signal. In this spirit, we developed a normal mode estimation algorithm using a partially spanning vertical line array (as opposed to a full water column spanning array) which we called modal-MUSIC. We were then able to use the normal modes estimated from noise for low-SNR localization with range-coherent MFP. In addition to his many technical contributions to the field of underwater acoustics, Bill had a gift for mentorship and was a joy to work for and with.