A method is introduced for using a single underwater sensor and knowledge o
f a shallow water environment to track a moving, multi-frequency, acoustic
source. The motion of the source creates a synthetic aperture that is adapt
ively matched against an acoustic propagation model in a manner similar to
matched held processing. However, in this case, many more unknowns must sim
ultaneously be determined. Results are presented for tracking using experim
ental data subject to unknown source speed, heading, level, and initial pos
ition. A new parallel adaptive method manages these simultaneous unknowns.
Perturbations in the sound speed based on empirical orthogonal functions ar
e used to extend the approach to operating when there also is uncertainty i
n this parameter. The performance of the tracking process with hydrophone d
ata for various signal-to-noise cases is considered. (C) 1999 Acoustical So
ciety of America. [S0001-4966(99)01209-6].