Bayesian tracking of multiple ocean acoustic sources with environmental uncertainties

Michael J. Wilmut, Stan E. Dosso


The article describes a Bayesian approach to the problem of simultaneous tracking of multiple acoustic sources in a shallow-water environment in which water- column and seabed properties are not well known. The Bayesian formulation is based on treating the environmental parameters, noise statistics, and locations and complex strengths of multiple sources as unknown random variables constrained by acoustic data formulated in terms of a likelihood function and be prior information. Closed form maximum-likelihood expressions for source strengths and noise variance at each frequency allow these parameters to be sampled implicitly, substantially reducing the dimensionality of the inversion. The multiple-source tracking procedure outlined in the previous section is demonstrated here with a two-source synthetic example involving shallow and deep sources moving along similar tracks.


Acoustics; Bayesian networks; Acoustic data; Bayesian approaches; Bayesian formulation; Bayesian tracking; Closed form; Deep sources; Environmental parameter; Environmental uncertainty; Likelihood functions; Multiple acoustic sources; Multiple source; Noise statistics; Noise variance; Ocean acoustics; Prior information; Seabed properties; Shallow-water; Source strength

Full Text:



  • There are currently no refbacks.