A computational Bayesian approach for localizing an acoustic scatterer in a stratified ocean environment

Abstract

A computational Bayesian method is presented for inference regarding the state of a submerged mobile object. The approach addresses the challenge of closely spaced multipath arrivals in refractive environments with uncertainty in ambient acoustic noise power. Vertical angles and Doppler frequencies of the arrival returns are jointly inferred and their posterior density is mapped to the object's range, depth, and speed through acoustic ray interpolation. The object is localized under the challenging constraint of a small receive vertical aperture. A case study with the classic Munk sound speed profile is presented to lend credence to the approach.