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

Area of Science:

• Underwater acoustics
• Computational Bayesian inference
• Mobile object state estimation

Background:

• Accurate localization of submerged mobile objects is crucial for various applications.
• Challenges include multipath arrivals in refractive environments and noise uncertainty.
• Limited vertical aperture of receivers complicates precise object detection.

Purpose of the Study:

• To develop a computational Bayesian method for inferring the state of submerged mobile objects.
• To address challenges posed by closely spaced multipath arrivals and uncertain acoustic noise power.
• To enable object localization with a small receive vertical aperture.

Main Methods:

• Joint inference of vertical angles and Doppler frequencies from arrival returns.
• Mapping posterior density to object range, depth, and speed via acoustic ray interpolation.
• Utilizing a computational Bayesian framework for state estimation.

Main Results:

• The method successfully infers object state parameters (range, depth, speed).
• Demonstrated effectiveness in refractive environments with multipath interference.
• Validated through a case study using the Munk sound speed profile.

Conclusions:

• The presented Bayesian method provides a robust approach for submerged object localization.
• It effectively handles complex acoustic conditions and sensor limitations.
• Offers a reliable tool for underwater surveillance and navigation.