Three-dimensional multiple-source focalization in an uncertain ocean environment.

Area of Science:

• Underwater acoustics
• Signal processing
• Bayesian inference

Background:

• Accurate source localization is crucial for underwater surveillance and exploration.
• Shallow water environments present challenges due to complex acoustic propagation and environmental variability.
• Existing methods often struggle with an unknown number of sources and uncertain environmental parameters.

Purpose of the Study:

• To develop a robust Bayesian focalization approach for three-dimensional (3D) source localization in shallow water.
• To handle scenarios with an unknown number of sources and uncertain environmental properties.
• To improve localization accuracy in the presence of multiple interfering sources.

Main Methods:

• Bayesian focalization using an adaptive hybrid optimization strategy.
• Minimization of the Bayesian Information Criterion (BIC).
• Metropolis and Gibbs sampling for source localization parameters (bearing, range, depth).
• Maximum-likelihood estimation for unknown source strengths and noise variance.
• An efficient source addition/deletion scheme for optimization.

Main Results:

• Demonstrated successful 3D localization of sources in shallow water with uncertain environments.
• The algorithm effectively handles an unknown number of sources.
• Accurate localization of a quiet source was achieved despite multiple interferers using a horizontal line array.
• Implicit sampling of source strengths and noise variance improved computational efficiency.

Conclusions:

• The proposed Bayesian focalization approach offers a powerful tool for 3D source localization in challenging shallow water conditions.
• The method's ability to manage unknown environmental properties and source counts enhances its practical applicability.
• This technique shows promise for applications requiring precise underwater source identification, even in complex acoustic scenarios.