Broadband matched-field processing: coherent and incoherent approaches.

Area of Science:

• Acoustics
• Signal Processing
• Ocean Engineering

Background:

• Matched-field methods compare physical models to data, with accuracy depending on model fidelity and data volume.
• Broadband methods enhance data quantity and stabilize estimation, but phase prediction challenges remain.
• Coherent vs. incoherent frequency combination is debated due to acoustic phase prediction difficulties.

Purpose of the Study:

• To propose a data-consistent model for observed acoustic signals.
• To analyze causes of performance degradation in matched-field processors.
• To introduce and evaluate a novel cross-frequency incoherent processor.

Main Methods:

• Developed a data-consistent signal model: deterministic channel structure multiplied by a random perturbation factor plus noise.
• Analyzed cross-frequency channel structure and perturbation decorrelation as sources of performance loss.
• Reviewed and compared existing Bartlett processors (incoherent, coherent normalized, matched-phase) with the proposed processor.

Main Results:

• Identified cross-frequency channel structure and perturbation decorrelation as key factors degrading processor performance.
• Demonstrated analytically that the proposed cross-frequency incoherent processor matches matched-phase processor performance at the ambiguity surface maximum.
• Showcased the proposed processor's ability to avoid phase estimation, leading to extremely low computational cost.

Conclusions:

• The proposed cross-frequency incoherent processor offers a computationally efficient alternative to the matched-phase processor.
• Accurate acoustic signal estimation can be achieved without explicit phase term estimation.
• This method advances broadband matched-field processing by addressing phase prediction challenges.