Two-dimensional Newtonized orthogonal matching pursuit compressive beamforming

Area of Science:

• Acoustics
• Signal Processing
• Array Signal Processing

Background:

• Conventional compressive beamforming struggles with acoustic sources located between grid points, leading to performance degradation due to basis mismatch.
• Accurate acoustic source localization is crucial in various fields, including sonar, radar, and medical imaging.

Purpose of the Study:

• To propose a novel two-dimensional Newtonized orthogonal matching pursuit compressive beamforming method to address the basis mismatch issue.
• To enhance the accuracy of acoustic source position and strength estimation in both near and far fields.
• To develop single and multiple snapshot versions of the algorithm applicable to arbitrary planar array geometries.

Main Methods:

• Constructs a maximum likelihood estimation model with source position and strength as parameters on a continuous 2D plane.
• Employs a two-dimensional Newton optimization combined with a feedback mechanism for precise coordinate refinement.
• Utilizes orthogonal matching pursuit for efficient sparse signal recovery.

Main Results:

• The proposed method effectively overcomes the basis mismatch problem, achieving high spatial resolution.
• Accurate estimation of acoustic source position and strength is demonstrated through simulations and experiments.
• The multiple snapshot version shows superior performance compared to the single snapshot version, particularly in low signal-to-noise ratio conditions.

Conclusions:

• The developed Newtonized orthogonal matching pursuit compressive beamforming offers a robust solution for acoustic source localization.
• The method provides accurate results without requiring prior knowledge of signal-to-noise ratio or regularization parameters.
• Increased number of snapshots significantly improves localization accuracy, highlighting the advantage of the multiple snapshot approach.