Physics based sparsity level determination for acoustic scattered far-field prediction.

Area of Science:

• Acoustics
• Signal Processing
• Compressed Sensing

Background:

• Sparse reconstruction with the equivalent source method shows potential for acoustic field prediction.
• Accurate acoustic field prediction requires knowledge or estimation of sparsity levels.

Purpose of the Study:

• To derive a lower bound for sparsity level in scattered far-field acoustic prediction.
• To determine the minimum number of measurements needed for accurate prediction using compressed sensing theory.

Main Methods:

• Derived the lower bound of sparsity level from the effective rank of the far-field transfer matrix.
• Utilized orthogonal matching pursuit with the pre-set sparsity hyperparameter.
• Applied compressed sensing theory to determine the minimum measurement count.

Main Results:

• Effectiveness demonstrated using simulated and tank data.
• Successfully combined physical propagation principles with compressed sensing.
• The proposed approach is easy to implement for practical applications.

Conclusions:

• The derived sparsity lower bound effectively guides sparse reconstruction for far-field acoustic prediction.
• This integrated approach offers an efficient and implementable solution for acoustic measurements.
• Validated the method's efficacy through experimental and simulated data.