A Novel Sparsity Adaptive Algorithm for Underwater Acoustic Signal Reconstruction

Area of Science:

• Signal Processing
• Acoustics
• Data Science

Background:

• Traditional algorithms for underwater acoustic signal reconstruction rely on known sparsity, limiting their effectiveness with compressed data.
• Existing methods often suffer from high computation times and suboptimal reconstruction quality.

Purpose of the Study:

• To propose a novel sparsity adaptive and variable step-size matching pursuit (SAVSMP) algorithm for precise underwater acoustic signal reconstruction.
• To address the limitations of existing algorithms in terms of accuracy and computational efficiency.

Main Methods:

• The SAVSMP algorithm estimates initial sparsity using the Restricted Isometry Property (RIP) criterion.
• Curve fitting refines sparsity estimation, preventing under- or overestimation, while adaptive step size ensures accurate sparsity level approximation.
• Atom selection via matching pursuit and filtering unsuitable atoms using the Least Squares Method enable precise signal reconstruction.

Main Results:

• The proposed SAVSMP algorithm demonstrates superior performance compared to traditional methods.
• Experimental results show significant improvements in reconstruction quality for underwater acoustic signals.
• The algorithm effectively reduces computation time while enhancing accuracy.

Conclusions:

• The SAVSMP algorithm offers a robust and efficient solution for reconstructing underwater acoustic signals from low-dimensional compressed data.
• This advancement is crucial for improving the performance of sonar systems and underwater acoustic analysis.