A waveguide invariant adaptive matched filter for active sonar target depth classification.

Area of Science:

• Acoustics
• Signal Processing
• Oceanography

Background:

• Distinguishing underwater targets from seafloor clutter is crucial in active sonar.
• Wideband active sonar data presents challenges for accurate target depth discrimination.
• The waveguide invariant property offers potential for improved target classification.

Purpose of the Study:

• To develop a robust method for depth discrimination of water column targets from bottom clutter in wideband active sonar.
• To utilize the waveguide invariant property for enhanced target classification.
• To ensure robustness against environmental variations.

Main Methods:

• Deriving multiple target snapshots by sub-sampling Short-Time Fourier Transform (STFT) coefficients.
• Defining a Waveguide Invariant Spectral Density Matrix (WI-SDM) for adaptive matched filtering.
• Employing a Waveguide Invariant Minimum Variance Filter (WI-MVF) with depth-dependent replicas from a normal mode model.
• Incorporating Environmental Perturbation Constraints (EPC) for robustness to environmental mismatch.

Main Results:

• Demonstrated effective depth classification of underwater targets using the proposed WI-SDM and WI-MVF.
• Achieved robust performance under uncertain environmental conditions, validated by simulation and real data.
• Presented Receiver Operating Characteristics (ROC) illustrating performance gains.

Conclusions:

• The waveguide invariant approach provides a robust and effective method for active sonar target depth classification.
• The developed technique successfully discriminates water column targets from bottom clutter.
• Environmental perturbation constraints enhance the reliability of depth classification in real-world scenarios.