Detection on a distributed network using information geometry.

Area of Science:

• Underwater acoustics
• Signal processing
• Information geometry

Background:

• Passive detection schemes are crucial for identifying targets in complex environments.
• Information geometry offers a novel framework for analyzing signal statistics.
• The Distributed Network Consensus experiment provided multi-sensor acoustic data.

Purpose of the Study:

• To evaluate a passive detection scheme based on information geometry using real-world underwater acoustic data.
• To assess the performance of geometric detectors in low signal-to-noise ratio (SNR) conditions.

Main Methods:

• Utilized a passive detection scheme employing information geometry.
• Interpreted signal excess as the non-Euclidean distance between probability distributions on a Riemannian manifold.
• Tested the scheme on underwater acoustic field data from the Distributed Network Consensus experiment.
• Artificially lowered the signal-to-noise ratio (SNR) to simulate challenging low SNR scenarios.

Main Results:

• The geometric detectors demonstrated robust performance.
• Effective target detection was achieved by identifying significant differences between signal and broadband background statistics.
• The method proved successful even when the SNR was artificially lowered.

Conclusions:

• Passive detection using information geometry is a viable and effective strategy for underwater acoustic surveillance.
• The geometric approach shows promise for reliable signal detection in challenging, low SNR environments.