The sound at an acoustic sensor in the ocean is amplitude modulated by various mechanisms that include the motion of the source, the time- and space-variability of the propagation medium, and the character of the receiver. We have discovered that the resultant modulation waveform, while complex, nevertheless contains information that is useful to passive sonar, and that can be exploited by suitably designed data analysis (signal processing) algorithms. We have shown in Phase I that these concepts have potential value to the ADS system, in the areas of brad band signal detection, signature formation and classification. The Phase II work will iteratively optimize the algorithms for ADS by exercising them on GFI data sets, measure their performance, and assess their system impact.
Keywords: SonarADSData AnalysisSignature FormationENVARSignal ProcessingDetection