Active Sonar Target Discrimination in Littoral Waters

By re-examining the basic physics of underwater signal propagation in the context of recently developed notions of state-space methods for dynamical systems, we have formulated a new active sonar signal-processing model. This model is based on morerealistic assumptions about the possible return waveforms, nongaussian noise, clutter and reverberation than existing approaches. Implementing a bank of optimized test statistics appropriate for mid-frequency, broadband waveforms, we have demonstrated a 5to 6 dB improvement over the conventional replica correlator for a developmental, at-sea active sonar system. The potential gains in tactical systems is larger as the data available for Phase I was not pre-processed consistent with our requirements,Simulations indicate gains of as much as 10 dB are possible in Phase II. Our processor is also easily integrated into existing systems and is computationally efficient, which makes it attractive for other applications. In this project, we will use ourpre-existing technology to specify, construct, and test a new class of active processors for waveforms of interest to the sponsor with the objective of demonstrating similar improvements for fleet sonars. The benefits in active sonar include increaseddetection range and higher detection confidence at a reduced false alarm rate. Further, these same benefits can be expected for pulse compression and broadband radar. This technology also has applications in seismic survey technologies.