SBIR-STTR Award

The most significant causes of poor detector performance are unrealistic and incomplete models for the true signal and noise statistics. Starting from first principles, we have developed a signal-processing model that incorporates more robust assumptions than is typical of published approaches for active sonar. Our new methods include robust estimators for removing reverberation and background noise to enhance detection of signals in difficult environments and a methodology for constructing a replica representation basis for classification of active sonar returns. An existing version of this processor has already demonstrated up to 10 dB deflection ratio gain on passive underwater ASW signals.