In today's littoral warfare environment, submarines will be called upon to operate in many diverse, challenging and complex threat environments, with many background targets, each emitting energy in several regions of the spectrum. Electronic support measures (ESM) is a tactically important tool in maintaining an understanding of this and sorting out the threats and targets from the background clutter. Tactical ESM systems are used to observe and classify/identify emitters such as radars, beacons, and other non-communications devices based on emission characteristics such as radio frequency, pulse repetition interval, pulse group repetition interval, pulse width, chirp, scan type, polarization, rf agility, etc. Techniques such as table lookup or clustering algorithms often fail in a complex environment, resulting in either misclassification or ambiguity. Fuzzy logic directly addresses this issue and has a major advantage over most other techniques such as expert systems or neural networks; an underlying mathematical basis which is derivable from first principles. Global and Penn State Applied Research Laboratory propose to integrate our in-depth background in sigint analysis and sigint data fusion with our demonstrated capabilities in fuzzy logic to prototype a fuzzy logic esm classification algorithm, perform a comparative analysis of its performance versus classical techniques, and develop a system level architecture for a fuzzy logic based classification system.
