THREAT IDENTIFICATION ERROR REDUCTION TECHNIQUES

A TEO-PHASE INVESTIGATION OF TECHNIQUES FOR IMPROVING THE CONFIDENCE LEVEL OF EMITTER DETECTION AND IDENTIFICATION IN ELECTRONIC WARFARE SYSTEMS IS PROPOSED. PHASE I WILL CONCENTRATE ON THE DEVELOPMENT OF LOW LEVEL IMAGE PROCESSING TECHNIQUES APPLIED TO HISTOGRAMMED RADAR WARNING RECEIVER PULSE DATA IN ORDER TO LOCATE OBVIOUS EMITTERS AND AREAS OF PROBABLE AMBIGUITY. PHASE II WILL APPLY EXPERT SYSTEM ANALYSIS TECHNIQUES TO THE RESULTS OF THE (PHASE I) IMAGE ANALYSIS AND THE RESULTS OF CONVENTIONAL PROCESSING TO REDUCE ERRONEOUS EMITTER DETECTIONS AND IDENTIFICATIONS TO AN ACCEPTABLE LEVEL. THE ALGORITHMS WILL BE DEVELOPED ON COMMERCIALLY AVAILABLE COMPUTERS AND TESTED BOTH AT AMHERST SYSTEMS USING IN-HOUSE ASSETS, AND AT THE U.S. AIR FORCE WRIGHT AERONAUTICAL LABORATORIES USING GOVERNMENT FURNISHED RECEIVERS/PROCESSORS.