SENSOR FUSION USING ELECTRO-OPTIC AND RADAR SIGNAL MODULATION DATA

THE NAVY F-14 FIGHTER CURRENTLY HAS TWO PRIMARY SYSTEMS WHICH CAN BE USED FOR THE IDENTIFICATION OF NON-COOPERATIVE AIR TARGETS: THE TELEVISION CAMERA SET (TCS) AND A RADAR SIGNAL MODULATION (RSM) PROCESSOR COUPLED TO THE AWG-9 FIRE CONTROL RADAR. CURRENTLY, DATA FROM THESE SYSTEMS IS PROCESSED INDEPENDENTLY, AND ANY FUSTION WHICH TAKES PLACE IS PERFORMED BY THE OPERATOR. THE OBJECTIVE OF THIS RESEARCH IS TO DEVELOP TECHNIQUES FOR FUSING THE ELECTRO-OPTIC (EO) DATA FROM THE TCS WITH THE RSM DATA FROM THE AWG-9. THREE FUSION TECHNIQUES WILL BE INVESTIGATED: - FUSION USING A FEEDFORWARD NEURAL NETWORK - FUSION USING A DEMSTER/SHAFER EVIDENTIAL REASONING METHOD - FUSION USING A EUCLIDIAN NEAREST NEIGHBOR CLASSIFIER. EACH OF THESE TECHNIQUES IS EXPECTED TO DISPLAY CERTAIN ADVANTAGES AND DISADVANTAGES IN ACCURACY/PERFORMANCE. THE FINAL EFFORT UNDER THIS PROPOSAL WILL THEREFORE BE TO DESIGN AND DEMONSTRATE A HYBRID FUSION SYSTEM WHICH DRAWS FROM THE STRENGTHS OF EACH OF THE THREE INDEPENDENT FUSION TECHNIQUES. AMELEX INTENDS TO TEAM WITH GRUMMAN AIRCRAFT SYSTEMS FOR THIS RESEARCH. AMELEX WILL CONTRIBUTE EXPERTISE IN RSM PROCESSING, NEURAL NETWORKS AND SENSOR FUSION, WHILE GRUMMAN WILL CONTRIBUTE EO TARGET RECOGNITION AND SENSOR FUSION CAPABILITIES.