AN ARTIFICIAL INTELLIGENCE NEURAL NET COMPUTER WHICH INTEGRATES DATA FROM ARTIS OPUS AND MK XV TO IDENTIFY AIR TARGETS

PRESENT AND PLANNED AIR-TO-AIR MISSILES HAVE EFFECTIVE RANGES THAT GREATLY EXCEED A PILOT'S ABILITY TO IDENTIFY AN AIRCRAFT. AS A RESULT, THE EFFECTIVENESS OF THESE WEAPONS IS GREATLY REDUCED. SYSTEMS SUCH AS ARTIS, OPUS AND MK XV IFF INCREASE THE RANGE AT WHICH AIRCRAFT CAN BE IDENTIFIED, BUT EACH HAS A CHARACTERISTIC PERFORMANCE ENVELOPED IN WHICH IT FUNCTIONS OPTIMALLY. WHAT IS NEEDED IS AN ARTIFICIAL INTELLIGENT SYSTEM INTEGRATION DEVICE WHICH CAN FUSE THE DATA FROM EACH OF THESE SYSTEMS TO PRODUCE A SINGLE RELIABLE AND UNAMBIGUOUS IDENTIFICATION OF AIR TARGETS. BY COMBINING DATA, IDS AT GREATER RANGE AND IN LESS OPTIMAL CONDITIONS SHOULD BE POSSIBLE. THIS PHASE I PROPOSAL IS FOR AN ARTIFICAL INTELLIGENCE PARALLELED PRO-CESSING COMPUTER BASED UPON NEURAL NETWORK ARCHITECTURES WHICH CAN PERFORM THIS SENSOR INTEGRATION WITH EXTREME SPEED AND ACCURACY. IN ADDITION, THIS RESEARCH WILL TEST THE FEASIBILITY OF USING THE SAME NEURAL NET DESIGN AS A PATTERN RECOGNIZER TO ENHANCE THE SPEED AND ACCURACY OF THE CURRENT ARTIS AND OPUS SIGNAL PROCESSING ALGORITHMS TO PRODUCE FASTER, MORE ACCURATE IDENTIFICATIONS AT THE INDIVIDUAL SYSTEM LEVEL THAN IS NOW POSSIBLE.