ASW SEARCH PLANNING

CURRENT CAPABILITIES TO UTILIZE ASW SYSTEMS AND FORMULATE SEARCH PLANS ARE LIMITED BY THE ABILITY TO PROCESS REAL-TIME TACTICAL DATA. AN AUTOMATED SYSTEM, EXPLORING STATE-OF-THE-ART DEVELOPMENTS IN NEURAL NETWORK AND PARALLEL PROCESSING TECHNOLOGY, WOULD ENHANCE THE ABILITY TO PROCESS INFORMATION IN A TIMELY MANNER, AND WOULD THUS RESULT IN AN ABILITY TO OPTIMIZE SYSTEM LINE-UP. WE PROPOSE TO PROVIDE AN IN-DEPTH ASSESSMENT OF CURRENT NEURAL NETWORK ALGORIZHMS AND PARALLEL PROCESSING ARCHITECTURES, AND THEIR SUITABILITY FOR IMPLEMENTATION IN A COMPUTER-BASED SYSTEM, BASED ON A RIGOROUS ANALYSIS OF THE REQUIREMENTS OF THE ASW SEARCH PLANNING PROBLEM. OUR REPORT WILL INCLUDE A RECOMMENDATION OF DESIGN REQUIREMENTS THAT WILL ENABLE A PROTOTYPE SYSTEM TO BE BUILT IN PHASE II OF THE PROJECT.