ALGORITHM DEVELOPMENT

THE OBJECTIVE OF THIS RESEARCH IS TO DEVELOP IMPROVED DETECTION AND CLASSIFICATION ALGORITHMS FOR TACTICAL AIRBORNE MILLIMETER WAVE (MMW) SEEKERS EMPLOYING POLARIMETRIC PROCESSING. THE PROPOSED RESEARCH IS BASED UPON NONCOHERENT TARGET INFORMATION (NTI) FEATURES (1) EXTRACTION BY WIDEBAND CIRCULAR POLARIZATION DUAL-CHANNEL ACTIVE MMW SEEKERS. THE NTI FEATURES ARE NEAR-COMPLETE IN TARGET INFORMATION SENSE AND THEY MAP THE PSEUDO-GEOMETRIC IMAGE (APPARENT SIZE AND APPARENT SHAPE) OF THE TARGET. THE RESEARCH IS FOCUSED ON THE INTEGRATED USE OF KNOWLEDGE-BASED/EXPERT SYSTEM AND ADAPTIVE SIGNAL PROCESSING TECHNIQUES. AN EXPERT SYSTEM IS ABLE TO MONITOR ALL SUBSYSTEM FUNCTIONS SIMULTANEOUSLY, FOR EXAMPLE, CROSS-COUPLING BETWEEN THE CHANNELS, THEN USE HEURISTICS, RULES AND AN APRIORI DATA BASE TO OPTIMIZE DETECTION AND CLASSIFICATION PERFORMANCE. EMPHASIS IS PLACED ON DEVELOPING AND EVALUATING THE JOINT CAPABILITY OF SYMBOLIC AND HEURISTIC METHODS IN AN EXPERT SYSTEM WITH ADAPTIVE SIGNAL PROCESSING TECHNIQUES BASED UPON THE ASSOCIATED UNDERLYING PRINCIPLES OF INFORMATION THEORY, ESTIMATION THEORY, DECISION THEORY AND MODERN CONTROL THEORY. ADDITIONALLY, TACTICAL SYSTEM IMPLEMENTATION REQUIRES REAL-TIME OPERATIONS, WEIGHT AND SIZE LIMITATIONS OF SIGNAL PROCESSING HARDWARE AND SOFTWARE.