E-3 SENSOR DATA FUSION ALGORITHM RESEARCH

ALGORITHMS NEEDED TO SOLVE DATA FUSTION IN THE E-3 ENVIRONMENT ARE GROUPED INTO FIVE CATEGORIES FOR THIS RESEARCH. THE PROPOSAL DESCRIBES TECHNICAL APPROACHES IN DETAIL FOR EACH: (1) TRACKING; THE USE OF THE DUAL-VELOCITY INTEGRATED ORNSTEIN-UHLENBECK MOTION MODEL KAMAN FILTER; (2) CORRELATION, THE USE OF STATISTICAL SCORING AND A UNIQUE SEQUENTIAL SOLUTION TO AN INTEGER PROGRAMMING FORMULATION OF THE MULTI-HYPOTHEIS, TRACK-TO-TRACK CORRELATION PROBLEM; (3) GRIDLOCK,THE USE OF A LEAST SQUARES SOLUTION METHODOLOGY FOR ESTIMATING BIAS ERRORS IN ALL REPORTING UNITS; (4) ERROR MANAGEMENT, THE USE OF OUTLIER DETECTION AND PROVISIONAL TRACKS; AND (5) USER INTERACTION, THE DISPLAY OF SINGLE-HYPOTHESIS DATA WHILE MAINTAINING MULTIPLE HYPOTHESIS INTERNALLY.