Basic Research for Level 2 Information Fusion

ORINCON's proposed effort will address fundamental issues in the research and development of Level 2 fusion, namely, development of a solid mathematical foundation on which to build algorithms for situation refinement that includes organization andstructure. We believe that the fundamental approach to fusion in Level 1 - that is, detection, prediction, association, hypothesis generation, hypothesis management, and update - is applicable to Level 2. Examining the key elements of Level 1 fusion, wecan identify the parallel components of Level 2 fusion and evaluate the development in the underlying fundamental theory. As part of our Phase I work, we have laid the foundation by considering metrics that arise in the setting of discrete sets, thusbuilding a theoretical foundation that will provide a setting in which to develop and evaluate metrics, and the application of these metrics to the Level 2 problems of detection, prediction, and association. In Phase II, it becomes necessary to extend thisconcept of measure between two discrete sets to understanding the statistical features used throughout the Detection-Associate-Update steps by defining probability measures that occur naturally from the defined metrics. Accomplishing these objectives willprovide a basis for rapid development of future Level 2 fusion.