Information Theory Analysis for Data Fusion

Complete theoretical development of random set unified data fusion, and place under theoretically solid information theory foundation suitable for publication. Findings: (1) Showed that optimal sensor allocation (redirection of the reallocatable sensors in a sensor suite) can be subsumed within the random set approach to information fusion, via generalization of nonlinear optimal control theory. (2) Showed that random set theory and information theory provides a common basis for performance evaluation in information fusion. Showed that parameters (e.g. target I.D. performance) can be measured in terms of information, and likewise for user defined constraints (e.g. subjective or multiple definitions of information). (3) Showed that both precise and ambiguous observations can be fused by generalizing Bayesian measurement models and the standard Bayesian recursive nonlinear filtering equations. (4) Seven chapters were completed and submitted for a book published in 1997 by Kluwer. (5) Organized a joint ONR/ARO scientific workshop on Applications and Theory of Random Sets.