In its simplest form, data fusion extracts information relevant to the commander's needs from all available sources and combines it into a complete picture of his tactical situation. In spite of the desire for a global solution to the data fusion problem, researchers have so far been successful in developing only partial solutions. Algorithms that have been developed for and tuned to a particular application and set of sensors tend to work reasonably well for the specific problem they were developed to address and poorly or not at all for other problems. These factors point to the need to develop a data fusion management technology that can select the algorithm best able to produce the information needed by the consumer and adapt it for the best performance/cost tradeoff. The proposed effort focuses on the development of enabling distributed data fusion management technology. It will formulate and evaluate candidate fusion management approaches based on advanced processing technologies and provide a feasibility demonstration of the most promising.
Keywords: DATA FUSION DISTRIBUTED INFERENCE FUSION MANAGEMENT EXPERT SYSTEMS NEURAL NETS FUZZY LOGIC MULTIPLE
