Decision Network Based BMD Battle Management Algorithms

"The IET Team, consisting of IET, Booz Allen Hamilton, Oregon State University, and George Mason University, proposes intelligent, integrated, BMD BM/C2 solution concept development, by applying Decision Network technology that is a decision-theoreticextension of Bayesian Network inference technology. The proposed approach consists of three steps: (1) Formulate multiple-agent, distributed BM/C2 problems in a large-scale Decision Network. (2) Generate a coarse Decision Network by applying appropriateDecision Network coarsening techniques such as the node aggregation method. (3) Finally let each local BM/C2 agent generate a finer Decision Network to optimally make local decisions based on locally available information. This three-step approach isbased on informational centralization, process distribution by replicating centralized decision-making design, coarse modeling by aggregation, and local refinement. The Phase I tasks include development of the solution concept following this approach, anddesign of the process for the algorithm development that is planned for the Phase II efforts. The result of this Phase I effort will be distributed BMD BM/C2 solution concepts based on Decision Network technology. These will include effectivemultiple-agent, distributed decision-making algorithms not only for BMD BM/C2 functions but also for any large-scale distributed decision-making system. The resulting technology will be applicable to m