Decision Network Based BMD Battle Management Algorithms

MDA's Project Hercules is developing a Ballistic Missile Defense System Decision Architecture that combines physics-based inference models with decision networks to provide a framework for advanced battle management, discrimination, and tracking algorithmsto address emerging BMD countermeasure regimes. Information Extraction and Transport, Inc. (IET) proposes Phase II development of a distributed processing version of decision network based battle management algorithms to support the evolution andtransition of technology developed by Project Hercules into deployed MDA systems. In order to solve difficulties associated with distributed, multiple-agent, decision-making systems, IET proposes an approach consisting of the following steps: 1) modify,as necessary, the information structure so that all the vital information is shared by all the relevant agents; 2) establish coordination among the decision-making agents so that all the decision agents use identical decision-making software to avoid anyinconsistency; and 3) make each agent solve a common global planning problem with a coarse problem resolution before solving its own BM/C2 decision problem with a finer resolution using detailed local information. This approach enables us to providepractical, yet very effective, near-optimal distributed solutions to BMDS BM/C2 DA. The proposed Phase II efforts will generate algorithm and software architecture based on this approach.