A systematic investigation of advanced modeling and stochastic control and scheduling methodologies is being undertaken that addresses one aspects of the weapons allocation problem, i.e. several platforms with assets of different character defending against a diverse collection of targets. Models for such scenarios lead to stochastic scheduling problems which cannot be handled by conventional analytical methods. Several different analytical approaches are being explored that have the potential for synthesis of effective engagement algorithms. Stochastic dynamical models are being designed to represent the interaction of the weapons platforms and the target systems during the post-boost and mid-course phases of operations. Based on these engagement models, a two-step approach is being used to develop the weapons allocation algorithms. A prototype set of algorithms is being developed using an optimization method called the stochastic gradient method. The algorithms computed by this method is serving as a baseline for the development of more representative strategies which reflect the operational structure of the battle management (BM) weapons allocation system. Stochastic scheduling models and a set of strategies called index rules are being used to derive effective and efficient engagement strategies for a BM system involving several weapons platforms responding to a large number of targets over an extended region of space.
