Modeling Human Decision Making and Agent-Based Modeling of C3 Architectures in Warfare Assessment Models

Toyon Research Corporation proposes to develop the capability within a constructive simulation to model military command and control decisions given information from many sources. In particular, Toyon will address the problem of intelligently redeploying forces given changing conditions reflected in the information provided by military sensors, including complex reactions such as learning from past experience. Toyon will use a layered approach; one layer will be a threat propagation layer, where single elements are modeled by a dynamic Bayesian network and relations between elements are modeled by a cellular-automaton based approach. The second layer will scalably simulate human command and control decisions by modeling each asset as an agent bound by simple local rules. In Phase I, we will develop these algorithms to specifically address the problem of re-deployment of ISR assets. We will show the utility of our work in a proof-of-concept scenario where the complex behavior of the cellular automaton and agent-based algorithms will allow a richer decision-making process than a simplistic rule-based decision tree. In Phase II, we shall design and implement a prototype model integrated within a mission-level simulation to model sophisticated C3 behaviors in a campaign-level scenario of utility to the Navy.