Toward a Theoretical Foundation for Multi-Agent Coordinated Decisions

A system of M agents coordinates if the agents possess and use information regarding the existence, decisions, or declsion-maklng strategies of each other. Inter-agent relationships may be expressed through a coordination function, which is a joint probability mass (or density) function defined over a 2M-dimenslonal coordination space. The coordination function expresses the joint relationships between all agents in the system as characterized by agent goals and abductive considerations such as cost, hazard, and resource consumption. The coordination function is used to derive two joint utilities, termed accuracy and rejectability, and Levi’s joint rule of epistemic utility is applied to identify the satisfieing set of decision vectors that represent jointly rational behavior for the system. By intersecting the sarisricing sets obtained under differing model as= sumptions, a robust set of satlsficing decisions may be obtained. Topic Areas: Cooperation and coordination; Conceptual and theoretical foundations.