Process algebra and constraint programming for modeling interactions in MAS

Abstract

We show how techniques from the realm of concurrent computation can be adapted for modeling the interactions of agents in multi-agent systems. In particular, we introduce a general process-algebraic approach to modeling multiagent systems. Our approach consists of an integration of the process algebras of Communicating Sequential Processes (CSP) and Concurrent Constraint Programming (CCP) for modeling the communication, synchronization and coordination in multiagent systems, including FIPA-style communication primitives and a programming language for group actions in a multi-agent system.