A Multi-agent Simulation Framework to Support Agent Interactions under Different Domains

Abstract

The ability to study complex systems has become feasible with the new intensive computing resources such as GPU, multi-core, clusters, and Cloud infrastructures. Many companies and scientific appli-cations use multi-agent modeling and simulation platforms to study com-plex processes where analytical approach is not feasible. In this paper, we use two negotiation protocols to generalize the interaction behaviors be-tween agents in multi-agent environments. The negotiation protocols are enforced by a domain-independent marketplace agent. In order to provide the agents with flexible language structure, a domain-dependent ontology is used. The integration of the domain-independent marketplace with the domain-dependent language ontology is accomplished through an auto-matic code generation tool. The tool simplifies deploying the framework for a specific domain of interest. Our methodology is implemented in FD-DEVS simulation environment and SES ontological framework.