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

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 applications use multi-agent modeling and simulation platforms to study complex processes where analytical approach is not feasible. In this paper, we use two negotiation protocols to generalize the interaction behaviors between 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 automatic 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.