An Adaptive Model of Service Composition Based on Policy Driven and Multi-Agent Negotiation

This paper presents an adaptive model of service composition to achieve the optimum profits in the emerging services market, which is based on policy driven and multi-agent negotiation. The negotiation rules will be defined in the policies to determine the conditions for composition and cooperation, such as the accepted range of QoS with corresponding price charge. An agent group will be established to achieve the policy driven negotiation process for service composition, which includes policy agent, evaluation agent and action agent. Policy agent will be responsible for acquiring and parsing negotiation rules from pre-defined policies. Action agent will be responsible for the interactions among the services to reach an agreement. During the negotiation process, evaluation agent will be responsible for doing the evaluation on the bid information received from action agent according to the negotiation rules. And, the action instructions will be generated and sent to action agent for the negotiation on the bids provided by services. Through such policy driven and multi-agent negotiation approach, it can address win-win in the composition and cooperation of the services