An Analyzable Model of Automated Service Negotiation

Negotiation is a key aspect of Service-Oriented Systems, which is rarely supported by formal models and tools for analysis. Often, service negotiation proceeds with timing, cost and resource constraints, under which the users and providers exchange information on their respective goals, until reaching a consensus. Consequently, a mathematically driven technique to analyze various ways to achieve such goals is beneficial. In this paper, we propose an analyzable negotiation model between service clients and providers, in our recently introduced language REMES and its corresponding textual service composition language HDCL. The model can be viewed as a negotiation interface for different negotiation strategies and protocols, which iterates until an agreement is reached. We show how to analyze the negotiation model against timing, cost and utility constraints, by transforming it into the Timed Automata formal framework. We illustrate our approach through an insurance scenario assuming a form of the Contract Net Protocol for web services.