An Automated Trust Negotiation Framework Based On Extended Game

After the long-term development of internet technology, the traditional software system has been overturned in the aspect of technological framework and theory. And it proceeds towards the direction of mass data and extensive concurrency rapidly. Around the distributed storage, Cloud Computing and virtualization framework, several researches, such as web semantics technology, incremental computation and Virtual Agent have come forward. The traditional CS system can't overtake and confront such complicated web application. P2P-based web application has gone with the stream and induced a number of services, such as Skype telephone, Thunder and kinds of streaming media tools.Because of the openness and distributivity of P2P web, and lack of core node to be in charge, there were a lot of fictitious or malicious nodes in the event of applications, resulting in the decreasing of P2P web service ability. And the way to establish the trust among the nodes safe and effectively has become the focus. The approval method for username/password can't adapt the equivalent web environment. Despite the fame system has solved the problem to some degrees, the need of real name certification would be apt to be deceived and attacked by the virtual node in the event of many nodes existing. Only the recent automatic trust negotiation web certification model could satisfy the need of safety, privacy and efficiency of P2P web, due to demand both sides establish trust by produce certifications alternately following the strategy.Automatic trust negotiation theory has described proceed of trust establishment by producing the right certification between the upsides entity, which could not only apply in P2P web environment. Existing theory only abstract the operation and logic, but not give a exact definition for the problems such as how to define the certification and strategy, the flow for the system consistency checking. Available studies have also elaborated it and advanced the some trust negotiation framework applied in different aspects, such as PSPL, TPL, Trust-X, KeyNote, etc. According to the researches before, our study discuss the possibility of developing the automatic trust negotiation with the game model, and then put forward a new frame Trust-Once, which is a automatic trust negotiation frame based expanding game model. This frame is fit for P2P, which has matched needs of efficiency and safety, described the details exactly both for strategy and definition for certification, flow of trust negotiation and consistency checking. It has also argued the advantage and disadvantage of available choices. In the end, our study mentioned the design and implement of Trust-Once system, and discuss the function of establishing the negotiation tree and consistency checking. Trust-Once negotiation frame was applicable to P2P web environment, which optimized the system with game theory, matched both the efficiency and safety in theory, and developed a model that could be engineered, all of which has provided the evidence for both the further theory research and the development of related intermediate product.