Research On The Application Of SLA Negotiation Mechanism Based On Game Theory In Cloud Computing

With the development of cloud computing, there are more and more providers, and the cloud computing market become more open and competitive. It is difficult for users to choose suitable provider, and the providers can’t provide the best service in the negotiated price. In the cloud computing environment, the SLA negotiation between providers and users is usually based on the traditional game theory. But for participants in the real economic life, it is difficult to meet the conditions of traditional game theory. In the enterprise competition and cooperation, participants show differentiation and have bounded rationality, and the lack of information caused by the complexity of economic environment and game theory is obviously. There are three aspects in the research of the above problems gradually as following:(1) In cloud computing environment, the providers and the users themselves can’t negotiate automatically, thus we proposed cloud service agent architecture based on dynamic game theory. Based on this architecture, due to the certain dynamic of the game interests of the game between providers and users, a framework of SLA negotiation based on evolutionary game theory was proposed. This framework can not only negotiate successfully between the parties, but also clearly show the dynamic evolutionary game process of participant groups and the internal.(2) A bilateral SLA negotiation model of cloud service agent based on dynamic game theory. Firstly, according to the negotiation process, nonlinear model, transaction period model and exponential function model were proposed and the contrast of the three models is analyzed. Secondly, according to the bilateral SLA negotiation model, Nash equilibrium point and satisfaction degree difference algorithm was proposed. This algorithm does not need to calculate all variable values in game matrix, thus the calculation steps greatly decrease. Finally, experimental results show that the users and the providers are able to achieve the same degree of satisfaction in the price and bandwidth.(3) A multi-factor SLA negotiation model based on evolutionary game theory. Firstly, this paper established a general two asymmetric game model, and deduced the replicator dynamics and evolutionary stable strategy. Secondly, this paper constructed an evolutionary game theory model between providers and users, and dynamically evolved the behavior of providers and users. Because Jacobi matrix reflects the optimal linear approximation of a differential equation and the given points, by analyzing the Jacobi matrix of the system whether the stable point of the system is evolutionary stable strategy can be judged. Finally,we simulated and analyzed the evolutionary game model.