Research of New Cloud Agent Game Model

In the cloud computing environment, each user expects to achieve higher quality of service. Different quality of service (QoS) requirements for users implies their different preferences. When each user tends to choose the resources with the prime computing power, some resources with strong computing power will be selected by multiple users, resulting in the overload of these resources and the decline of comprehensive performance. At this time, the overall resource allocation of the system is not optimal. Accordingly, a number of users have formed a competitive relationship with some key resources. Aiming at those problems, a cloud agent game model has been designed in this paper. Under the premise of satisfying the user’s QoS multi-objective optimization requirements, all cloud agents in the cloud computing system form a decentralized self organizing cloud agent system. The cloud proxy is created along with the request of the individual task scheduling, and the cloud proxy obtains the optimal resource through the mode of autonomous competition game. Then, revocation occurs as the users’ task scheduling completes, while the whole task scheduling process has dynamic scalability and adaptability, so that the overall system achieves the interest of the whole.