Research And Analysis On Prediction Model Of Cloud User Behavior Analysis Based On Game Theory

With the flourish development of cloud computing technology, the new theory and new technology of cloud computing is more and more approved by people over time. In the process of moving toward mainstream, new products of cloud computing technology is also misappropriation of traditional network technology products market. However, as is limited by the cloud itself, the technical conditions and operating modes, safety issues have gradually revealed and is widely concerned by scholars and the public. How to address existing deficiencies and unknown risks in a better way, help it a better and faster development, are the current problems to be solved.In this paper, we use the game theory and focus on user behavior as the research point, put the cloud users and cloud service providers as opposing players. First, we build an prediction model of cloud user behavior analysis based on incomplete information static game theory, solving the feasible equilibrium in the model through equivalent method, then analyse the effect on cloud user behavior on different profit and loss condition and obtain reasonable security management program. Secondly, for the differences between analyse conditions of incomplete information static game theory and real situation, we build an prediction model of cloud user behavior analysis based on incomplete information dynamic game theory, put historical behavior of cloud user as calculated basis, inferred assess from cloud provider to cloud user, and update assess from cloud user after repeated game, to reach an conclusion more accurately. Thirdly, for the deficiency of traditional game theory analyse conditions, we build an prediction model of cloud user behavior analysis based on evolutionary game theory, regard cloud user and cloud provider as opposite population, depict strategy selection trend of participator behavior with replicator dynamics equation, describe convergence trend of dynamic situation with evolutionary stable strategy, verify local stability of stable points with Jacobian matrix, analyse the effect on cloud user behavior on different profit and loss condition, and gives out reasonable security management suggestions for manager consider. Through a case study, we also designed an experiment over analysis for all the 3 models to show the effectiveness of the proposed method.Focus on cloud user behavior security analysis and prediction lead cloud user behavior more normative before any abnormal behavior, inhibit abnormal behavior in an early time, make cloud system security more driving and behavior foreseeable, the use of game theory on user behavior security provides more choices for management to formulate strategies, and also provides a new idea for security management.