Research On Resource Optimization Of Mobile Cloud Gaming Based On Game Theory

Cloud gaming is a brand new cloud service model that runs games on the server and returns game video to the player client over the Internet for a new gaming experience.In the cloud game scenario,on the one hand,due to the large demand for hardware resources in the game,when the player connecting the cloud game concurrently reaches a certain scale,it will exert tremendous pressure on the data center;on the other hand,different players play the game.Quality has different needs,such as resolution,etc.In this case,how to minimize the loss of QoE(Quality of Experience)of the global user's cloud game service quality experience in the case of limited data center resources is a huge The challenge.This paper first introduces the model of edge computing and data center,points out the quantitative index of user QoE in the scene of mobile cloud game,and then shows that the optimization problem of distributed data center is an NP-hard level problem,and thus the conventional method is obtained.It is difficult to give a solution in a limited time.Therefore,this paper combines the Nash equilibrium and the potential game to find the potential function corresponding to the QoE index,which proves that the multi-user computing unloading game is a potential game,and will definitely converge to the Nash equilibrium,thus obtaining fast optimization results.Finally,this paper compares game theory with other common resource optimization algorithms to prove the excellent performance brought by the game algorithm.This paper studies the problem of data center resource allocation in mobile edge computing scenarios.By applying the game algorithm to cloud game scenarios,it can quickly give a solution to distributed optimization problems.In addition,the game algorithm proposed in this paper has very good performance,which quickly converges to the optimal value,and the time complexity is O(MlogM),where M is the physical server of the data center.In order to verify the performance of the algorithm,this paper based on the data collected by the crawler and the self-developed cloud game platform to conduct experiments,compare various common data center resource optimization strategies,such as sequential polling,random greed,etc.,from the global user QoE loss indicators for analysis,The algorithm of this paper compares the above strategy to reduce the QoE loss by at least 54%.