Research On Edge User Allocation Strategy Based On Edge Server Performance And Task Computing Intensity

Edge user allocation problem is to allocate app users from the perspective of app service providers,and it aims to make users get lower latency and more stable services.By considering the limited computing resources of the edge server,the signal coverage area,and the user’s demand for service quality,the existing research can well solve the edge user allocation problem.However,the impact of cost control and quality of experience caused by the complexity of user tasks and the performance of edge servers cannot be ignored.These two factors have not been properly solved in existing research,so the edge user allocation problem is still challenging.Therefore,in this paper,to attack this problem,the edge user allocation problem is modeled as a combinatorial optimization problem.The detailed research contents are as follows:(1)Under the pay-as-you-go payment model,the overall user’s time cost is closely related to the cost paid by the app service provider.Therefore,in view of the influence of the computational intensity of user tasks and the performance of edge servers on user time cost,this thesis considers these two factors in the problem of edge user allocation.By building models,the problem is described as a combinatorial optimization problem.Then,an optimization method is given to minimize the overall user time cost to solve this problem.And the problem is proved to be NP-Hard problem.For large-scale scenarios,an approximation method is given to seek suboptimal solutions,and the efficiency of the approximation method is demonstrated.(2)In the real world,the user experience quality is not only related to the computational intensity of user tasks and the computing performance of edge servers,but also to the change in signal strength caused by the distance between the user and the edge server.Therefore,in the edge user allocation problem,this thesis aims to optimize the overall user experience quality.In addition to considering the computational intensity of user tasks and the performance of edge servers,the impact of signal strength on user experience quality is also considered.Through the analysis,the influence coefficient model of task intensity and signal lightness on quality of experience,the model of user’s ideal quality of experience and the user’s actual quality of experience model are given.Then,an approximation method suitable for large-scale scenarios is given,and the complexity and approximation rate of the algorithm are analyzed.Finally,experiments are designed to verify the effectiveness of this method.