Research On Resource Allocation Algorithm Based On Mobile Edge Computing Offloading

Cloud computing,as a large data analysis and processing platform,provides powerful technical support for the development of economy and society with its massive storage capacity,powerful computing capacity and dynamic resource scheduling characteristics.However,the traditional cloud processing methods have gradually appeared such problems as load bottleneck,delay problem,large amount of network bandwidth occupied by long-distance data transmission,limited terminal computing resources,etc.Therefore,a large number of resources close to users and applications are needed for computing and data processing,so as to meet the real-time,high reliability and low cost requirements of Internet of Things applications.Mobile Edge Computing proposed in 2014,brings computing and storage resources to the edge of mobile networks,and can run high-demand applications on users to meet the strict delay requirements,so as to obtain low-latency cloud services,and solve the problems of high latency and high energy consumption in cloud computing.However,MEC is still in its early stage of development,and the related technologies are not mature enough.Especially in the process of offloading tasks to edge servers,how to make offloading decisions,when to offload,where to offload,and how to allocate resources after offloading tasks are all the urgent problems to be solved by MEC technology.In this dissertation,game theory and some classical algorithms are used to solve the problem of offloading decision and resource allocation in the process of task offloading.The specific research contents are as follows:1.A resource allocation algorithm based on Lagrange duality is proposed to solve the problem of low throughput caused by low resource utilization when users offload tasks to MEC servers.We first explore the problem of computing offloading in multi-channel environment.Then an optimization problem is proposed to maximum the overall throughput of the offloading system,where the power control and interference are taken into consideration.We transform the original optimization problem into a convex optimization problem,and use the Lagrangian dual algorithm to establish the KKT condition.By continuous iterative operation,the user's transmit power is optimally achieved.Simulation results show that our proposed scheme not only guarantees the communication quality,but also improves the system throughput and spectrum utilization.2.In order to improve the energy efficiency at the edge server and the utilization of resources by users.We proposed a heterogeneous network of macro cells and small cells equipped with edge servers,to deal with the problem of resource utilization and offloading efficiency between edge servers and small cells.We combine the stackelberg game model to define the MEC edge server as the leader,and the small cell base station as the follower.We denote the utility of small cell base station as the energy efficiency using the edge sever,and express the utility of the MEC sever as the overhead of computing offloading.A utility function based on benefit and overhead is designed,and the utility function satisfies the condition of the concave function,guarantees the existence of the Nash equilibrium point.Then the corresponding algorithm for the edge server and the small cell base station are designed.Finally,the optimal strategy and Nash equilibrium are achieved.The simulation results show that the proposed algorithm is reasonable during the offloading process,and the optimal resource allocation of leaders and followers is realized.