Cooperative Offloading Strategy Based On Mobile Edge Computing

With the development of technology,a diverse variety of computing-intensive applications are emerging and attracting great attention.However,due to the restrictions of the MDs on size,weight,battery life,and heat dissipation,the gap between the capability of limited computing resources and demand for executing complex applications is gradually increasing.MEC system has been introduced to bring computing and storage resources in close physical proximity of the wireless end devices.MEC server helps to process the computing job offloaded by users within the coverage so as to solve the problem of insufficient computing resource of mobile device.First of all,the offloading strategy is determined by the wireless access conditions.Considering that millimeter wave will be an important communication carrier in 5g era,the channel of MEC system is modeled as a large-scale MIMO system under rice channel,and the uplink reachable rate of the system is analyzed and verified by simulation in this thesis.Based on this,according to Shannon's theorem,the offloading strategies determined by the channel conditions are given when the number of antennas tends to infinity or any value.With the combination offloading strategy and channel conditions,the strategy can be controlled by adjusting the physical layer parameters.Then,this thesis uses the game of system profit maximization to model the interactive decision-making process between users,and solves the congestion control problem caused by multi-user sharing MEC system computing resources.The queue model is used to model the processing of computing tasks,and a utility function is proposed based on the unloading strategy related to channel conditions in the thesis.In the game,ICRM and SCRM are proposed according to whether or not the core cloud server is involved in the calculation.The sufficient and necessary conditions for the existence of the unique positive Nash equilibrium solution are analyzed respectively under the both modes,and the distributed algorithm is used to solve the optimal offloading strategy.Simulation results show that the performance of the two models is excellent,and ICRM can effectively improve the profit of users and system and reduce congestion compared with SCRM.In this thesis,we study the job collaborative offloading strategy between user and MEC server and between user and user in MEC system,so as to make full use of the computing resources of mobile device,MEC server and central cloud server.And a new direction for later research on the job offloading strategy in MEC system is provided.