User's Mobility Based Tasks Offloading And Migration Schemes With Edge Computing

Mobile Edge Computing(MEC)technology is considered as one of the key technologies of 5G networks.By deploying MEC servers at the edge of the network,exploiting the abundant computing resources,it can provide data processing services for different kinds of users,and improve the quality of service experience.The MEC server is generally deployed at the access point or base station.Affected by the factors such as hardware cost,the computing resources of the server is limited,so only a limited number of users can be served.In addition,due to the small coverage area of the server,the total benefits obtained from the MEC server are greatly affected by users' mobility.Therefore,how to design an effective task offloading and resource allocation strategy with a resource-constrained MEC server based on users' mobility to maximize the total system benefits is the main target of this thesis.Due to users' mobility and the limited coverage of the MEC server,before the task offloaded to the server is processed,the user may have already left the service scope of MEC.At this time,the task results cannot be sent directly to the user,and need to be transmitted to the target edge computing server through a macro base station or cloud server.Then transmitte it back to the user,the process of which is called task migration.The process of migration will cause additional overhead of time and resource,affecting the overall efficiency of the system.In view of the above problem,this thesis first studies the tasks offloading and migration issues of edge computing in the Internet of Vehicles.By analyzing the driving route of the vehicle,the driving time of the vehicle within the coverage of a single RSU can be deduced.By jointly optimizing the offloading and resource allocation strategies,task migration can be avoided and maximize the total system revenue.In this thesis,the original problem is decomposed into two sub-problems:the task offloading problem and the computational resource allocation problem.A Gini Coefficient based task offloading decision algorithm and a resource allocation algorithm based on Lagrange multiplier method are used to solve the problem.Simulation results show that the proposed scheme can well avoid task migration,reduce delay and energy consumption,and improve overall system revenue.Next,for the general users with mobile devices,a novel mobility-aware offloading and migration scheme with MEC is proposed to maximize the total revenue of MEs.The probabilistic model of users'sojourn time is used to describe the user's mobility.Combined with the probability density function of user's sojourn time,the problem of maximizing the total revenue of MEs is formulated,which integrates the offloading decision and resource allocation to reduce the probability of migration and ultimately to maximize the user's total revenue expectations.In order to solve this mixed integer nonlinear programming problem,this thesis uses genetic algorithm to obtain a suboptimal solution.Simulation results show that the proposed scheme can effectively reduce the task migration probability,avoid extra costs caused by migration,and thereby improve the overall revenue of users.