Research On Joint Task Offloading And Resource Allocation Algorithm For Mobile Edge Computing Systems

The rapid development of Internet and smart devices promotes the emerging of various new applications,i.e.,augmented reality,virtual reality and natural language processing.However,the resource-intensive feature of those applications poses great challenges to the computation capability of smart devices.Mobile edge computing(MEC)was proposed to address this issue.Through deploying high performance MEC servers at radio access networks,MEC allows users to offload their tasks to the MEC servers,which then conduct task execution for the users,hence,the task execution latency and energy consumption of smart devices can be reduced efficiently,the quality of service of users could be improved significantly.It is of great significance to design efficient task offloading and resource allocation strategies for MEC systems considering the characteristics of tasks and the available resources jointly.In this thesis,the joint task offloading and resource allocation algorithms for MEC systems are investigated,and the main contents are as follows:The concepts and architecture of MEC are introduced.The related technologies and major senarios of MEC are then described,and the existing task offloading and resource allocation algorithms for MEC systems are summaried and analyzed.A cost optimization based joint task offloading and resource allocation algorithm is proposed for device-to-device(D2D)enabled cellular MEC systems.By defining the task execution cost as the weighted sum of task execution latency and energy consumption and considering constraints including task offloading,task partition,transmission rate and resource allocation,the joint optimization problem is formulated as a task execution cost minimization problem.As the formulated optimization problem is a mixed integer nonlinear programming(MINLP)problem,which can not be solved conveniently,we proposed a heuristic algorithm and transform the original optimization problem into two subproblems,i.e.,task offloading subproblem and resource allocation subproblem.By solving the two subproblems by means of Kuhn-Munkres(K-M)algorithm and Lagranian dual method,respectively,the joint optimization strategies could be obtained.An energy consumption optimization based joint task offloading and resource allocation algorithm is proposed for dense-networking cellular MEC systems.Through defining the task execution energy consumption as the maximal task execution energy consumption of the system-wide users and considering constraints including task offloading,power allocation,transmission rate and computation resource allocation,the joint optimization problem is formulated as a maximal task execution energy consumption minimization problem.As the formulated optimization problem is a MINLP problem,which cannot be solved directely and conveniently,we proposed a heuristic algorithm and transform the original optimization problem into two subproblems,i.e.,power allocation subproblem,task offloading and computation capability allocation subproblem.By solving the two subproblems through fractional programming method,Lagranian dual method,variable relaxation and substitution,and upper bound substitution method,respectively,the joint optimization strategies could be obtained.