Resource Allocation And Tasks Scheduling In Mobile Edge Computing System

With the continuous development of communication technology and the prosperity of mobile Internet,the number of intelligent applications and smart mobile devices has made rapid progress.As applications become more and more intelligent,they are more and more sensitive to the demand of energy consumption and processing delay.However,the limited resources of mobile devices are increasingly difficult to meet these requirements,so Mobile Edge Computing(MEC)came into being.Mobile edge computing is a technology that sinks computing resources to the edge so that mobile devices can offload their computing tasks to the edge server for processing.As a technology that can assist mobile devices to deal with computationcentralized or delay-sensitive applications,compared to mobile cloud computing,mobile edge computing has the advantages of low latency,high bandwidth,mobility support,security privacy protection and personalized services,which effectively enhances the computing capacity of mobile devices and saves battery energy consumption.It is an important technology in 5G technology.In mobile edge computing system,the core problems need to be solved include: mobile device offloading strategy,communication and computing resource allocation and tasks processing scheduling.In this paper,the convex optimization theory and tasks scheduling theory are used to successfully put forward a method for multi-device multitask to realize moving edge computing in time-division duplex mode and frequencydivision duplex mode.The main research contents are as follows:1)This paper considers the three stages of the task simultaneously,namely,uploading,processing and downloading stages,as well as their time allocation and energy consumption.Combined with the two communication modes of TDD and FDD,the algorithm of how to allocate time slots or spectrum resources in different communication modes is given.2)This paper takes the limitation of parallel processing capability of edge servers into account,and presents an algorithm of execution sequence for tasks scheduling and the uploading and downloading execution sequence when edge servers only have limited parallel processing capability.3)This paper considers tasks offloading strategy of mobile devices,the uploading,processing and downloading stages of tasks and their execution sequence,time slots of uploading and downloading resources allocation in TDD mode,physical spectrum of uploading and downloading resources allocation in FDD mode,uploading and receiving power consumption of mobile devices,and the scheduling of computing resources synthetically.In addition,The resource allocation and task scheduling algorithms in TDD and FDD communication modes are presented.