Computing And Communication Resources Scheduling Of Three-tier Processing Architecture In Mobile Edge Network

With the development of mobile networks and the popularity of mobile smart devices(smartphones,tablet computers,etc.),mobile applications(interactive games,virtual reality,natural language processing,etc.)have been experiencing explosive growth.The limited computing capacity and battery capacity of mobile devices have been unable to meet the computing needs of various mobile applications.How to solve the conflict between the limited computing and storage resources of mobile devices and the huge resource requirements of mobile applications is facing great challenges.Mobile Edge Computing as an effective technical approach to deal with such a challenge can greatly improve the computing power of mobile devices and reduce computing energy consumption.In the mobile edge network system,the user offloads the computing task to the edge server through the wireless cellular network,alleviating the computation load of the mobile device and utilizing the characteristics of the mobile edge network close to the terminal user to provide the terminal user with the computing capability nearby.Therefore,the computation offloading problem in mobile edge networks has been studied.However,most of the existing related researches are about the task two-tier processing architecture(local & SBS,or local & MBS),and rarely consider the task three-tier processing architecture(local & SBS & MBS).Compared with two-tier processing architecture,the three-tier processing architecture is more advantageous in terms of delay and computing power since its MEC server is arranged at both the small base station and the macro base station.Meanwhile,the problem of computation offloading decisions and resource allocation in the task three-tier processing architecture is also more complicated.Therefore,this paper mainly studies the problem of computation offloading and resource allocation of the task three-tier architecture in the mobile edge network.The main tasks are as follows:Firstly,taking the system utility as the optimization goal,the optimization problem of maximizing system utility of the task three-tier architecture in the mobile edge network is proposed.In order to solve the problem,this paper first presents an average resource allocation scheme to achieve the decoupling between task offloading and resource allocation by allocating communications and computing resources to users in advance,and then make task offloading decisions.Specifically,pre-allocating the channel and SBS's computing resources to users,thereby the optimization problem was simplified into a 0-1 integer linear programming problem for only solving the offloading variables,and then solving the problem.Finally,the simulation compares the task three-tier processing architecture(local & SBS & MBS)used in this paper with two other types of two-tier processing architecture(including local & SBS and local & MBS)with the changes of power consumption weights,task delay constraints,user's number on the system utility and system power consumption.The simulation results show that the performance of the task three-tier processing architecture is superior to the two-tier processing architecture in both system metrics,which also reflects the advantages of the three-tier processing architecture used in this paper.Secondly,there is a serious waste of system resources for the average resource allocation scheme.This paper further proposes a resource allocation scheme based on task offloading priority of the task three-tier architecture in the mobile edge network.The scheme first ensures that tasks meeting latency constraints and with minimumpower consumption are processed locally,and designs a task offloading priority function that takes into consideration multiple factors such as the amount of data required to process the task,task delay constraints,computing power requirements,and channel conditions,etc..For non-locally processed tasks,task offloading priority function is used to make task offloading decisions and resource allocation.Finally,the simulation also compares the results of the resource allocation scheme based on task offloading priority and the average resource allocation scheme on the system utility and system power consumption with the changes of power consumption weights,task delay constraints,and user's number.The simulation results show that the proposed resource allocation scheme based on task offloading priority has higher system utility and lower system power consumption than the average resource allocation scheme,and is more suitable for multi-task offloading system.