Researches On Integrated Management And Control For Edge Computing And Communication

Mobile edge computing(MEC),which deploys powerful computing resources at the radio network edge nodes(such as small base stations,macro base stations,wireless access nodes,wireless network controllers,and so on)close to the mobile terminals,aims to achieve immersive edge cloud service environment with ultra low latency and low energy consumption.Under MEC,the limited mobile terminal can transmit its complex application request to the edge cloud and then receive the application results(i.e.,application offloading),which helps to extend mobile terminal’s computing capacity as well as its battery life.As one of the key technologies of 5G,MEC gains much attentions from the academia and industry.And the newly emerging Internet of Things(Io T)applications(such as high-definition video,augmented reality,virtual reality and so on)are blooming based on the rapid development of the MEC,which promotes to a great developing in areas such as the super transportation,smart home,holographic radio and so on.Consequently,MEC has great potential to support the comprehensive development for the economy and society.However,due to the factors such as the size of the physical space,the network deployment cost and the network maintenance cost,the edge computing and the communication resources in MEC are still limited,differentiated and dynamic,which may fail to achieve high-quality response service for the newly emerging Io T applications with diverse,latencysensitive,energy-hungry and computing-intensive features.In order to provide ultra lowlatency and low energy consumption response for the diversified Io T applications anytime and anywhere,this dissertation focuses on the developing of the integrated management and control technology for the edge computing and communication,which helps to design reasonable application response scheme for the mobile terminals.Actually,an available edge resource control and management integrated scheme should not only meet the differentiated needs for the diversified Io T applications,but also ensure fair and reasonable scheduling for the differentiated resources,which would lead to a great challenge in the development of MEC.In this dissertation,excavating the competing and sharing relationship between the mobile terminals for limited computing and communication resources,exploring the complex”supply-demand” relationship of computing between MEC servers and differentiated terminals,and weighing the MEC ”computing utility-communication utility-user experience”are fully studied.And the development of the adaptive edge computing and communication joint management technology for the diversified Io T applications are studied to alleviate the impact of ”resource bottlenecks” on high-quality network service.The goal of this dissertation is to achieve green and quick application response on the mobile terminal,as well as to improve computing resource utilization.This dissertation is organized as follows:1.By fully exploring the complex competing and sharing relationship among the network mobile terminals for the limited edge computing and communication resources,an online resource coordinating and allocating scheme was proposed,which helps to effectively solve the real-time contradiction between the random application requests and time-varying network resources.Considering random application requests,uncertain terminal processing status,time-varying edge computing and communication status more comprehensively,a personalized and adjustable application offloading model based on the updating system data and service queues is designed,which aims to minimize the application response delay and energy consumption simultaneously.Then,based on the Lyapunov theory,variable substitution technique as well as the resource provision priority mechanism,an online resource coordinating and allocating scheme is designed for solving the non-convex and non-smooth target offloading problem with low complexity.Finally,the theoretical and simulation results show that our online resource coordinating and allocating scheme can adaptively adjust the real-time offloading scheme for the mobile terminal to meet the personalized service requirements and ensure high-quality application response with low latency and low energy consumption.2.By fully exploring the ”supply-demand” relationship of the computing resources between MEC servers and the differentiated application requests,a fair service enabled pricing and allocating algorithm(FS PAA)is proposed in this dissertation,which helps to maximize the earnings of edge cloud servers for providing computing capacity and the utility of the mobile terminal by making offloading for energy saving simultaneously,which also helps to improve the utilization of computing.Firstly,a local CPU self-adjusting energy consumption model embedded energy saving utility model of application offloading is proposed under the constraints of the maximum application latency.Then,based on Stackelberg game theory,convex optimization and classification discussion methods,FS PAA is proposed.Note that,the fair service enabled pricing and allocating algorithm reveals the explicit mathematics relationship between the optimal terminal’s offloading decision and the edge computing pricing,as well as the explicit relationship between the optimal computing pricing and maximum computing resource provisioning,which helps to provide an effective reference for the actual edge computing resource deployment and configuration.Simulation results show that our FS PAA can ensure fair offloading services for the different application service requests,and also ensure more users to obtain green offloading services.And the utilization of MEC computing resources under FS PAA can be improved greatly.3.By reasonably weighing the computing utility,communication utility,and the user experience under MEC,an integrated coordinating and scheduling scheme for edge computing and communication is proposed to reduce the response delay of the diversified Io T applications,balance the heterogeneous user experience,and also expand network service capabilities(i.e.,the number of the served users).By fully considering the diversified application requests(including different types of application request,personalized user requirements,etc),limited multi-dimensional cloud resources,limited backhaul capacity and the antenna deployment on the small base station,a MEC composite utility maximization model is designed,in which our designed punishment dominated incentive mechanism for quick application response is embedded.Based on Lagrange dual theory and greedy configuring idea,a lowcomplexity multi-dimensional edge computing and communication resources coordinating and scheduling scheme is proposed for efficiently solving the mixed integer nonlinear composite utility maximization problem.Simulation results show that the edge computing and communication integrated scheduling scheme can achieve quick response for practical diversified Io T applications,and also expand the system service capabilities for more users to obtain satisfied service.