Research On Energy-saving And High-efficiency Joint Optimization Of Resources In Mobile Edge Computing System

With the rapid development of the Internet of things technology and 5G mobile communication technology,a variety of mobile services and applications have become a part of people's daily life,such as online games,live network,virtual reality,augmented reality and so on.Although these emerging mobile services and applications greatly enrich people's lives,they also occupy the huge computing,storage,network and battery resources of intelligent mobile devices.Mobile edge computing emerges as an efficient solution,which offloads computing tasks to the edge server,uses the powerful computing power of the edge server to expand the resources of intelligent mobile devices,thus alleviats the problems,i.e.high energy consumption and low computation frequency,caused by the lack of resources of intelligent mobile devices.As a new computing mode after cloud computing,mobile edge computing can sink the computing capacity of cloud center to the edge of network,and intelligent mobile devices can interact with edge servers in a short distance,so as to meet the demand of low latency and low power consumption for mobile services and applications.With the rapid development of Internet of Things,5G,Artificial Intelligence,Big Data and other fields,mobile edge computing will show more and more important value and become an indispensable support technology in the field of wireless communication.Considering the lack of energy efficiency of intelligent mobile devices in mobile edge computing,there are many researches on joint optimization computing and communication resource allocation on the purpose of reducing energy consumption.However,with the combination of mobile applications such as online games and virtual reality,and technologies such as Artificial Intelligence and Big Data,the battery of intelligent mobile devices can no longer meet the needs of such mobile applications with complex data processing functions.Meanwhile,the battery's life will directly affect the user's experience of mobile services.Therefore,based on the premise of ensuring the user service experience,this paper studies the goal of minimizing the energy consumption which includes the following contents: how to realize the energysaving and efficient resource optimization scheme for the task dependent mobile application in the mobile edge computing;how to realize the joint optimization mechanism of efficient double-layer computing and communication resources in the mobile edge computing;how to implement energy-saving and efficient relay selection strategy and resource allocation strategy in D2D-enabled mobile edge computing.The specific contributions of this paper are as follows:Firstly,aiming at the research of computing offloading of mobile applications composed of multiple functional modules or tasks with dependency in mobile edge computing environment,this paper proposes a model based on DAG task dependency graph,and considers the impact of the returned results of partially offloaded tasks on the offloading strategy.In order to maximize the utilization of resources and minimize the energy consumption,we establish a problem of minimizing the energy consumption of mobile devices and the optimization problem of joint optimization of offloading ratio,transmission power and CPU frequency is formulated.The formulation of optimization problem combined with convex optimization technology is transformed into the problem of constrained nonlinear equations,and an optimal solution algorithm based on binary search algorithm is proposed.The simulation results show that the proposed offloading strategy significantly reduces the energy consumption.Secondly,aiming at the problem that the communication energy consumption is too large in the long-distance task offloading of low-performance devices such as intelligent wearable devices,a mobile edge computing system combined with device collaboration is proposed.The low-performance device offloads part of the computing tasks to high-performance devices as corporation devices such as smart phones.After receiving the offloading data,the high-performance device processes part of the data,and the rest is transmitted to nearby edge server through wireless network,forming a two-tier computing unloading system.In order to minimize the overall energy consumption of mobile devices,an iterative optimization algorithm based on the block coordinate descent method is proposed to jointly optimize the computing and communication resources of collaboration nodes,edge servers and intelligent wearable devices.Simulation results show that the proposed iterative optimization algorithm significantly reduces the energy consumption of mobile devices,i.e.10%,and the execution time of the algorithm is relatively little.Finally,aiming at the problem of long-distance task offloading in mobile edge computing,this chapter proposes an optimization problem to minimize the energy consumption of mobile devices by jointly optimizing the relay selection strategy and resource allocation strategy in D2D-enabled mobile edge computing system under the condition of satisfying the constraints of computing,communication resource and delay,which is fomulated as an integer-mixed nonconvex optimization problem,and a twostage optimization algorithm is proposed.By using convex optimization techniques,such as discrete variable relaxation and linearization,the original problem is transformed into a convex optimization problem.In the first stage of the algorithm,Lagrange multiplier method is used to solve the problem,and the optimal relay selection strategy is obtained.In the second stage of the algorithm,the optimal resource allocation strategy is obtained by combining convex optimization technology and relay selection strategy.Simulation results show that the proposed two-stage joint optimization algorithm has less energy consumption,i.e.10%-20%,and better performance while ensuring the quality of service.In conclusion,in view of the above three research problems,this paper puts forward the corresponding mobile edge computing framework for the research problem,the communication,calculation and other system models involved in the formalization of the problem are described and then the derivation and simplification in the problem solving process are listed in detail,finally,simulation experiments are given to verify the correctness and effectiveness of the proposed algorithm.