Energy-saving Strategy For Mobile Edge Computing By Collaborative Processing Task Strategy On Base Station Group

In the Mobile Edge Computing(MEC)system,the Smart Mobile Device(SMD)can offload the task to the edge base stations for reducing energy consumption and processing time of the tasks.The existing researches on MEC mainly focus on the energy consumption and task processing delay of smart mobile devices,and there are few researches on cooperative processing tasks among base stations.In the existing researches on base station cooperative processing task,the centralized model is mainly used for task assignment,but the data interaction cost of centralized processing is neglected in the model,and the centralized task assignment is not suitable for high realtime mobile edge computing system.In this thesis,we study the task offloading strategy and computing resource allocation scheme among base stations to minimize the total energy consumption of base stations.The distributed task offloading strategy and computing resource allocation scheme without global information are given.In addition we analyze the disadvantage of the offloading strategy without global information.A Dynamic Voltage Scaling technique is introduced for the scenario of the base station task queues with low load.We establish the energy consumption model of the base station group cooperative processing task.Furthermore,we respectively design the task offloading strategy and the computing resource allocation scheme.The threshold of emergency task is given for determining whether to offload task.In order to minimize the energy consumption of the base station group,we use the optimization method to design the task offloading strategy,and design the computing source allocation scheme with the dynamic planning.Through the experiments,it is shown that the base stations can obtain 30%~40% energy savings by using Task-offloading Decision Algorithm(TDA)and Task Fragments Handle Algorithm(TFHA).For the case of base station task queues with high load,we introduce the Dynamic Voltage Scaling technology,and establish the energy consumption model of the corresponding base station group cooperative processing task.The task offloading strategy and the computing resource allocation scheme are jointly considered.We propose a new offloading judgment condition which has the larger scope of application.Through the analysis of the THFA,the energy consumption model of the base station group cooperative processing task is divided into two types.We establish two nonlinear problems with the goal to minimize energy consumption of the base station.We design the Load Balancing Algorithm(LBA)and Breakpoint Searching Algorithm(BSA)to solve the problems.Experiments show that in a 10? 10 cellular network with high load.It can reduce 15% energy consumption of the entire network by using LBA and BSA.At the same time,the offloading conflicting rate of the unpredicted real-time task offloading strategy under different load conditions of the base station network is demonstrated.