Research On Data Caching And Task Scheduling In Edge Environment

Cloud computing effectively solves various disadvantages of traditional computing model,such as extremely limited computing power and storage capacity of terminal devices,and greatly promotes the development of big data and other related technologies.Compared with cloud computing,edge computing has many characteristics,such as low latency,low energy consumption,high bandwidth and so on,which has attracted the attention of a large number of researchers and developed a large number of research work in recent years,such as task scheduling,computing offloading,data caching,etc.The existing related research usually ignores the impact of data caching on task scheduling,and considers them separately,which is difficult to achieve the purpose of optimization.In addition,in the actual application scenario,there are usually dependencies between multiple tasks to be scheduled,and the execution order of some subtasks cannot be determined,which is usually ignored in the existing related work.In view of the above challenges,this work will study the joint optimization of data caching and task scheduling for the edge computing environment,and in view of the dependency and importance between tasks,the task scheduling will be discussed and studied.The main work of this paper includes the following parts:(1)Several factors affecting data caching and replacement are discussed and analyzed,namely,priority,replacement cost and data size.Then a calculation method of data caching value based on information entropy theory is given,and then a multi-indicator considering caching replacement algorithm is put forward.The experimental results show that the algorithm proposed can effectively improve the caching hit rate of data.This work provides support for the joint optimization of data caching and task scheduling.(2)The joint optimization of data caching and task scheduling in edge environment is studied.Based on the caching replacement method mentioned above,a hierarchical task scheduling method is proposed.This method schedules the current task request according to the priority of the nearest edge server,adjacent edge server and cloud server.In addition,for task requests scheduled to adjacent edge server or cloud server,the required data is cached to the edge server nearest to the terminal device in real time during the scheduling process,so as to schedule the task requests submitted by the terminal device to the nearest edge server for processing as much as possible in the process of subsequent task scheduling.Experimental results show that the method proposed can effectively reduce the average response delay and total energy consumption.(3)The task scheduling is discussed and studied for the dependency and importance between tasks.Firstly,a method to obtain the task scheduling sequence of DAG graph is proposed.Then,a new task scheduling method is proposed by the fusion of ant colony algorithm and simulated annealing algorithm.The experimental results show that the method proposed can effectively reduce the average response delay and total energy consumption of tasks.In addition,the influence of importance on task scheduling sequence is further discussed and analyzed in this paper.This paper verifies the effectiveness and efficiency of the proposed method through a large number of experiments.Compared with the existing related solutions,the proposed method can effectively improve the performance of data caching and task scheduling,and has good theoretical research significance and practical application value.