Research On Scheduling And Storage Strategy For Terminal Upstream Data Flows In Edge Computing

Since the development of computer network technology,new types of network interconnection technologies have been produced continuously.Network interconnection is no longer part of the interconnection of a single electronic computer.It has become a scene where everything in the world can connect.With the advance of network terminal equipment,the network access volume of terminal equipment is huge,the functions become more abundant,and the data upload volume of the terminal is increasing day by day.Although centralized cloud computing can solve intensive computing tasks,in the application scenario of the Internet of Everything,problems such as high latency,high bandwidth,and network congestion are needed to occur.Under the additional demand,edge computing is proposed.Edge computing places computing resources in a distributed manner at the edge of the network close to the data source,and the data can be handled locally or locally.Edge computing can keep up with the shortcomings of cloud computing,improve computing agility,and reduce the pressure of network broadband transmission.Edge computing offers the opportunity to solve the limited resource of terminal equipment,improve the network computing processing environment,and improve user service quality.Although the idea of edge computing is conducive to coping with the current status of the Internet of Things,it is associated with the problem of resource management in the new network structure.The edge server must reasonably schedule and allocate its resources when providing computing or storage services for terminal devices.Resulting from the extreme density of terminals,complex functions,and strong heterogeneity in edge computing,the scheduling,and storage of upstream data flows pose additional challenges to the business logic of edge servers.In this paper,we study the scheduling and storage strategies of the upstream data flow of the edge server in the three-layer edge computing model of the cloud,edge,and end.The main contents of this study are:(1)The business logic structure of upstream data flow processing in the edge server is proposed.In the network architecture,edge computingis between the cloud computing service and the terminal equipment and has the role of inheritance.Combining the attributes of edge computing and the network environment,the main business logic of the edge server for upstream data flow is designed.(2)The organizational structure model of upstream data flow scheduling in the edge server is proposed.There is a lot of data interaction between the network terminal layer and the edge server,and the amount of data in the network is huge.The edge server uses the organizational model of data flow scheduling to manage the upstream data flow formed in the network to achieve the unified management of terminal data.(3)The time slice self-adaptive scheduling(TSAS)algorithm for upstream data flow management in the edge server is proposed.The algorithm is used to schedule and control a large number of data flows that reach the edge server node to prevent congestion problems at the data flow reception.Experiments show that the TSAS algorithm can achieve efficient and reasonable data flow scheduling processing.(4)The hierarchical storage strategy for side-end cooperative data flow is proposed.The local processor's demand for data flows directly affects the storage strategy of the edge server.The designed data flow hierarchical storage algorithm can perform hierarchical storage management on data flows.Both the edge and the terminal use this data storage strategy,which not only makes full use of the storage resources but also improves the search time of the data.