Optimizing Edge Networks For Smart Classroom:Node Placement And Task Scheduling

The smart classroom is expected to improve teaching effectiveness and teaching quality from many dimensions,and has broad application space in future teaching activities.With further integration of the computer information technology,artificial intelligence technology,the Internet of things technology and education,the future of smart classroom will further change the teaching process.This paper first looks into the development of future smart classrooms: introducing a large number of sensors,smart cameras,wearable devices,and VR/AR devices to collect large data produced in teaching activities,and to provide teaching evaluation for the teaching process through large data processing and analysis based on artificial intelligence / deep learning algorithm,real time teaching feedback,teaching materials recommendation and other intelligent guidance services.In view of the development trend of the smart classroom,the current smart classroom management system is relatively independent of the sharing of resources,and the computing ability of a single smart classroom is not sufficient to support the intelligent guidance service based on artificial intelligence / deep learning.This paper focuses on solving the problem of insufficient computing resources in smart classroom and the scheduling of dynamic computing resources in smart classroom.This paper first analyzes the characteristics of the application of the intelligent classroom,and puts forward a data processing model based on edge computing,and constructs software defined management platform to manage the resources such as the edge computing and the network bandwidth required by the smart classroom.This paper also presents an edge computing selection model,which quantitatively analyzes the applicable scene of the data processing model based on the edge computing,and provides theoretical support for the optimization of the proposed edge network management platform for the smart classroom.In order to solve the problem of insufficient computing resources in the edge network,an edge node placement model is proposed to guide the deployment of edge nodes in the smart classroom scene,and the simulation experiment is constructed based on the distribution of the teaching buildings and the campus network topology in a Chinese famous university,and the performance of the deployment scheme of the edge nodes in the mobile edge computing scene is performed.In this paper,an improved BFDA algorithm and a task scheduling algorithm based on linear programming model are proposed in this paper.According to the data variation of data source nodes,the computing resources of the edge network task scheduling are carried out and simulation experiment based on the campus network topology is constructed.Compared to the traditional BFDA algorithm,significant improvement in performance is achieved.