Research on Key Technology of Edge-Node Resource Scheduling Based on Linear Programming

Abstract

With the arrival of the 5G era, centralized cloud computing is gradually shifting to edge computing. However, due to the characteristics of distributed multi-deployment of edge computing, it is difficult to ensure that all edge-node clusters have enough resources. Faced with the limited resources of edge-node cluster, how to carry out reasonable resource scheduling and provide users with high-quality services (high quality and low delay) has become the focus of current research. Aiming at this problem, in this paper, the edge-node resource scheduling based on linear programming is the key technology being studied. It is compared with conventional algorithm-based edge-node resource scheduling system. In the five deployments of the edge node resource scheduling system under the conventional algorithm, 8 virtual machines are used. The edge node resource scheduling system under the linear programming algorithm uses 7 virtual machines for 3 deployments and 8 virtual machines for 2 deployments. It can be seen that the results affirm the feasibility of edge node resource scheduling based on linear programming, and make due contributions to current research.