Usage-Oriented Resource Allocation Strategy In Edge Computing Environments

Edge-computing using distributed computing architecture has solved the problem of massive data transmitting. By letting multiple nodes being computed locally and upload afterward to clouds, data processing gained great progress efficiently. Edge-computing involved virtualizing techniques, but due to the heaviness of virtualization, people prefer another lighter technique, the Container. Furthermore, while any node is installed with dockers, the node can normally execute containerized applications. This article simulates the edge-computing environment through Kubernetes. By adding a new predicate strategy to ensure the efficiency of job allocating, two built-in algorithms in the Kubernetes are compared with the proposed approach. Experimental results show the performance improvement after adopting the proposed approach.