Task Classification and Scheduling Based on K-Means Clustering for Edge Computing

• Published in: Wireless personal communications Vol. 113; no. 4; pp. 2611 - 2624
• Main Authors: Ullah, Ihsan, Youn, Hee Yong
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.08.2020; Springer Nature B.V
• Subjects: Algorithms; Classification; Cloud computing; Cluster analysis; Clustering; Communications Engineering; Computer Communication Networks; Computer simulation; Edge computing; Engineering; Heterogeneity; Internet of Things; Networks; Nodes; Performance enhancement; Queues; Queuing theory; Resource utilization; Schedules; Signal,Image and Speech Processing; Task scheduling; Vector quantization; Task scheduling; K-means algorithm; Internet of Things; Task classification; Edge computing
• ISSN: 0929-6212; 1572-834X
• DOI: 10.1007/s11277-020-07343-w

The rapid evolution of Internet of Things and cloud computing have endorsed a novel computing paradigm called edge computing. Here tasks are processed by edge devices before sent to the cloud to reduce the computational latency and overhead of cloud server. In edge computing efficient classification and distribution of the tasks among the constituent nodes is a challenging issue because of their resource limitedness and heterogeneity. In this paper a novel scheme named KTCS (K-means Clustering-based Task Classification and Scheduling) is proposed which classifies the task based on the type of resource requirement in terms of CPU, I/O, or COMM before distributed to the edge node. Using the K-means algorithm modeled with the M / M / c queuing theory, the proposed scheme efficiently schedules and assigns the task so that the utilization of the edge devices can be increased. The simulation result reveals that the proposed scheme significantly improves the performance of edge nodes in terms of task execution time and resource utilization.