Resource and History-Aware IoT Task Scheduling in Volunteer Assisted Fog Computing

Abstract

Internet of Things (IoT) has revolutionized the way in which we collect, process, and utilize data of different systems. To make the best use of this technology and handle emergency situations in a timely manner, the data collected by the IoT devices have to be processed in real-time sometimes. Owing to limited capabilities of IoT devices, fog computing, which is a distributed and scalable computing paradigm, has evolved to facilitate computing, storage and communications resources at the edge of the network. Since fog nodes have limited resources, with the increase in the number of delay sensitive IoT applications, some of the delay sensitive processing tasks have to forwarded to the cloud, which involves a huge delay. To address this issue, it has been proposed to combine volunteer computing with fog computing, where unutilized computing resources of the devices in the coverage areas of fog nodes are utilized to execute delay sensitive tasks. This paper addresses task scheduling in volunteer assisted fog computing. A volunteer selection metric that forwards the delay sensitive tasks to the most efficient volunteer is developed. And a task scheduling method capable of supporting delay sensitive IoT applications in the presence of mobile volunteers is also developed. The performance evaluation shows that the proposed method results in a lower delay and lower network usage compared to the existing literature.