Component-based Scheduling for Fog Computing

Abstract

Cloud computing have established the utility computing paradigm as a standard for application development and execution. As heterogeneity in applications requirements become a norm, fog computing has emerged recently to introduce computing capacity layers between the edge and the cloud, creating a hierarchy of computing power that can be used as a utility to run highly heterogeneous applications. However, in order to make this layered infrastructure a reality, new resource management mechanisms are necessary. In this paper we propose a component-based scheduler that considers application requirements heterogeneity as well as the fog-cloud computing hierarchy to improve applications execution in a cloud-fog computing infrastructure. The proposed algorithm takes into account the delay-priority of applications when taking scheduling decision on the fog-cloud infrastructure. We evaluate the proposal in a simulator, and preliminary results suggest the component-based scheduling algorithm is able to reduce average delays for applications with stricter requirements.