Fog-Based Computing Architecture for Enhancing Secondary Distribution Grid Services Management

Abstract

In monitoring the health of secondary distribution networks (SDNs), power utility providers have faced an increasing need to deploy intelligent solutions with affordable sensing and data-driven technologies. Existing manual-based approaches are not capable of collecting large volumes of real-time operational data to achieve significant monitoring of the SDN network for reliable power distribution. Effective monitoring would require real-time sensing, scalable high-performance computing, and appropriate grid-based applications designed for efficient data processing. This paper presents a computing architecture for grid services monitoring to enhance real-time fault management in SDN. The architecture leverages wireless sensor networks, a hybrid cloud-fog computing architecture, and a heuristic-based application coordination mechanism to efficiently manage grid applications. Experimental results indicate that coordination mechanism improved workload distribution by up to 70% in fog nodes and to 40% in the cloud. A fog-based architecture provided low latency improvements of 70% compared with that of cloud-only architectures. This signifies that most of the data processing was pushed to the local fog nodes, which is crucial for distributed fault management applications.