Research Article
A Dual Ring Architecture Using Controllers for Better Load Balancing in a Fog Computing Environment
@INPROCEEDINGS{10.1007/978-3-031-23116-2_11, author={Birane Koundoul and Youssou Kasse and Fatoumata Balde and Bamba Gueye}, title={A Dual Ring Architecture Using Controllers for Better Load Balancing in a Fog Computing Environment}, proceedings={Innovations and Interdisciplinary Solutions for Underserved Areas. 5th EAI International Conference, InterSol 2022, Abuja, Nigeria, March 23-24, 2022, Proceedings}, proceedings_a={INTERSOL}, year={2023}, month={2}, keywords={Fog computing Load balancing IoT Quality of service System performance}, doi={10.1007/978-3-031-23116-2_11} }- Birane Koundoul
Youssou Kasse
Fatoumata Balde
Bamba Gueye
Year: 2023
A Dual Ring Architecture Using Controllers for Better Load Balancing in a Fog Computing Environment
INTERSOL
Springer
DOI: 10.1007/978-3-031-23116-2_11
Abstract
Fog Computing is a paradigm that extends cloud computing by bringing network and cloud resources closer to the edge. This means that points of presence are placed close to end users for easy access and to enable delay-sensitive applications to have minimal response times. This Fog layer preprocesses data as close to the sensors as possible. However, with the increasing demand of IoT, even when close to sensors, fog nodes tend to be overloaded, compromising the response times of latency-restricted IoT (Internet of Things) applications, and therefore also compromising the quality of user experience. However, the limited storage and processing capacity leads us to ask the question: wouldn’t load balancing be an asset in the Fog Computing environment to avoid overwhelming some fog nodes? It is in this sense that we will exploit load balancing in the Fog Computing environment. We will propose an algorithm for selecting the fog node with the best resource (BRFC) to handle the request. A controller node is placed at each zone to manage data access and placement. This will prevent some nodes from being overloaded or underloaded in order to improve response time, system performance, throughput, cost and even energy consumption.
