A Novel Federated Fog Architecture Embedding Intelligent Formation

• Post by: Ahmad Hammoud
• 26 June 2021
• Comments off

Abstract

Network delays cause disturbance and reduction in the Quality-of-Service (QoS) for Internet-of-Things (IoT) while end-users are running critical real-time services. In parallel, federated fogs are not effective when formed without considering the performance perceived by the end-users. This article presents a novel architecture for the federated fog concept and proposes an adaptive and intelligent federation formation approach using Genetic Algorithms and Machine Learning models. Fog federations serve as a solution for fog providers to offer the required QoS they serve. Such a concept allows efficient distribution of load among multiple fog providers that share their resources. Throughout this process, the issue of QoS deterioration, due to local overloads, is relatively solved. Hence, the end-users can enjoy a delay-free experience when using real-time applications. Real data is used to evaluate the proposed architecture and formation mechanism. The results show a notable improvement in the throughput as well as a decrease in the response time for the services requested.

Authors’ Notes

The contributions of this paper are:

• Devising a novel federated fog architecture that embeds all the participating entities and takes into account real-life parameters and constraints within a fog environment. To the best of our knowledge, this is the first attempt at addressing those aspects within a comprehensive fog federation architecture.
• Elaborating an adaptive federation formation process that is based on Genetic and Intelligent models. The proposed scheme achieves efficient results in terms of the number of satisfied users.
• Proposing a Machine Learning model for evaluating the fitness of the federations and dynamically adapting the Genetic Model for improving its results from one generation to another.