Role of Machine Learning in Resource Allocation of Fog Computing

Abstract

Fog computing is a fundamental facilitating technology in the future of networks. It broadens cloud computing services to its network edge, so that fog computing is able to bear the load of different emerging applications like IoT, blockchain, and big data without incurring high latency or cost of bandwidth consumption. In order to reach the optimal capacity of fog computing, designing an incentive mechanism for its service provider is essential. Over the recent years, we have noticed steady growth in the adoption of Machine Learning not only for the enhancement of fog computing applications but also for providing fog services, which include better resource management, improved security, reduced energy consumption, latency and traffic modelling. There hasn’t been a study yet, which investigates the role of Machine Learning in fog computing paradigm and this is what our current research aims to shed light on. Machine Learning application used for fog computing needs high layers services profound analytics, a strong end-user as well as smart responses for its designated tasks. Here, in this paper we work on presenting a comprehensive study that underlines the current advancements in Machine Learning techniques associated with the management of three important aspects of fog computing: accuracy, resource, security, and, as well as highlighting the role of Machine Learning in edge computing.