REVIEW OF EDGE COMPUTING AND RESOURCE MANAGEMENT: CHALLENGES AND STATE OF THE ART SOLUTIONS

Abstract

The Internet of Things with its discovery for linking billions of devices and static devices to aid with numerous applications in real time has make cloud computing paradigms encounter major challenges which includes jitter, extreme latency, non-encouraging location- cognizance and mobility. In addressing such challenges, edge computing paradigm and related architecture have been developed to move digital services from the central cloud to network edge. Observing its disruption, IoT connected devices are forecasted to reach 500 billion by the year 2030, even as the world mobile traffic is anticipated to be in the rise by 2022. Given the advantages of the edge computing paradigm, the edge devices are to some extend resource constrained were an efficient resource management is essential to make edge computing a reality. The study proposed a framework for efficient resource management at the network edge based on deep reinforcement learning. The experimental results reveals that the EECFRM framework based on the TD3 algorithm converges in comparison with the DDPG based algorithm as presented.