A review on edge computing with data analysis and IoT techniques

Abstract

The Internet of Things (IoT) has permeated people’s daily lives, giving critical measurement and data collection capabilities to inform everyone’s choices. Edge computing has developed as a new paradigm for addressing IoT and localized computing requirements as a solution for mitigating resource congestion. These methods will produce huge volumes of important data at a network edge, necessitating not just immediate data processing but also smart data analysis to properly exploit the edge big data potential. Due to their limited calculation and high latency capability, both on-device computing and traditional cloud computing cannot adequately handle this issue. The purpose of this paper is to review recent efforts on machine learning and deep learning empowered edge computing applications and to provide insights into how to influence deep learning advances to ease edge applications from different domains, namely, smart city, smart industry, smart multimedia, and smart transportation. In a comparative analysis, cloud Machine Learning (ML) has the highest 98.04% accuracy and Convolutional Neural Network (CNN) ResNet-50 has 81.06% accuracy. Additionally, emphasize the critical research issues and promising research areas associated with them. This paper will inspire additional research and contributions to this promising area.