Edge computing: Architecture, Applications and Future Perspectives

TLDR

Fast advances in mobile internet and IoT create challenges for centralized cloud computing such as high latency, low spectral efficiency, and incompatible machine‑type communication. The article aims to resolve these issues by shifting cloud functions to edge devices and provides an overview of mobile edge computing, cloudlets, and fog computing. The authors describe the standardization, principles, architecture, and utilities of the three edge technologies, discuss fog computing radio access network features, and outline unresolved issues and future research directions. In radio access networks, mobile edge computing differs from fog computing.

Abstract

The fast advancements in the fields of mobile internet and the internet of things (IoT) have caused several serious challenges for the traditional centralized cloud computing like large latency, small spectral efficiency (SE), and incompatible machine type of communication. Aimed at resolving the mentioned issues, several innovative technologies have been developed to shift the functions of the centralized cloud computing to the edge device of the network. Various edge computing techniques based on diverse origins have been established to decline the latency while improving SE, and supporting the massive machine-type communications. The present article offers an overview on three edge computing technologies: mobile edge computing, cloudlets, and fog computing. Specifically, standardizing procedures, principle, architecture, and utility of the mentioned technologies will be addressed. In terms of radio access network, the mobile edge computing difference from the fog computing was described. Features of fog computing radio access networks will be addressed as well. In the end, unsolved issues and future research topics will be discussed.