A Survey of Fog Computing-Based Resource Allocation Approaches: Overview, Classification, and Opportunity for Research

Abstract

    Recently, cloud computing has affected a large part of the computer industry, including software companies and internet service providers. It has proven efficient in managing tasks for applications. Despite its popularity, cloud computing does not meet the requirements of applications because it faces many limitations, such as high latency and bandwidth bottlenecks. These limitations will significantly affect applications sensitive to delays. To meet this challenge, fog computing is introduced as an extension to cloud computing. It improves quality of service (QoS) for applications that suffer from latency by keeping resources and services close to the end-user. How to efficiently and fairly allocate the available resources, e.g., CPU, bandwidth, and memory, between different requested tasks is a complex challenge. The main goal of this paper is to study the concepts of fog computing, architecture, environment, and metrics that affect resource allocation in fog computing. It also summarizes the classification of modern resource allocation approaches based on QoS metrics (2017–2023). On the other hand, highlighting the pros and cons of these studies as well as future research directions to develop different approaches.