Machine learning-based solutions for resource management in fog computing

Fog computing is a paradigm that offers distributed and diverse resources at the network edge to fulfill the quality of service requirements. However, effectively managing these resources has become a significant challenge due to the dynamic nature of user demands and the distributed and heterogeneous characteristics of fog computing. Consequently, managing resources based on accurately predicting dynamic user demands and resource availability using machine-learning methods becomes demanding. In this study, we conduct a comprehensive analysis of existing literature that leverages machine learning-based approaches to address resource management challenges in fog computing. These challenges encompass resource provisioning, application placement, scheduling, resource allocation, task offloading, and load balancing. The examined literature is thoroughly compared based on their employed strategies, objective metrics, tools, datasets, and techniques. Furthermore, we identify research gaps in resource management issues and propose future directions for advancing the field.