Systematic Analysis and Taxonomy of AI/ML Based Resource Management in Fog and Edge Computing

Abstract

Fog and edge computing have emerged as pivotal paradigms to address latency, bandwidth, and privacy challenges associated with cloud-centric architectures. These decentralized computing models bring computation closer to data sources, enabling real-time and context-aware services. Effective resource management in fog and edge environments is critical to optimize performance, reduce energy consumption, and ensure Quality of Service (QoS). Recent advances in Artificial Intelligence (AI) and Machine Learning (ML) have demonstrated significant potential in enhancing resource allocation, task scheduling, load balancing, and fault tolerance in these distributed systems. This paper presents a systematic analysis and taxonomy of AI/ML-based resource management techniques in fog and edge computing environments. We survey the state-of-the-art research published before 2019, categorizing approaches based on learning algorithms, resource types, and application domains. Our taxonomy highlights the use of supervised, unsupervised, and reinforcement learning techniques applied to various resource management challenges. We analyze their methodologies, datasets, performance metrics, and deployment scenarios. Furthermore, this work identifies current research gaps and challenges, including scalability, heterogeneity handling, security concerns, and real-time adaptability. The paper aims to guide researchers and practitioners in understanding the landscape of AI/ML-driven resource management and foster the development of robust, efficient, and intelligent solutions tailored for fog and edge computing platforms.