Traffic Management Based on Cloud and MEC Architecture with Evolutionary Approaches towards AI: A Review

Abstract

This review paper explores the significance of machine learning (ML), deep learning (DL), reinforcement learning (RL), and deep reinforcement learning (DRL) techniques in improving traffic management based on cloud and mobile edge computing (MEC) architectures. The key findings and contributions of this review highlight the potential of these techniques for transforming traffic management systems through data-driven decision-making, adaptive control, and optimization. The challenges identified in this field include data availability and quality, scalability and computational requirements, privacy and security concerns, and ethical considerations. In conclusion, ML, DL, RL, and DRL techniques, in conjunction with cloud and MEC architectures, have significant implications for improving traffic management. Their ability to process and analyse large-scale and real-time traffic data enables improved traffic flow, reduced congestion, enhanced energy efficiency, and enhanced overall transportation system performance.