Artificial Intelligence Techniques for Strategy Design for Energy-Efficient Data Offloading in 6G-Enabled Vehicular Edge Computing Networks Using Double Deep Q-Network: Trends and Challenges

Abstract

The rapid advancement of intelligent transportation systems and the emergence of sixth-generation (6G) networks have intensified the demand for efficient computation in vehicular environments. Vehicular Edge Computing (VEC) has become a promising paradigm to address latency-sensitive and computation-intensive applications by enabling task offloading from vehicles to nearby edge servers. However, achieving energy-efficient data offloading in dynamic vehicular networks remains a significant challenge due to high mobility, varying channel conditions, and limited computational resources. Artificial Intelligence (AI), particularly Deep Reinforcement Learning (DRL), has emerged as an effective solution for optimizing offloading decisions in such complex environments. Among DRL techniques, the Double Deep Q-Network (DDQN) has gained attention for its ability to reduce overestimation bias and improve learning stability. Recent studies demonstrate that DDQN-based strategies outperform traditional optimization approaches in minimizing energy consumption and latency while adapting to dynamic network conditions.  This paper presents a comprehensive review of AI-driven strategies for energy-efficient data offloading in 6G-enabled vehicular edge computing networks. It analyses recent advancements (2020–2023), highlights key trends, and identifies critical challenges. Furthermore, it explores hybrid AI techniques and multi-agent frameworks for improving scalability and efficiency. Finally, the paper outlines future research directions toward developing intelligent, adaptive, and secure offloading mechanisms in next-generation vehicular networks.