Enabling Safe Autonomous Vehicular Coordination through 6G, Internet of Vehicle and AI Synergy

Abstract

The transition toward fully autonomous vehicular ecosystems requires a fundamental rethinking of wireless networks, sensing, and distributed intelligence. This article surveys the foundational technologies, architectural elements, and performance requirements that will enable 6G-enabled Internet of Vehicles (IoV) for safe cooperative autonomy. We trace the evolution of mobile generations from 1G to 6G, highlighting why 5G falls short of sub-millisecond latency, extreme reliability, and integrated sensing-communication (ISAC) demanded by Level-4/5 autonomy. Key enabling technologies — terahertz spectrum, reconfigurable intelligent surfaces (RIS), edge-native AI, integrated sensing and communication, and blockchain-based trust mechanisms — are examined with emphasis on their role in reducing end-to-end latency and improving situational awareness. We discuss AI paradigms (onboard inference, federated learning, graph neural networks, and deep reinforcement learning) for perception, trajectory prediction, cooperative decision-making, and network resource allocation. Representative latency-critical use cases (collision avoidance, platooning, intersection orchestration, and cooperative perception) are analyzed to derive concrete latency, reliability, and compute placement requirements. Finally, the paper outlines open research challenges including scalable distributed learning, secure low-latency trust frameworks, energy-aware AI at the edge, and standardization gaps. This work provides researchers and system designers with an integrated roadmap linking 6G physical-layer innovations to AI methods and IoV architectures required to realize real-time, cooperative autonomous driving at city scale.