Privacy-preserving Machine Learning in Internet of Vehicle Applications: Fundamentals, Recent Advances, and Future Direction
This review systematically examines how Privacy-Preserving Machine Learning (PPML) techniques can protect sensitive vehicle data in Internet of Vehicle (IoV) systems. It analyzes foundational methods like Federated Learning and Homomorphic Encryption, detailing their application across traffic management, autonomous driving, and communication domains. The study concludes that future success requires hybrid frameworks to balance privacy guarantees with the real-time performance demands of vehicles.