Privacy-preserving Machine Learning in Internet of Vehicle Applications: Fundamentals, Recent Advances, and Future Direction

2025-03-03

AI Summary

To address escalating privacy risks arising from widespread machine learning (ML) deployment in the Internet of Vehicles (IoV), this paper establishes a comprehensive privacy-preserving machine learning (PPML) theoretical framework and technical system tailored for IoV. We propose a four-layer IoV architecture and a three-domain application taxonomy, and—first in the field—introduce a systematic PPML technique classification alongside a three-dimensional adaptability evaluation framework balancing privacy, utility, and efficiency. Integrating secure multi-party computation, homomorphic encryption, differential privacy, and federated learning, we validate our approach through modeling in key IoV scenarios: cooperative perception, edge resource scheduling, and V2X services. The work yields a full-ML-lifecycle PPML technology roadmap, identifies six cross-cutting challenges, and proposes five actionable research directions—providing both theoretical foundations and practical pathways for standardization and engineering implementation of privacy protection in IoV.

Technology Category

• Machine Learning: Privacy
• Humans and AI: Human-in-the-loop Machine Learning
• Computer Vision: Multi-modal Vision

Application Category

• Security and Privacy: Security and privacy of machine learning and AI applications
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems

Abstract

Machine learning (ML) has revolutionized Internet of Vehicles (IoV) applications by enhancing intelligent transportation, autonomous driving capabilities, and various connected services within a large, heterogeneous network. However, the increased connectivity and massive data exchange for ML applications introduce significant privacy challenges. Privacy-preserving machine learning (PPML) offers potential solutions to address these challenges by preserving privacy at various stages of the ML pipeline. Despite the rapid development of ML-based IoV applications and the growing data privacy concerns, there are limited comprehensive studies on the adoption of PPML within this domain. Therefore, this study provides a comprehensive review of the fundamentals, recent advancements, and the challenges of integrating PPML into IoV applications. To conduct an extensive study, we first review existing surveys of various PPML techniques and their integration into IoV across different scopes. We then discuss the fundamentals of IoV and propose a four-layer IoV architecture. Additionally, we categorize IoV applications into three key domains and analyze the privacy challenges in leveraging ML for these application domains. Next, we provide an overview of various PPML techniques, highlighting their applicability and performance to address the privacy challenges. Building on these fundamentals, we thoroughly review recent advancements in integrating various PPML techniques within IoV applications, discussing their frameworks, key features, and performance evaluation in terms of privacy, utility, and efficiency. Finally, we identify current challenges and propose future research directions to enhance privacy and reliability in IoV applications.

Problem

• Addressing privacy challenges in ML for IoV applications.
• Reviewing PPML techniques for privacy in IoV systems.
• Proposing future research to enhance IoV privacy and reliability.

Innovation

• Four-layer IoV architecture proposed
• PPML techniques categorized and analyzed
• Recent PPML advancements in IoV reviewed