SecureV2X: An Efficient and Privacy-Preserving System for Vehicle-to-Everything (V2X) Applications

2025-08-26

AI Summary

To address privacy leakage risks—particularly of location and sensitive biomedical data (e.g., EEG)—arising from machine learning deployments in V2X systems, this paper proposes the first efficient multi-agent secure neural network inference framework tailored for vehicular networks. The framework integrates secure multi-party computation (MPC), lightweight encryption, and distributed model partitioning to enable privacy-preserving deep learning inference across vehicle–edge–cloud collaboration. Its core innovation lies in drastically reducing communication rounds and computational latency while supporting large-scale concurrent private inference. Experimental evaluation demonstrates substantial improvements: on driver fatigue detection, inference throughput increases 9.4×, communication rounds decrease 143×, and total communication volume drops 16.6×; on red-light violation detection, it achieves nearly 100× speedup over state-of-the-art baselines—without compromising security or practical deployability.

Technology Category

• Machine Learning: Hardware-aware ML
• Computer Vision: Multi-modal Vision
• Multiagent Systems: Adversarial Agents

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Security and Privacy: Security and privacy of machine learning and AI applications
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Autonomous driving and V2X technologies have developed rapidly in the past decade, leading to improved safety and efficiency in modern transportation. These systems interact with extensive networks of vehicles, roadside infrastructure, and cloud resources to support their machine learning capabilities. However, the widespread use of machine learning in V2X systems raises issues over the privacy of the data involved. This is particularly concerning for smart-transit and driver safety applications which can implicitly reveal user locations or explicitly disclose medical data such as EEG signals. To resolve these issues, we propose SecureV2X, a scalable, multi-agent system for secure neural network inferences deployed between the server and each vehicle. Under this setting, we study two multi-agent V2X applications: secure drowsiness detection, and secure red-light violation detection. Our system achieves strong performance relative to baselines, and scales efficiently to support a large number of secure computation interactions simultaneously. For instance, SecureV2X is $9.4 imes$ faster, requires $143 imes$ fewer computational rounds, and involves $16.6 imes$ less communication on drowsiness detection compared to other secure systems. Moreover, it achieves a runtime nearly $100 imes$ faster than state-of-the-art benchmarks in object detection tasks for red light violation detection.

Problem

• Addresses privacy concerns in V2X machine learning systems
• Protects sensitive user data like location and medical information
• Enables secure neural network inferences for V2X applications

Innovation

• Secure multi-agent system for V2X
• Privacy-preserving neural network inferences
• Efficient secure computation interactions