Homomorphic Encryption Preserves Vehicle Data Privacy in Real-World Tests

The increasing reliance on data from connected vehicles presents a significant challenge to user privacy. Yet, this data is crucial for improving traffic flow and safety. Abdullah Al Mamun, Kyle Yates, and Antsa Rakotondrafara, along with colleagues at Clemson University, address this issue by experimentally evaluating the potential of post-quantum homomorphic encryption (HE) for vehicle-to-everything (V2X) communication. Their research represents the first real-world assessment of three leading HE schemes, Brakerski-Fan-Vercauteren, Brakerski-Gentry-Vaikuntanathan, and Cheon-Kim-Kim-Song, applied to practical scenarios such as vehicle counting and average speed aggregation. The team’s findings demonstrate that HE can realistically protect sensitive data in intelligent transportation systems, offering viable latency for applications like traffic monitoring and regional analysis, and paving the way for secure and privacy-preserving data processing in future connected vehicle technologies.

Proactive ITS Cybersecurity Needs Robust Defenses

Intelligent Transportation Systems (ITS) increasingly rely on data exchange between vehicles and infrastructure, creating vulnerabilities to cyberattacks. These systems, designed to enhance safety and efficiency, become potential targets, necessitating robust cybersecurity measures. The growing complexity of ITS, coupled with the increasing connectivity of vehicles, expands the potential attack surface and elevates the risk of malicious interference. Consequently, ensuring the cybersecurity of ITS is paramount for maintaining public safety and the reliable operation of transportation networks. Current approaches often focus on detecting and mitigating attacks after they occur, which can be reactive and insufficient to prevent significant disruptions.

Proactive cybersecurity strategies, anticipating and preventing attacks before they impact system performance, are therefore crucial. Existing methods frequently lack the ability to comprehensively assess the cascading effects of cyberattacks on interconnected transportation systems, hindering effective risk management. This research addresses the need for a proactive and comprehensive cybersecurity framework for ITS by developing a novel methodology for assessing and mitigating the cascading impacts of cyberattacks. The primary objective is to create a system-level model that captures the interdependencies within ITS, allowing for the prediction of how cyberattacks spread through the network, ultimately providing transportation agencies with a tool to proactively manage cybersecurity risks.

Privacy Preserving Data Analysis in ITS

This research explores the feasibility and application of Homomorphic Encryption (HE) within Intelligent Transportation Systems (ITS). HE allows computations to be performed on encrypted data without decrypting it first, crucial for protecting sensitive information. ITS generates vast amounts of data concerning location, speed, and driving behavior, and protecting this data is paramount. HE offers a solution by enabling data analysis and sharing while maintaining confidentiality. The research explores various HE schemes, including leveled and fully HE (FHE), and their suitability for different ITS applications.

A major challenge is the significant computational overhead associated with HE, and research focuses on optimizing HE implementations for real-time ITS applications. The threat of quantum computers breaking current encryption algorithms is also addressed, with research exploring post-quantum HE schemes to ensure long-term security. HE applications are investigated in areas including privacy-preserving control of traffic signals, improving traffic flow, secure data sharing, and analyzing traffic patterns without compromising individual privacy.

Post-Quantum Encryption Enables Vehicle Data Analysis

Intelligent Transportation Systems (ITS) increasingly rely on data from vehicles to manage traffic and improve safety, creating significant privacy concerns. Researchers have investigated homomorphic encryption (HE) as a solution, and have now conducted the first real-world evaluation of three post-quantum HE schemes, Brakerski-Fan-Vercauteren (BFV), Brakerski-Gentry-Vaikuntanathan (BGV), and Cheon-Kim-Kim-Song (CKKS), within vehicular communication scenarios. The study focused on two common applications: counting vehicles and aggregating average speeds, testing performance over both Wi-Fi and wired connections. The results demonstrate that BFV and BGV are well-suited for applications where some delay is acceptable, such as monitoring intersections or analyzing regional traffic patterns, achieving end-to-end latencies under 10 seconds.

While CKKS introduces more computational overhead, it remains viable for periodically aggregating numerical data, like speed readings, offering a trade-off between precision and performance. This demonstrates that HE can be practically deployed in ITS environments with appropriate security levels, provided that the specific latency requirements of each application are considered. Importantly, this research moves beyond simulations by using actual hardware and network connections, providing a realistic assessment of HE’s impact on system performance. Previous work often relied on theoretical models or partially homomorphic encryption, whereas this study utilizes post-quantum schemes designed to withstand future threats. The findings confirm that HE offers a foundational tool for building secure and privacy-preserving systems for connected vehicles, enabling data processing without compromising sensitive information.

Homomorphic Encryption Enables Secure Traffic Analysis

This study presents the first real-world evaluation of three post-quantum secure homomorphic encryption (HE) schemes, Brakerski-Fan-Vercauteren (BFV), Brakerski-Gentry-Vaikuntanathan (BGV), and Cheon-Kim-Kim-Song (CKKS), within intelligent transportation systems. Experiments conducted using both Wi-Fi and Ethernet networks demonstrate the practical feasibility of these schemes for latency-tolerant applications involving vehicle-generated data, such as congestion monitoring and regional traffic analysis. The results indicate that BFV and BGV offer suitable performance for applications requiring updates within seconds, while CKKS remains viable for periodic aggregation of numerical data, despite experiencing higher computational overhead. The research confirms that HE can enable secure and privacy-preserving vehicle-to-everything communication under current security standards, provided application latency requirements align with each scheme’s capabilities. Future work will focus on optimizing HE parameters, exploring hardware acceleration, and conducting real-world deployment trials to improve scalability and validate performance.