Efficient FSS-based Private Statistics for Traffic Monitoring
Zhichao Wang, Qi Feng, Min Luo, Xiaolin Yang, Zizhong Wei · 2024
The emergence of traffic monitoring systems aims to improve living standards, reduce environmental pollution and minimize the economic impact of traffic accidents. These systems mainly function by collecting user data for analysis and formulating corresponding countermeasures. However, the widespread collection of data inevitably raises concerns about the leakage of personal privacy, which is a key safety issue in traffic monitoring. Many researchers are now focusing on traffic flow data statistics, and researchers such as Mohammadali have used homomorphic encryption to design models that enable statistics while protecting user privacy. However, the associated encryption, decryption, and communication overheads are substantial. Therefore, we have moved away from traditional models and encryption/decryption algorithms, proposing a new, efficient, and lightweight system that leverages Distributed Point Functions (DPF) and Distributed Comparison Functions (DCF). This approach not only preserves user privacy but also significantly reduces overhead. Our system is built on a client dual-server framework, designed to withstand attacks from partially sincere adversaries, enabling third-party access to aggregated data or data exceeding a threshold of t, without compromising individual user privacy. Compared to existing related work, our method offers a substantial improvement in operational efficiency while maintaining the same level of communication overhead. Our research provides an advanced solution for aggregation scenarios in traffic monitoring systems, offering lower costs while ensuring robust user privacy. This work lays a strong foundation for the future development of comprehensive traffic monitoring systems.