Characterizing and Optimizing Differentially-Private Techniques for High-Utility, Privacy-Preserving Internet-of-Vehicles

Yicun Duan, Junyu Liu, Xiaoxing Ming, Wangkai Jin, Zilin Song, Xiangjun Peng

Designed, developed and optimized HUT, a Differential Privacy (DP) enabled algorithm for high data utility and privacy protection in the context of Internet-of-Vehicles (IoV), by leveraging local batch aggregation and order constraints. Evaluated HUT with state-of-the-art algorithms, which achieves 96.9% reduction in information loss while maintaining strong mathematically-guaranteed privacy protection.