Location data publishing mechanism in semi-trusted vehicle networks based on differential privacy

Vehicular ad-hoc networks (VANETs) are essential in mobile ad-hoc network applications, utilizing vehicle location data. However, safeguarding the privacy and security of this data is challenging, especially with untrusted roadside units. To address this, we propose a novel differential privacy-based location data publishing mechanism. It consists of three key components: dynamic group leader selection, data aggregation, and privacy-preserving data publishing. The group leader selection algorithm is based on directional thresholds, hop cycles, and lifespans, assuming semi-trusted vehicles. Additionally, the differential privacy-based algorithm uses map grid segmentation, privacy budget recovery, and a filtering mechanism to ensure location privacy. Comparative experiments show that our scheme outperforms existing models in total hops, group leader lifetime, and mean absolute error.