TPRT: a trajectory publishing scheme for the Internet of Vehicles based on radix tree

Abstract The location and trajectory information generated in the Internet of Vehicles (IoV) provides sufficient data support for road network planning. However, there is much sensitive information in these data, and publishing them directly will lead to serious privacy leakage risks. Therefore, this paper proposes a trajectory publishing scheme for the Internet of Vehicles based on radix tree (TPRT). Firstly, the trajectory data are divided into multiple location planes based on timestamps and processed using clustering and generalization. This approach addresses the issue of real-life trajectory data having almost no identical prefixes, making it more suitable for storage in radix tree. Then, the directed graph and the shortest path algorithm are utilized to synthesize a new trajectory dataset for publication. Subsequently, a radix tree structure that satisfies differential privacy is defined. In comparison to the research method employing a prefix tree, the radix tree not only captures the spatiotemporal characteristics of the trajectories but also reduces space consumption. Finally, a novel method of noise addition is proposed. In contrast to the traditional layer-by-layer noise addition approach, our method reduces the cost of noise addition and enhances data availability. Experimental results demonstrate that TPRT exhibits superior data availability compared to the baseline methods.