International Conference on Computer and Knowledge Engineering

Home / 15th International Conference on Computer and Knowledge Engineering

Synthetic Trajectory Sharing Indoors under Privacy Constraints

Authors :

Mahdi Soltanpour¹ Vahideh Moghtadaiee² Mina Alishahi³

1- Cyberspace research institute, Shahid Beheshti University, Tehran, Iran 2- Cyberspace research institute, Shahid Beheshti University, Tehran, Iran 3- Department of Computer Science, Open Universiteit Amsterdam, The Netherlands

Keywords :

Local Differential Privacy،Indoor Trajectory،Trajectory Sharing

Abstract :

User trajectories are essential for services such as route prediction, behavior analysis, and location-based recommendations in indoor navigation systems. However, sharing these trajectories raises serious privacy concerns. While Wi-Fi signals are commonly used for indoor localization, they often require collecting sensitive movement data. To address this, local differential privacy (LDP) offers a solution by perturbing data locally on user devices, eliminating reliance on trusted servers. This paper proposes an LDP-based indoor navigation framework that focuses on protecting full user trajectories. By integrating floor plans and wall structures, we develop a probabilistic model based on principal trajectories to generate synthetic paths that balance privacy and utility. Experiments on two indoor datasets show our method achieves strong balance between privacy and localization accuracy.