Differentially Private Adaptive Noise Injection DP-ANI

This section explains the details and the steps of the official implementation of our paper:

The original paper experiments a private dataset however as an example, this repository applies DPMM on San Francisco city taxi dataset. However since the proposed model relies on the road segment classes (functional class) it requires to have a dedicated road network.

Pre-requisites

There are several main packages that requires DPMM to work. We listed the important packages below.

• GeoPandas
• OSMnx
• NetworkX
• Shapely
• UTM converter

DP-ANI Aggregated Mobility Network Generation

Follow these steps to reproduce the results of our paper.

• Run experiments with

  map_matcher_traj_clean_2024.py

  This step generates clean link trajectories that will be used for noise injection.

• Run experiments with

  map_matcher_noise_everyOD_2024.py

  This step generates adds noise to the origin destinations of the road segment with proposed DP-ANI.

Notes:

1. The road network file is a confidential graph used internally for experiment reproduction. As a result, it is not included in the repository.
2. map_matcher_noise_everyOD_2024.py only generates the proposed DPANI method with different epsilon values. It does not reproduce baseline experiments.

Citation

If you find our work useful in your research, please cite our paper:

@article{haydari2021differential,
  title={Differential Privacy in Aggregated Mobility Networks: Balancing Privacy and Utility},
  author={Haydari, Ammar and Chuah, Chen-Nee and Zhang, Michael and Macfarlane, Jane and Peisert, Sean},
  journal={arXiv preprint arXiv:2112.08487},
  year={2021}
}

Copyright Notice

Differentially Private Adaptive Noise Injection (DP-ANI) Copyright (c) 2025, The Regents of the University of California, through Lawrence Berkeley National Laboratory (subject to receipt of any required approvals from the U.S. Dept. of Energy) and University of California, Davis. All rights reserved.

If you have questions about your rights to use or distribute this software, please contact Berkeley Lab's Intellectual Property Office at IPO@lbl.gov.

NOTICE. This Software was developed under funding from the U.S. Department of Energy and the U.S. Government consequently retains certain rights. As such, the U.S. Government has been granted for itself and others acting on its behalf a paid-up, nonexclusive, irrevocable, worldwide license in the Software to reproduce, distribute copies to the public, prepare derivative works, and perform publicly and display publicly, and to permit others to do so.