TraCS

Code for paper: [PETS'26] TraCS: Trajectory Collection in Continuous Space under Local Differential Privacy

Contributions:

• The first method to collect 2D trajectories in continuous space under pure LDP.
• Demonstrations of the benefits of operating directly in continuous spaces (rather than discretizing).

Reproductions

This project was developed with Python 3.13 and uses uv for package management. Ensure you have uv installed.

Python packages: To install the required dependencies (in pyproject.toml), cd to the root directory and run:

uv sync

This will create a virtual environment and install all the necessary Python packages with Python version specified in .python-version (3.12 in this case).

To reproduce the results in the paper, you can change the directory to experiments/continuous_space or experiments/discrete_space and run the corresponding script. For example, running

uv run experiment_1_1.py

will print the results of the first experiment in the paper. The csv files containing the results will be saved in the results/ directory, where also contains the drawing scripts for the figures in the paper.

Code Structure

The code is organized as follows:

• src/ contains the source code for the TraCS mechanisms and other methods.
  • src/ldp_mechanisms.py contains the LDP mechanisms used in this paper.
  • src/perturbation_tracs.py contains the implementation of TraCS-D and TraCS-C for each location.
  • src/methods.py wraps the TraCS methods and other methods for a trajectory.
  • src/utilities.py contains some utility functions.
• tests/ contains the testing code for main classes and methods.
• experiments/ contains the code for the experiments in the paper.

In file src/ldp_mechanisms.py, we classify the LDP mechanisms into PiecewiseMechanism and DiscreteMechanism classes. Class PiecewiseMechanism encapsulates the piecewise-based mechanisms for the circular domain and the linear domain, and each perturbation mechanism is implemented as a method of the class. e.g.

perturbed_direction = PiecewiseMechanism(private_val, epsilon).circular_perturbation()

instantiates a PiecewiseMechanism object with the circular domain and perturbs the private value private_val with the given epsilon.

Location-level API. In src/perturbation_tracs.py, TraCS-D and TraCS-C are implemented as the classes DirectionDistancePerturbation and CoordinatePerturbation, respectively. To instantiate a TraCS-D object, use:

tracs_d = DirectionDistancePerturbation(ref_location, location, epsilon, epsilon_d, x_max, y_max)
tracs_d_perturbed = tracs_d.perturb()

Here, ref_location and location denote the reference location and the location to be perturbed, respectively. epsilon is the total privacy budget, while epsilon_d is the portion of the budget allocated to the direction. x_max and y_max specify the bounds of the spatial domain (see Algorithm 1 in the paper).

Implementation tested corrected; see directory tests/ for the testing code.

Trajectory-level API. In file src/methods.py, we wrap the TraCS methods and other methods for a trajectory as functions. For example, to perturb a trajectory with TraCS-D, we can use

perturbed_traj_tracs_d = tracs_d(private_traj, epsilon, epsilon_d, x_max, y_max)

where private_traj is the trajectory to be perturbed, and the other parameters are the same as those for the location-level API.

Freedom of Usage

This project is licensed under the MIT License for freedom of usage and distribution. Hope this paper and code can help you in your research or work.