Robust Fingerprint of Location Trajectories Under Differential Privacy

Area of Science:

• Computer Science
• Data Privacy
• Cybersecurity

Background:

• Location-based services generate valuable data, but sharing it poses privacy risks and concerns over unauthorized redistribution.
• Existing differentially-private mechanisms often degrade data utility, hindering analysis.
• Protecting sensitive user movement patterns and locations is crucial.

Purpose of the Study:

• To develop a novel fingerprinting scheme for location datasets.
• To simultaneously ensure data privacy through differential privacy and detect unauthorized redistribution.
• To enhance the utility of differentially-private location data.

Main Methods:

• A novel fingerprinting scheme is proposed, integrating a utility-focused post-processing technique.
• The scheme incorporates a sampling method based on the post-processing to preserve spatiotemporal correlations.
• Differential privacy is maintained throughout the process due to immunity to post-processing.

Main Results:

• The proposed scheme effectively identifies unauthorized redistribution of location datasets.
• It provides strong differential privacy guarantees, remaining robust against various attacks like flipping and collusion.
• Experimental results show high fingerprinting robustness and improved data utility compared to existing methods.

Conclusions:

• The novel fingerprinting scheme offers a robust solution for secure location data sharing.
• It successfully balances privacy protection with data utility, benefiting data analysis.
• The approach is resilient to sophisticated attacks, ensuring data integrity and accountability.