Mobility-Aware Privacy-Preserving Mobile Crowdsourcing.

Area of Science:

• Computer Science
• Mobile Computing
• Cybersecurity

Background:

• Mobile crowdsourcing services leverage smartphones and location-based services (LBS).
• Existing privacy techniques protecting actual trajectories are vulnerable to inferential attacks.
• Data leakage from failed privacy protection threatens users' location-related private information.

Purpose of the Study:

• To propose a privacy-preserving mobile crowdsourcing solution using predicted trajectories.
• To address the challenge of participating in mobile crowdsourcing without revealing actual locations.

Main Methods:

• Developed TMarkov, a mobility-aware trajectory-prediction solution.
• Incorporated time-partitioning into the Markov model to capture time-varying user behavior.
• Employed Gibbs Sampling for unbiased estimation due to data incompleteness.

Main Results:

• TMarkov characterizes dynamic mobile behaviors for privacy-preserving crowdsourcing.
• Predicted mobility-aware spatiotemporal trajectories enable participation without actual location data.
• Experiments demonstrate TMarkov effectively balances privacy preservation and data usability.

Conclusions:

• TMarkov offers a novel approach to location privacy in mobile crowdsourcing.
• The method enhances user privacy by utilizing predicted trajectories.
• TMarkov provides a practical solution for secure participation in mobile crowdsourcing applications.