Location Privacy for Mobile Crowd Sensing through Population Mapping.

Area of Science:

• Computer Science
• Mobile Computing
• Data Privacy

Background:

• Opportunistic sensing utilizes mobile devices for context data collection, raising significant user privacy concerns due to embedded time and location information.
• Existing methods struggle to balance data utility with robust de-anonymization protection, even after data scrubbing.

Purpose of the Study:

• To propose and evaluate a novel spatiotemporal blurring mechanism for enhancing user privacy in opportunistic sensing applications.
• To enable efficient, local data anonymization before transmission, mitigating risks associated with centralized processing.

Main Methods:

• Developed a privacy-preserving technique based on tessellation and clustering for spatiotemporal data blurring.
• Implemented a probabilistic k-anonymity model allowing adjustable privacy-quality trade-offs.
• Evaluated the algorithm's effectiveness using real-world mobility traces.

Main Results:

• The proposed spatiotemporal blurring mechanism effectively protects user privacy against system-level de-anonymization attempts.
• The tessellation and clustering approach provides controllable levels of location privacy and data quality.
• Local blurring demonstrated efficiency, eliminating the need for an online anonymization server.

Conclusions:

• The novel spatiotemporal blurring technique offers a viable solution for privacy-preserving opportunistic sensing.
• The system effectively balances the need for context reporting with the imperative of safeguarding user anonymity.
• This approach enhances the security and trustworthiness of mobile crowdsensing applications.