SaTC: CORE: Small: Customizable Geo-Obfuscation to Protect Users' Location Privacy in Mobile Crowdsourcing

University Of North Texas, Denton TX

Investigators

Abstract

Geo-obfuscation is a privacy protection technique that allows users to report obfuscated locations instead of their precise locations in location-based applications such as mobile crowdsourcing. While geo-obfuscation offers strong privacy guarantees, the location errors that arise due to obfuscation can affect the quality of the service. Current research on this topic focuses on minimizing quality loss through optimization techniques, but due to the high computation overhead, it is limited to a small number of locations. Consequently, existing works restrict the scope of geo-obfuscation to either low granularity in a large-scale region or high granularity in a small-scale region, leaving little room for users to customize geo-obfuscation to their specific requirements. The project's novelties are to design a user-centric geo-obfuscation in multi-granularity and multi-scale, which can be customized based on users' diverse demands. The proposed research significantly enhances the flexibility and scalability of existing geo-obfuscation techniques, thereby advancing their application in various mobile contexts. The project's broader significance and importance are to positively impact the wide adoption of many advanced crowdsourcing applications while ensuring individual privacy. In addition, the educational efforts of this project encompass curriculum development, K-12 outreach activities, and the engagement of underrepresented students in research. The project comprises three research tasks. Firstly, the project designs time-efficient algorithms capable of calculating geo-obfuscation for a wide range of location granularity and scales, covering millions of locations in a hierarchy that includes counties, cities, streets, buildings, and rooms. Secondly, based on the developed geo-obfuscation algorithms, the project aims to customize geo-obfuscation according to the demands of a large group of mobile users, taking into account their diverse mobility features and contextual information across different scenarios. Finally, the project identifies potential threats to customizable geo-obfuscation at various granularity levels. On the basis of the insights gained from the threat models, the project designs countermeasures to better protect users' location privacy against the new threats. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.