Privacy Enhancing Architecture for RFID Labels
Project Manager: Dr. Sandro Etalle
Faculty of Electrical Engineering, Mathematics and Computer Science
Tel.: +31 53 489 1195
Project website: PEARL
The goal of this project is to develop tools and methodologies as well as their theoretical foundations for using RFID systems while preserving the user's privacy. To achieve this, a number of challenges have to be met.
Due to resource constraints a roaming agent faces a dilemma between legitimate use and hostile tracking. Therefore, the first challenge concerns the modelling of privacy properties in this setting. A sound engineering of a privacy respecting architecture requires a formal definition of the involved privacy notions. Such a formalisation precisely documents the requested privacy properties and lays the ground for further analysis and certification. A formal model of the involved privacy properties will be based on an analysis of the problem domain and a literature study.
The second research challenge is to develop new privacy enhancing protocols for the extremely resource constrained RFID environment, where in principle only hashes or custom encryption routines are available. Several problems need to be solved. For one thing, RFID tags should implement a simple yet reliable access control mechanism to limit access to authorised parties only. Current RFID tags respond predictably when queried by a transceiver, enabling invasive tracking opportunities, that have to be prevented. The so called "blocker tags" provide only a primitive solution to the privacy problem, useless for many applications in the ambient world that require RFIDs to be continuously present.
The third research challenge is to shape a context in which the user can check the privacy policies enforced by the RFID-based application. To tackle this, first we have to investigate which sort of privacy policies can be deployed given the kind of hardware. Secondly, we need to design new specific privacy languages and to investigate the possibility of automatically checking when two policies are (ir)reconcilable.
The fourth challenge concerns securing the integration of RFID tags and the backoffice applications that will support them. To this end we need to design and verify new secure protocols for data access. This will lead to a combination of two research fields, namely cryptography and datamining.
Project duration: January 2007 until January 2011
Project budget: 644 K-€
Number of person/years: 2.5 fte
Participants: UT, Radboud University Nijmegen, TU Delft, TU/e
Project budget CTIT: 165 K-€
Number of person/years CTIT: 1.2 fte
Involved groups: Distributed and Embedded Systems (DIES)