You are here: Resources > FIDIS Deliverables > Profiling > D7.7: RFID, Profiling, and AmI > 

D7.7: RFID, Profiling, and AmI

RFID, RFID systems* and Identification  Title:
NON-INTERACTIVE AUTHENTICATION AND TRACKING USING RFID
 RFID systems*, AmI-systems and Security

 

Non-interactive Authentication and Tracking using RFID

RFID tags* can be read out remotely from various distances, depending on the used radio frequencies and other physical effects (cp. Annex, chapter ). In the case that a RFID tag* is linked directly (implant) or indirectly (via a product) to a person, it is probable that the person will not be able to observe that the tag has been read, who controls the reader*, how and to what purpose the data from the tag are being processed etc. Like certain types of biometrics (especially behavioural biometrics, but also face recognition; cp. Gasson, Meints Warwick (2005) for further information on Biometrics) RFID can be used for unobserved and non-interactive authentication of persons. This kind of authentication is also called ‘passive authentication’ in non-academic literature.

From a privacy perspective unobserved and non-interactive authentication is of special interest, as this kind of authentication undermines effective exercise of the right of individuals for data self-determination. Although we know of proposed technical measures that allow for example the detection of certain types of RFID tags* and methods to log access to RFID (Floerkemeier, Schneider, Langheinrich, 2004), we do not know of any generally usable and reliable technical method to prevent unobserved and non-interactive authentication of persons via today’s RFID systems*. All known approaches, such as the Faraday cages or so-called blocker tags (Juels, Rivest, Szydlo, 2003), work for known and obvious tags or specific frequencies only. In this context a digital-rights-management-like (DRM*-like) approach for data collected by any RFID reader* on the world has been discussed (Molnar, Soppera, Wagner, 2005) recently, but given the enormous problems with today’s DRM*-approaches this seems neither easy to implement technically, nor easily enforceable on the market.

Rieback, Crispo and Tanenberg (2005) suggest a so called RFID Guardian. This is a PDA like device allowing for detection of tags and readers and supporting among others access management for RFID tags through external readers by acting as RFID proxy. In the context of this concept today only basic functions such as RFID tag* and reader* detection and jamming (for jamming c.p. chapter ) are available in a prototype. Most of the planned access management functions are not supported by todays mainly used RFID tags*.

Another known approach is the implementation of a cryptographic key that restricts the access to the data on an RFID tag* of the type ‘microcontroller’ (cp. Annex, chapter ). Microcontrollers could potentially reduce or take away the need for centralised reference databases for some areas of application of RFID. But today’s implementation of microcontrollers, for example in Machine Readable Travel Documents (MRTDs), show a number of severe technical and conceptual problems, such as cryptographic weaknesses and the access to the data via a - from the perspective of the bearer of the RFID tag* non-trusted – reader*. Future developments of microcontrollers certainly can improve the situation, but as microcontrollers will be much more expensive compared to today’s simple tags used in SCM it is not very likely that they will substitute simple tags in the near future.

RFID today, for example, is used to track goods in the supply chain as already explained. In cases where the link between an RFID tag* and a person is stable, RFID tags* can be used to track persons as well, leading to profiles on their movement (when has she or he been where?). To facilitate this, a network of readers* with known reader* locations is needed. While some experts argue that such a ubiquitous reader* network is not very realistic, others claim that via customer loyalty programs and the corresponding data exchange of many participating shop operators, a far reaching reader* network could easily be established (Garfinkel, Rosenberg, 2005).  

 

 

RFID, RFID systems* and Identification  fidis-wp7-del7.7.RFID_Profiling_AMI_02.sxw  RFID systems*, AmI-systems and Security
Denis Royer 7 / 43