Leonid Bolotnyy

Physically Unclonable Function-Based Security and Privacy in RFID Systems

Radio frequency identification (RFID) is an increasingly popular technology that uses radio signals for object identification. Tracking and authentication in RFID tags have raised many privacy and security concerns. On the other hand, known privacy and security cryptographic defenses are too hardware-expensive to incorporate into low-cost RFID tags. In this paper, we propose hardware-based approaches to RFID security that rely on physically unclonable functions (PUFs). These functions exploit the inherent variability of wire delays and parasitic gate delays in manufactured circuits, and may be implemented with an order-of-magnitude reduction in gate count as compared with traditional cryptographic functions. We describe protocols for privacy-preserving tag identification and secure message authentication codes. We compare PUFs to digital cryptographic functions, address other uses of PUFs to enhance RFID security and suggest interesting directions for future research. The proposed solutions are efficient, practical, and appropriate for low-cost RFID systems

Generalized "Yoking-Proofs" for a Group of RFID Tags

Recently Ari Juels suggested a yoking-proof where a pair of radio-frequency identification (RFID) tags are both read within a specified time bound, and left open for future research the problem of generating a proof for larger groups of tags. We generalize his protocol by developing a proof which ensures that a group of tags is read within a certain time period. The tags generate such a proof even if the reader is untrusted. The proof is improbable to forge, and is verifiable off-line by a trusted verifier. Juels problem formulation does not lake privacy into account and the resulting protocol offers no privacy to the tags. We modify the problem statement to require the yoking-proof to maintain privacy, and we give a protocol for this new anonymous yoking problem, along with suggestions for speed ups

The Case for Multi-Tag RFID Systems

Radio frequency identification (RFID) is a promising technology for automated non-line-of-sight object identification. However, factors such as object occlusions, metal/liquid opaqueness, environmental conditions, and radio noise degrade the overall availability, reliability, and dependability of RFID systems. We show that simply increasing the number of readers does not adequately address these issues. Instead, we propose tagging each object with multiple tags, and provide definitive experimental data showing that this strategy dramatically improves the effectiveness of RFID systems in the face of radio noise and other interfering factors. We solidify the case for multi-tag RFID systems by addressing obstacles to reliable object detection, and analyzing how multi-tags improve tag detection, even in the presence of (radio-opaque) metals and liquids. We discuss applications that will benefit considerably from multi-tags, and propose careful RFID system design through the deployment of appropriate types of multi-tags and anti-collision algorithms. We also analyze the economics of multi-tag RFID systems and argue that the benefits of multi-tags can substantially outweigh the costs in many current applications, and that this trend will become even more pronounced in the future.

Multi-tag RFID systems

Successful object identification is the primary objective of Radio Frequency Identification (RFID) technology. Yet, a recent major study by Wal-Mart has shown that object detection probability can be as low as 66%. We propose the tagging of objects with multiple tags to address the fundamental issue of object detectability. We show that this strategy dramatically improves the efficacy of RFID systems, even in the face of radio noise and other interfering factors. We define different types of multi-tag systems and examine their benefits using analytics, simulations, and experiments with commercial RFID equipment. We investigate the effect of multi-tags on anti-collision algorithms, and develop several techniques that enable multi-tags to enhance RFID security. We suggest new promising applications of multi-tags, ranging from improving patient safety to preventing illegal deforestation. We analyse the economics of multi-tag RFID systems and argue that the benefits of multi-tags will continue to increasingly outweigh their costs in many applications.

Multi-tag radio frequency identification systems

We propose and analyze the effects of attaching more than one RFID tag to each object. We define different types of multi-tag systems and examine their benefits, both analytically and empirically. We also analyze how multi-tags affect some existing tag singulation algorithms. We show how multi-tags can serve as security enhancers, and propose several new promising applications of multi-tags, such as preventing illegal deforestation.

Randomized pseudo-random function tree walking algorithm for secure radio-frequency identification

Privacy and security are two main concerns in radio frequency identification (RFID) systems. We first extend the analysis of the randomized tree walking algorithm for RFID tag collision avoidance, which is secure against passive adversaries. Then, we devise a new randomized pseudo-random function (PRF) tree walking algorithm, which is secure against active eavesdroppers and allows for the efficient interrogation of many tags. Our algorithm accommodates the addition and removal of tags from the system, and dynamically adapts to security and privacy policy changes.

Generalized "Yoking-Proofs" and Inter-Tag Communication

Some radio-frequency identification (RFID) scenarios require a proof of action (e.g., that a group of objects tagged with RFID tags were identified simultaneously). For example, pharmaceutical distributors may want to prove that a bottle of medicine was sold together with its instructions leaflet [Juels, 2004]; manufacturers may want to prove that safety devices were sold together with a tool or that a number of matching parts were delivered simultaneously; banking centers or security stations may want to prove that several forms of ID were read simultaneously; meeting organizers may want to prove that a group of people were present together at a meeting, etc. Third-parties can verify the validity of such proofs. For examples above, the verifying third parties can be regulatory agencies, company headquarters, etc. We seek to ensure that if a group of tags is not read nearly-simultaneously, an entity (RFID reader) will not be able to forge a proof that they were by constructing a valid forged proof. Inspired by this problem, Ari Juels developed a protocol that creates such a proof for a pair of RFID tags [Juels, 2004]. He left open for future research the problem of generalizing his protocol to three or more tags. We develop a methodology that generalizes yoking-proof protocols to arbitrarily large groups of tags. We also define an anonymous yoking problem and propose an efficient solution for it [Bolotnyy & Robins, 2006]. Finally, we show how these yoking protocols can be sped up.