Dominique Unruh

Zero-Knowledge in the Applied Pi-calculus and Automated Verification of the Direct Anonymous Attestation Protocol

We devise an abstraction of zero-knowledge protocols that is accessible to a fully mechanized analysis. The abstraction is formalized within the applied pi-calculus using a novel equational theory that abstractly characterizes the cryptographic semantics of zero-knowledge proofs. We present an encoding from the equational theory into a convergent rewriting system that is suitable for the automated protocol verifier ProVerif. The encoding is sound and fully automated. We successfully used ProVerif to obtain the first mechanized analysis of (a simplified variant of) the Direct Anonymous Attestation (DAA) protocol. This required us to devise novel abstractions of sophisticated cryptographic security definitions based on interactive games. The analysis reported a novel attack on DAA that was overlooked in its existing cryptographic security proof. We propose a revised variant of DAA that we successfully prove secure using ProVerif.

Compromising Reflections-or-How to Read LCD Monitors around the Corner

We present a novel eavesdropping technique for spying at a distance on data that is displayed on an arbitrary computer screen, including the currently prevalent LCD monitors. Our technique exploits reflections of the screens optical emanations in various objects that one commonly finds in close proximity to the screen and uses those reflections to recover the original screen content. Such objects include eyeglasses, tea pots, spoons, plastic bottles, and even the eye of the user. We have demonstrated that this attack can be successfully mounted to spy on even small fonts using inexpensive, off-the-shelf equipment (less than 1500 dollars) from a distance of up to 10 meters. Relying on more expensive equipment allowed us to conduct this attack from over 30 meters away, demonstrating that similar attacks are feasible from the other side of the street or from a close-by building. We additionally establish theoretical limitations of the attack; these limitations may help to estimate the risk that this attack can be successfully mounted in a given environment.

Universally composable quantum multi-party computation

The Universal Composability model (UC) by Canetti (FOCS 2001) allows for secure composition of arbitrary protocols. We present a quantum version of the UC model which enjoys the same compositionality guarantees. We prove that in this model statistically secure oblivious transfer protocols can be constructed from commitments. Furthermore, we show that every statistically classically UC secure protocol is also statistically quantum UC secure. Such implications are not known for other quantum security definitions. As a corollary, we get that quantum UC secure protocols for general multi-party computation can be constructed from commitments.

Towards key-dependent message security in the standard model

Standard security notions for encryption schemes do not guarantee any security if the encrypted messages depend on the secret key. Yet it is exactly the stronger notion of security in the presence of key-dependent messages (KDM security) that is required in a number of applications: most prominently, KDM security plays an important role in analyzing cryptographic multi-party protocols in a formal calculus. But although often assumed, the mere existence of KDM secure schemes is an open problem. The only previously known construction was proven secure in the random oracle model. We present symmetric encryption schemes that are KDM secure in the standard model (i.e., without random oracles). The price we pay is that we achieve only a relaxed (but still useful) notion of key-dependent message security. Our work answers (at least partially) an open problem posed by Black, Rogaway, and Shrimpton. More concretely, our contributions are as follows: 1. We present a (stateless) symmetric encryption scheme that is information-theoretically secure in face of a bounded number and length of encryptions for which the messages depend in an arbitrary way on the secret key. 2. We present a stateful symmetric encryption scheme that is computationally secure in face of an arbitrary number of encryptions for which the messages depend only on the respective current secret state/key of the scheme. The underlying computational assumption is minimal: we assume the existence of one-way functions. 3. We give evidence that the only previously known KDM secure encryption scheme cannot be proven secure in the standard model (i.e., without random oracles).

Computationally sound verification of source code

Increasing attention has recently been given to the formal verification of the source code of cryptographic protocols. The standard approach is to use symbolic abstractions of cryptography that make the analysis amenable to automation. This leaves the possibility of attacks that exploit the mathematical properties of the cryptographic algorithms themselves. In this paper, we show how to conduct the protocol analysis on the source code level (F# in our case) in a computationally sound way, i.e., taking into account cryptographic security definitions. We build upon the prominent F7 verification framework (Bengtson et al., CSF 2008) which comprises a security type-checker for F# protocol implementations using symbolic idealizations and the concurrent lambda calculus RCF to model a core fragment of F#. To leverage this prior work, we give conditions under which symbolic security of RCF programs using cryptographic idealizations implies computational security of the same programs using cryptographic algorithms. Combined with F7, this yields a computationally sound, automated verification of F# code containing public-key encryptions and signatures. For the actual computational soundness proof, we use the CoSP framework (Backes, Hofheinz, and Unruh, CCS 2009). We thus inherit the modularity of CoSP, which allows for easily extending our proof to other cryptographic primitives.

Non-Interactive Zero-Knowledge Proofs in the Quantum Random Oracle Model

We present a construction for non-interactive zero-knowledge proofs of knowledge in the random oracle model from general sigma-protocols. Our construction is secure against quantum adversaries. Prior constructions (by Fiat-Shamir and by Fischlin) are only known to be secure against classical adversaries, and Ambainis, Rosmanis, Unruh (FOCS 2014) gave evidence that those constructions might not be secure against quantum adversaries in general.

OAEP Is Secure under Key-Dependent Messages

Key-dependent message security, short KDM security, was introduced by Black, Rogaway and Shrimpton to address the case where key cycles occur among encryptions, e.g., a key is encrypted with itself. We extend this definition to include the cases of adaptive corruptions and arbitrary active attacks, called adKDM security incorporating several novel design choices and substantially differing from prior definitions for public-key security. We also show that the OAEP encryption scheme (using a partial-domain one-way function) satisfies the strong notion of adKDM security in the random oracle model.The OAEP construction thus constitutes a suitable candidate for implementating symbolic abstractions of encryption schemes in a computationally sound manner under active adversaries.

A Formal Language for Cryptographic Pseudocode

Game-based cryptographic proofs are typically expressed using pseudocode, which lacks a formal semantics. This can lead to ambiguous specifications, hidden mistakes, and even wrong proofs. We propose a language for expressing proofs that is expressive enough to specify all constructs occurring in cryptographic games, including probabilistic behaviors, the usage of oracles, and polynomial-time programs. The language is a probabilistic higher-order lambda calculus with recursive types, references, and support for events, and is simple enough that researchers without a strong background in the theory of programming languages can understand it. The language has been implemented in the proof assistant Isabelle/HOL.

Random oracles and auxiliary input

We introduce a variant of the random oracle model where oracle-dependent auxiliary input is allowed. In this setting, the adversary gets an auxiliary input that can contain information about the random oracle. Using simple examples we show that this model should be preferred over the classical variant where the auxiliary input is independent of the random oracle. In the presence of oracle-dependent auxiliary input, the most important proof technique in the random oracle model--lazy sampling--does not apply directly. We present a theorem and a variant of the lazy sampling technique that allows one to perform proofs in the new model almost as easily as in the old one. As an application of our approach and to illustrate how existing proofs can be adapted, we prove that RSA-OAEP is IND-CCA2 secure in the random oracle model with oracle-dependent auxiliary input.

Comparing Two Notions of Simulatability

In this work, relations between the security notions standard simulatability and universal simulatability for cryptographic protocols are investigated. A simulatability-based notion of security considers a protocol π as secure as an idealization τ of the protocol task, if and only if every attack on π can be simulated by an attack on τ. Two formalizations, which both provide secure composition of protocols, are common: standard simulatability means that for every π-attack and protocol user H, there is a τ-attack, such that H cannot distinguish π from τ. Universal simulatability means that for every π-attack, there is a τ-attack, such that no protocol user H can distinguish π from τ. Trivially, universal simulatability implies standard simulatability. We show: the converse is true with respect to perfect security, but not with respect to computational or statistical security. Besides, we give a formal definition of a time-lock puzzle, which may be of independent interest. Although the described results do not depend on any computational assumption, we show that the existence of a time-lock puzzle gives an even stronger separation of standard and universal simulatability with respect to computational security.