skip to main content

Title: QuantumHammer: A Practical Hybrid Attack on the LUOV Signature Scheme
Post-quantum schemes are expected to replace existing public-key schemes within a decade in billions of devices. To facilitate the transition, the US National Institute for Standards and Technology (NIST) is running a standardization process. Multivariate signatures is one of the main categories in NIST's post-quantum cryptography competition. Among the four candidates in this category, the LUOV and Rainbow schemes are based on the Oil and Vinegar scheme, first introduced in 1997 which has withstood over two decades of cryptanalysis. Beyond mathematical security and efficiency, security against side-channel attacks is a major concern in the competition. The current sentiment is that post-quantum schemes may be more resistant to fault-injection attacks due to their large key sizes and the lack of algebraic structure. We show that this is not true. We introduce a novel hybrid attack, QuantumHammer, and demonstrate it on the constant-time implementation of LUOV currently in Round 2 of the NIST post-quantum competition. The QuantumHammer attack is a combination of two attacks, a bit-tracing attack enabled via Rowhammer fault injection and a divide and conquer attack that uses bit-tracing as an oracle. Using bit-tracing, an attacker with access to faulty signatures collected using Rowhammer attack, can recover secret key bits more » albeit slowly. We employ a divide and conquer attack which exploits the structure in the key generation part of LUOV and solves the system of equations for the secret key more efficiently with few key bits recovered via bit-tracing. We have demonstrated the first successful in-the-wild attack on LUOV recovering all 11K key bits with less than 4 hours of an active Rowhammer attack. The post-processing part is highly parallel and thus can be trivially sped up using modest resources. QuantumHammer does not make any unrealistic assumptions, only requires software co-location (no physical access), and therefore can be used to target shared cloud servers or in other sandboxed environments. « less
Authors:
; ;
Award ID(s):
1814406
Publication Date:
NSF-PAR ID:
10295733
Journal Name:
CCS '20: 2020 ACM SIGSAC Conference on Computer and Communications Security, Virtual Event, USA, November 9-13, 2020
Sponsoring Org:
National Science Foundation
More Like this
  1. The Schnorr signature scheme is an efficient digital signature scheme with short signature lengths, i.e., $4k$-bit signatures for $k$ bits of security. A Schnorr signature $\sigma$ over a group of size $p\approx 2^{2k}$ consists of a tuple $(s,e)$, where $e \in \{0,1\}^{2k}$ is a hash output and $s\in \mathbb{Z}_p$ must be computed using the secret key. While the hash output $e$ requires $2k$ bits to encode, Schnorr proposed that it might be possible to truncate the hash value without adversely impacting security. In this paper, we prove that \emph{short} Schnorr signatures of length $3k$ bits provide $k$ bits of multi-usermore »security in the (Shoup's) generic group model and the programmable random oracle model. We further analyze the multi-user security of key-prefixed short Schnorr signatures against preprocessing attacks, showing that it is possible to obtain secure signatures of length $3k + \log S + \log N$ bits. Here, $N$ denotes the number of users and $S$ denotes the size of the hint generated by our preprocessing attacker, e.g., if $S=2^{k/2}$, then we would obtain secure $3.75k$-bit signatures for groups of up to $N \leq 2^{k/4}$ users. Our techniques easily generalize to several other Fiat-Shamir-based signature schemes, allowing us to establish analogous results for Chaum-Pedersen signatures and Katz-Wang signatures. As a building block, we also analyze the $1$-out-of-$N$ discrete-log problem in the generic group model, with and without preprocessing.« less
  2. Large-scale quantum computing is a significant threat to classical public-key cryptography. In strong "quantum access" security models, numerous symmetric-key cryptosystems are also vulnerable. We consider classical encryption in a model which grants the adversary quantum oracle access to encryption and decryption, but where the latter is restricted to non-adaptive (i.e., pre-challenge) queries only. We define this model formally using appropriate notions of ciphertext indistinguishability and semantic security (which are equivalent by standard arguments) and call it QCCA1 in analogy to the classical CCA1 security model. Using a bound on quantum random-access codes, we show that the standard PRF- and PRP-basedmore »encryption schemes are QCCA1-secure when instantiated with quantum-secure primitives. We then revisit standard IND-CPA-secure Learning with Errors (LWE) encryption and show that leaking just one quantum decryption query (and no other queries or leakage of any kind) allows the adversary to recover the full secret key with constant success probability. In the classical setting, by contrast, recovering the key uses a linear number of decryption queries, and this is optimal. The algorithm at the core of our attack is a (large-modulus version of) the well-known Bernstein-Vazirani algorithm. We emphasize that our results should *not* be interpreted as a weakness of these cryptosystems in their stated security setting (i.e., post-quantum chosen-plaintext secrecy). Rather, our results mean that, if these cryptosystems are exposed to chosen-ciphertext attacks (e.g., as a result of deployment in an inappropriate real-world setting) then quantum attacks are even more devastating than classical ones.« less
  3. Security of machine learning is increasingly becoming a major concern due to the ubiquitous deployment of deep learning in many security-sensitive domains. Many prior studies have shown external attacks such as adversarial examples that tamper the integrity of DNNs using maliciously crafted inputs. However, the security implication of internal threats (i.e., hardware vulnerabilities) to DNN models has not yet been well understood. In this paper, we demonstrate the first hardware-based attack on quantized deep neural networks–DeepHammer–that deterministically induces bit flips in model weights to compromise DNN inference by exploiting the rowhammer vulnerability. DeepHammer performs an aggressive bit search in themore »DNN model to identify the most vulnerable weight bits that are flippable under system constraints. To trigger deterministic bit flips across multiple pages within a reasonable amount of time, we develop novel system-level techniques that enable fast deployment of victim pages, memory-efficient rowhammering and precise flipping of targeted bits. DeepHammer can deliberately degrade the inference accuracy of the victim DNN system to a level that is only as good as random guess, thus completely depleting the intelligence of targeted DNN systems. We systematically demonstrate our attacks on real systems against 11 DNN architectures with 4 datasets corresponding to different application domains. Our evaluation shows that DeepHammer is able to successfully tamper DNN inference behavior at run-time within a few minutes. We further discuss several mitigation techniques from both algorithm and system levels to protect DNNs against such attacks. Our work highlights the need to incorporate security mechanisms in future deep learning systems to enhance the robustness against hardware-based deterministic fault injections.« less
  4. Anne Canteaut, Yuval Ishai (Ed.)
    It is convenient and common for schemes in the random oracle model to assume access to multiple random oracles (ROs), leaving to implementations the task—we call it oracle cloning—of constructing them from a single RO. The first part of the paper is a case study of oracle cloning in KEM submissions to the NIST Post-Quantum Cryptography standardization process. We give key-recovery attacks on some submissions arising from mistakes in oracle cloning, and find other submissions using oracle cloning methods whose validity is unclear. Motivated by this, the second part of the paper gives a theoretical treatment of oracle cloning. Wemore »give a definition of what is an “oracle cloning method” and what it means for such a method to “work,” in a framework we call read-only indifferentiability, a simple variant of classical indifferentiability that yields security not only for usage in single-stage games but also in multi-stage ones. We formalize domain separation, and specify and study many oracle cloning methods, including common domain-separating ones, giving some general results to justify (prove read-only indifferentiability of) certain classes of methods. We are not only able to validate the oracle cloning methods used in many of the unbroken NIST PQC KEMs, but also able to specify and validate oracle cloning methods that may be useful beyond that.« less
  5. Quantum computing challenges the computational hardness assumptions anchoring the security of public-key ciphers, such as the prime factorization and the discrete logarithm problem. To prepare for the quantum era and withstand the attacks equipped with quantum computing, the security and cryptography communities are designing new quantum-resistant public-key ciphers. National Institute of Standards and Technology (NIST) is collecting and standardizing the post-quantum ciphers, similarly to its past involvements in establishing DES and AES as symmetric cipher standards. The NIST finalist algorithms for public-key signatures are Dilithium, Falcon, and Rainbow. Finding common ground to compare these algorithms can be difficult because ofmore »their design, the underlying computational hardness assumptions (lattice based vs. multivariate based), and the different metrics used for security strength analyses in the previous research (qubits vs. quantum gates). We overcome such challenges and compare the security and the performances of the finalist post-quantum ciphers of Dilithium, Falcon, and Rainbow. For security comparison analyses, we advance the prior literature by using the depth-width cost for quantum circuits (DW cost) to measure the security strengths and by analyzing the security in Universal Quantum Gate Model and with Quantum Annealing. For performance analyses, we compare the algorithms’ computational loads in the execution time as well as the communication costs and implementation overheads when integrated with Transport Layer Security (TLS) and Transmission Control Protocol (TCP)/Internet Protocol (IP). Our work presents a security comparison and performance analysis as well as the trade-off analysis to inform the post-quantum cipher design and standardization to protect computing and networking in the post-quantum era.« less