International Association
for Cryptologic Research

Fully homomorphic encryption (FHE) is an advanced cryptography technique to allow computations (i.e., addition and multiplication) over encrypted data. After years of effort, the performance of FHE has been significantly improved and it has moved from theory to practice. The transciphering framework is another important technique in FHE to address the issue of ciphertext expansion and reduce the client-side computational overhead. To apply the transciphering framework to the CKKS FHE scheme, a new transciphering framework called the Real-to-Finite-Field (RtF) framework and a corresponding FHE-friendly symmetric-key primitive called HERA were proposed at ASIACRYPT 2021. Although HERA has a very similar structure to AES, it is considerably different in the following aspects: 1) the power map x → x3 is used as the S-box; 2) a randomized key schedule is used; 3) it is over a prime field Fp with p > 216. In this work, we perform the first third-party cryptanalysis of HERA, by showing how to mount new algebraic attacks with multiple collisions in the round keys. Specifically, according to the special way to randomize the round keys in HERA, we find it possible to peel off the last nonlinear layer by using collisions in the last-round key and a simple property of the power map. In this way, we could construct an overdefined system of equations of a much lower degree in the key, and efficiently solve the system via the linearization technique. As a esult, for HERA with 192 and 256 bits of security, respectively, we could break some parameters under the same assumption made by designers that the algebra constant ω for Gaussian elimination is ω = 2, i.e., Gaussian elimination on an n × n matrix takes O(nω) field operations. If using more conservative choices like ω ∈ {2.8, 3}, our attacks can also successfully reduce the security margins of some variants of HERA to only 1 round. However, the security of HERA with 80 and 128 bits of security is not affected by our attacks due to the high cost to find multiple collisions. In any case, our attacks reveal a weakness of HERA caused by the randomized key schedule and its small state size.

A secure n-bit tweakable block cipher (TBC) using t-bit tweaks can be modeled as a tweakable uniform random permutation, where each tweak defines an independent random n-bit permutation. When an input to this tweakable permutation is fixed, it can be viewed as a perfectly secure t-bit random function. On the other hand, when a tweak is fixed, it can be viewed as a perfectly secure n-bit random permutation, and it is well known that the sum of two random permutations is pseudorandom up to 2n queries.A natural question is whether one can construct a pseudorandom function (PRF) beyond the block and the tweak length bounds using a small number of calls to the underlying tweakable permutations. A straightforward way of constructing a PRF from tweakable permutations is to xor the outputs from two tweakable permutations with c bits of the input to each permutation fixed. Using the multi-user security of the sum of two permutations, one can prove that the (t + n − c)-to-n bit PRF is secure up to 2n+c queries.In this paper, we propose a family of PRF constructions based on tweakable permutations, dubbed XoTPc, achieving stronger security than the straightforward construction. XoTPc is parameterized by c, giving a (t + n − c)-to-n bit PRF. When t < 3n and c = t/3 , XoTPt/3 becomes an (n + 2t/3 )-to-n bit pseudorandom function, which is secure up to 2n+2t/3 queries. It provides security beyond the block and the tweak length bounds, making two calls to the underlying tweakable permutations. In order to prove the security of XoTPc, we extend Mirror theory to q ≫ 2n, where q is the number of equations. From a practical point of view, our construction can be used to construct TBC-based MAC finalization functions and CTR-type encryption modes with stronger provable security compared to existing schemes.

The main goal of this work is to construct authenticated encryption (AE) hat is both committing and leakage-resilient. As a first approach for this we consider generic composition as a well-known method for constructing AE schemes. While the leakage resilience of generic composition schemes has already been analyzed by Barwell et al. (Asiacrypt’17), for committing security this is not the case. We fill this gap by providing a separate analysis of the generic composition paradigms with respect to committing security, giving both positive and negative results: By means of a concrete attack, we show that Encrypt-then-MAC is not committing. Furthermore, we prove that Encrypt-and-MAC is committing, given that the underlying schemes satisfy security notions we introduce for this purpose. We later prove these new notions achievable by providing schemes that satisfy them. MAC-then-Encrypt turns out to be more difficult due to the fact that the tag is not outputted alongside the ciphertext as it is done for the other two composition methods. Nevertheless, we give a detailed heuristic analysis of MAC-then-Encrypt with respect to committing security, leaving a definite result as an open task for future work. Our results, in combination with the fact that only Encrypt-then-MAC yields leakage-resilient AE schemes, show that one cannot obtain AE schemes that are both committing and leakage-resilient via generic composition. As a second approach for constructing committing and leakage-resilient AE, we develop a generic transformation that turns an arbitrary AE scheme into one that fulfills both properties. The transformation relies on a keyed function that is both binding, i.e., it is hard to find key-input pairs that result in the same output, and leakage-resilient pseudorandom.

QARMAv2 is a general-purpose and hardware-oriented family of lightweight tweakable block ciphers (TBCs) introduced in ToSC 2023. QARMAv2, as a redesign of QARMAv1 with a longer tweak and tighter security margins, is also designed to be suitable for cryptographic memory protection and control flow integrity. The designers of QARMAv2 provided a relatively comprehensive security analysis in the design specification, e.g., some bounds for the number of attacked rounds in differential and boomerang analysis, together with some concrete impossible differential, zerocorrelation, and integral distinguishers. As one of the first third-party cryptanalysis of QARMAv2, Hadipour et al., [HGSE24] significantly improved the integral distinguishers of QARMAv2, and provided the longest concrete distinguishers of QARMAv2 up to now. However, they provided no key recovery attack based on their distinguishers. This paper delves into the cryptanalysis of QARMAv2 to enhance our understanding of its security. Given that the integral distinguishers of QARMAv2 are the longest concrete distinguishers for this cipher so far, we focus on integral attack. To this end, we first further improve the automatic tool introduced by Hadipour et al. [HSE23,HGSE24] for finding integral distinguishers of TBCs following the TWEAKEY framework. This new tool exploits the MixColumns property of QARMAv2 to find integral distinguishers more suitable for key recovery attacks. Then, we combine several techniques for integral key recovery attacks, e.g., Meet-in-the-middle and partial-sum techniques to build a fine-grained integral key recovery attack on QARMAv2. Notably, we demonstrate how to leverage the low data complexity of the integral distinguishers of QARMAv2 to reduce the memory complexity of the meet-in-the-middle technique. As a result, we successfully present the first concrete key recovery attacks on reduced-round versions of QARMAv2. This includes attacking 13 rounds of QARMAv2-64-128 with a single tweak block (T = 1), 14 rounds of QARMAv2-64-128 with two independent tweak blocks (T = 2), and 16 rounds of QARMAv2-128-256 with two independent tweak blocks (T = 2), all in an unbalanced setting. Our attacks do not compromise the claimed security of QARMAv2, but they shed more light on the cryptanalysis of this cipher.

The linear layer of block ciphers plays an important role in their security In particular, ciphers designed following the wide-trail strategy use the branch number of the linear layer to derive bounds on the probability of linear and differential trails. At FSE 2014, the LS-design construction was introduced as a simple and regular structure to design bitsliced block ciphers. It considers the internal state as a bit matrix, and applies alternatively an identical S-Box on all the columns, and an identical L-Box on all the lines. Security bounds are derived from the branch number of the L-Box.In this paper, we focus on bitsliced linear layers inspired by the LS-design construction and the Spook AEAD algorithm. We study the construction of bitsliced linear transformations with efficient implementations using XORs and rotations (optimized for bitsliced ciphers implemented on 32-bit processors), and a high branch number. In order to increase the density of the activity patterns, the linear layer is designed on the whole state, rather than using multiple parallel copies of an L-Box. Our main result is a linear layer for 128-bit ciphers with branch number 21, improving upon the best 32-bit transformation with branch number 12, and the one of Spook with branch number 16.

This paper focuses on equivalences between Generalised Feistel Networks (GFN) of type-II. We introduce a new definition of equivalence which captures the concept that two GFNs are identical up to re-labelling of the inputs/outputs, and give a procedure to test this equivalence relation. Such two GFNs are therefore cryptographically equivalent for several classes of attacks. It induces a reduction o the space of possible GFNs: the set of the (k!)2 possible even-odd GFNs with 2k branches can be partitioned into k! different classes.This result can be very useful when looking for an optimal GFN regarding specific computationally intensive properties, such as the minimal number of active S-boxes in a differential trail. We also show that in several previous papers, many GFN candidates are redundant as they belong to only a few classes. Because of this reduction of candidates, we are also able to suggest better permutations than the one of WARP: they reach 64 active S-boxes in one round less and still have the same diffusion round that WARP. Finally, we also point out a new family of permutations with good diffusion properties.

Impossible differential cryptanalysis is very important in the field of symmetric ciphers. Currently, there are many automatic search approaches to find impossible differentials. However, these methods have two underlying assumptions: Markov cipher assumption and key independence assumption. Actually, these two assumptions are not true in ARX ciphers, especially lightweight ones. In this paper, we study the impossible differentials in ARX cipher under weak keys for the first time. Firstly, we propose several accurate difference propagation properties on consecutive two and three modular additions. Then, these properties are applied to four typical local constructions composed of two consecutive modular additions, two modular additions with a rotation operation, xoring secret key or constant in the middle, to find impossible differentials under weak keys or special constants. What’s more, we propose a more accurate difference propagation property on three consecutive modular additions. It can be used to find impossible differentials on more complex local constructions under weak keys or special constants. In practical ciphers, these impossible differentials on local constructions can be used to find contradictions. Lastly, combining our new findings with traditional automatic search methods for impossible differentials, we propose a framework to find impossible differentials in ARX ciphers under weak keys. As applications, we apply the framework to SPECK-32/64, LEA and CHAM-64/128. As a result, we find two 8-round impossible differentials for SPECK-32/64 under 260 weak keys, and one 11-round impossible differential for LEA under 2k−1 weak keys, where k is the key size. These impossible differentials can start from any round. Furthermore, we find two 22-round impossible differentials for CHAM-64/128 under 2127 weak keys starting from certain rounds. As far as we know, all these impossible differentials are longer than previous ones.

The Nostradamus attack was originally proposed as a security vulnerability for a hash function by Kelsey and Kohno at EUROCRYPT 2006. It requires the attacker to commit to a hash value y of an iterated hash function H. Subsequently, upon being provided with a message prefix P, the adversary’s task is to identify a suffix S such that H(P∥S) equals y. Kelsey and Kohno demonstrated a herding attack requiring O(√n · 22n/3) evaluations of the compression function of H, where n represents the output and state size of the hash, placing this attack between preimage attacks and collision searches in terms of complexity. At ASIACRYPT 2022, Benedikt et al. transform Kelsey and Kohno’s attack into a quantum variant, decreasing the time complexity from O(√n · 22n/3) to O( 3√n · 23n/7). At ToSC 2023, Zhang et al. proposed the first dedicated Nostradamus attack on AES-like hashing in both classical and quantum settings. In this paper, we have made revisions to the multi-target technique incorporated into the meet-in-the-middle automatic search framework. This modification leads to a decrease in time complexity during the online linking phase, effectively reducing the overall attack time complexity in both classical and quantum scenarios. Specifically, we can achieve more rounds in the classical setting and reduce the time complexity for the same round in the quantum setting.

Integral, impossible-differential (ID), and zero-correlation (ZC) attacks are three of the most important attacks on block ciphers. However, manually finding these attacks can be a daunting task, which is why automated methods are becoming increasingly important. Most automatic tools regarding integral, ZC, and ID attacks have focused only on finding distinguishers rather than complete attacks. At EUROCRYPT 2023, Hadipour et al. proposed a generic and efficient constraint programming (CP) model based on satisfiability for finding ID, ZC, and integral distinguishers. This new model can be extended to a unified CP model for finding full key recovery attacks. However, it has limitations, including determining the contradiction location beforehand and a cell-wise model unsuitable for weakly aligned ciphers like Ascon and PRESENT. They also deferred developing a CP model for the partial-sum technique in key recovery as future work.In this paper, we enhance Hadipour et al.’s method in several ways. First, we remove the limitation of determining the contradiction location in advance. Second, we show how to extend the distinguisher model to a bit-wise model, considering the internal structure of S-boxes and keeping the model based on satisfiability. Third, we introduce a CP model for the partial-sum technique for the first time. To show the usefulness and versatility of our approach, we apply it to various designs, from strongly aligned ones like ForkSKINNY and QARMAv2 to weakly aligned ones such as Ascon and PRESENT, yielding significantly improved results. To mention a few of our results, we improve the integral distinguisher of QARMAv2-128 (resp. QARMAv2-64) by 7 (resp. 5) rounds, and the integral distinguisher of ForkSKINNY by 1 round, only thanks to our cell-wise distinguisher modelings. By using our new bit-wise modeling, our tool can find a group of 2155 5-round ID and ZC distinguishers for Ascon in only one run, taking a few minutes on a regular laptop. The new CP model for the partial-sum technique enhances integral attacks on all SKINNY variants, notably improving the best attack on SKINNY-n-n in the single-key setting by 1 round. We also enhance ID attacks on ForkSKINNY and provide the first analysis of this cipher in a limited reduced-round setting. Our methods are generic and applicable to other block ciphers.

Recently, there has been a surge of interest in the security of authenticated encryption with associated data (AEAD) within the context of key commitment frameworks. Security within this framework ensures that a ciphertext chosen by an adversary does not decrypt to two different sets of key, nonce, and associated data. Despite this increasing interest, the security of several widely deployed AEAD schemes has not been thoroughly examined within this framework. In this work, we assess the key committing security of several AEAD schemes. First, the AEGIS family, which emerged as a winner in the Competition for Authenticated Encryption: Security, Applicability, and Robustness (CAESAR), and has been proposed to standardization at the IETF. A now outdated version of the draft standard suggested that AEGIS could qualify as a fully committing AEAD scheme; we prove that it is not the case by proposing a novel attack applicable to all variants, which has been experimentally verified. We also exhibit a key committing attack on Rocca-S. Our attacks are executed within the FROB game setting, which is known to be one of the most stringent key committing frameworks. This implies that they remain valid in other, more relaxed frameworks, such as CMT-1, CMT-4, and so forth. Finally, we show that applying the same attack techniques to Rocca and Tiaoxin-346 does not compromise their key-committing security. This observation provides valuable insights into the design of such secure round update functions for AES-based AEAD schemes.

It is known that the sponge construction is tightly indifferentiable from a random oracle up to around 2c/2 queries, where c is the capacity. In particular, it cannot provide generic security better than half of the underlying permutation size. In this paper, we aim to achieve hash function security beating this barrier. We present a hashing mode based on two b-bit permutations named the double sponge. The double sponge can be seen as the sponge embedded within the double block length hashing paradigm, making two permutation calls in parallel interleaved with an efficient mixing function. Similarly to the sponge, the permutation size is split as b = r+c, and the underlying compression function absorbs r bits at a time. We prove that the double sponge is indifferentiable from a random oracle up to around 22c/3 queries. This means that the double sponge achieves security beyond the birthday bound in the capacity. In addition, if c > 3b/4, the double sponge beats the birthday bound in the primitive size, to our knowledge being the first hashing mode based on a permutation that accomplices this feature.

DCT is a beyond-birthday-bound (BBB) deterministic authenticated encryption (DAE) mode proposed by Forler et al. in ACISP 2016, ensuring integrity by redundancy. The instantiation of DCT employs the BRW polynomial, which is more efficient than the usual polynomial in GCM by reducing half of the multiplication operations. However, we show that DCT suffers from a small stretch problem similar to GCM. When the stretch length τ is small, choosing a special m-block message, we can reduce the number of queries required by a successful forgery to O(2τ/m). We emphasize that this attack efficiently balances space and time complexity but does not contradict the security bounds of DCT. Finally, we propose an improved scheme named Robust DCT (RDCT) with a minor change to DCT, which improves the security when τ is small and makes it resist the above attack.

For Arithmetization-Oriented ciphers and hash functions Gröbner basis attacks are generally considered as the most competitive attack vector. Unfortunately, the complexity of Gröbner basis algorithms is only understood for special cases, and it is needless to say that these cases do not apply to most cryptographic polynomial systems. Therefore, cryptographers have to resort to experiments, extrapolations and hypotheses to assess the security of their designs. One established measure to quantify the complexity of linear algebra-based Gröbner basis algorithms is the so-called solving degree. Caminata & Gorla revealed that under a certain genericity condition on a polynomial system the solving degree is always upper bounded by the Castelnuovo-Mumford regularity and henceforth by the Macaulay bound, which only takes the degrees and number of variables of the input polynomials into account. In this paper we extend their framework to iterated polynomial systems, the standard polynomial model for symmetric ciphers and hash functions. In particular, we prove solving degree bounds for various attacks on MiMC, Feistel-MiMC, Feistel-MiMC-Hash, Hades and GMiMC. Our bounds fall in line with the hypothesized complexity of Gröbner basis attacks on these designs, and to the best of our knowledge this is the first time that a mathematical proof for these complexities is provided. Moreover, by studying polynomials with degree falls we can prove lower bounds on the Castelnuovo-Mumford regularity for attacks on MiMC, Feistel-MiMC and Feistel-MiMCHash provided that only a few solutions of the corresponding iterated polynomial system originate from the base field. Hence, regularity-based solving degree estimations can never surpass a certain threshold, a desirable property for cryptographic polynomial systems.

Lightweight cryptographic constructions are often optimized on multiple aspects that put the security bounds to the limit. In this respect, it is important to obtain security bounds that are tight and give an accurate and exact indication of the generic security. However, whereas for black-box security bounds it has become common practice to argue tightness of security bounds, for leakage resilience security bounds this is not the case. This is unfortunate, as for leakage resilience results, tightness is even more important as there is already a lossiness incurred in capturing the actual leakage by a theoretical model in the first place.In this work, we consider the SuKS (Suffix Keyed Sponge) PRF construction and investigate tightness of the leakage resilience bound of Dobraunig and Mennink (ToSC 2019). We observe that, although their black-box security result is tight, their leakage resilience bound is not tight in their bounded leakage term λ. We observe that this is caused by the fact that parts of the security bound contain a term covering multicollisions and a term covering leakage, but an adversary is unable to combine both. We next consider improved security of the SuKS for two types of leakage: fixed position leakage, where the adversary directly learns the value of λ bits of a secret state, and Hamming weight leakage, where the Hamming weight of a fixed part of the state is leaked. For fixed position leakage, a very generous form of bounded leakage, we improve the original bound by making wise use of the multicollision limit function of Daemen et al. (ASIACRYPT 2017). For the more realistic setting of Hamming weight leakage, we structurally revisit the multicollision limit function analysis by including Hamming weight in the computation, a problem that is difficult on its own due to the non-uniform character of this type of leakage. In both cases, we improve and tighten the leakage resilience bound of Dobraunig and Mennink. The improved bound for the SuKS has immediate consequences for the leakage resilience of the NIST lightweight cryptography competition finalist ISAP v2, an authenticated encryption scheme that uses the SuKS internally.

A deterministic random bit generator (DRBG) generates pseudorandom bits from an unpredictable seed, i.e., a seed drawn from any random source with sufficient entropy. The current paper formalizes a security notion for a DRBG, in which an attacker may make any legal sequence of requests to the DRBG and sometimes compromise the DRBG state, but should still not be able to distingush DRBG outputs from ideal random bits. The paper proposes XDRBG, a new DRBG based on any eXtendable Output Function (XOF) and proves the security of the XDRBG in the ideal-XOF model. The proven bounds are tight, as demonstrated by matching attacks. The paper also discusses the security of XDRBG against quantum attackers. Finally, the paper proposes concrete instantiations of XDRBG, employing either the SHAKE128 or the SHAKE256 XDRBG. Alternative instantiations suitable for lightweight applications can be based on ASCON.