IACR News
Updates on the COVID-19 situation are on the
Announcement channel.
Here you can see all recent updates to the IACR webpage. These updates are also available:
18 July 2023
Szczecin, Poland, 24 June - 26 June 2024
Event date: 24 June to 26 June 2024
17 July 2023
Lichao Wu, Amir Ali-pour, Azade Rezaeezade, Guilherme Perin, Stjepan Picek
Profiling side-channel analysis has gained widespread acceptance in both academic and industrial realms due to its robust capacity to unveil protected secrets, even in the presence of countermeasures. To harness this capability, an adversary must access a clone of the target device to acquire profiling measurements, labeling them with leakage models. The challenge of finding an effective leakage model, especially for a protected dataset with a low signal-to-noise ratio or weak correlation between actual leakages and labels, often necessitates an intuitive engineering approach, as otherwise, the attack will not perform well.
In this paper, we introduce a deep learning approach that does not assume any specific leakage model, referred to as the multibit model. Instead of trying to learn a representation of the target intermediate data (label), we utilize the concept of the stochastic model to decompose the label into bits. Then, the deep learning model is used to classify each bit independently. This versatile multibit model can align with existing leakage models like the Hamming weight and Most Significant Bit leakage models while also possessing the flexibility to adapt to complex leakage scenarios. To further improve the attack efficiency, we extend the multibit model to simultaneously attack all 16 subkey bytes, which requires negligible computational effort. Based on our preliminary analysis, two of the four considered datasets could only be broken using a Hamming Weight leakage model. Using the same model, the proposed methods can efficiently crack all key bytes across four considered datasets. Our work, thus, signifies a significant step forward in deep learning-based side-channel attacks, showcasing a high degree of flexibility and efficiency without any presumption of the leakage model.
In this paper, we introduce a deep learning approach that does not assume any specific leakage model, referred to as the multibit model. Instead of trying to learn a representation of the target intermediate data (label), we utilize the concept of the stochastic model to decompose the label into bits. Then, the deep learning model is used to classify each bit independently. This versatile multibit model can align with existing leakage models like the Hamming weight and Most Significant Bit leakage models while also possessing the flexibility to adapt to complex leakage scenarios. To further improve the attack efficiency, we extend the multibit model to simultaneously attack all 16 subkey bytes, which requires negligible computational effort. Based on our preliminary analysis, two of the four considered datasets could only be broken using a Hamming Weight leakage model. Using the same model, the proposed methods can efficiently crack all key bytes across four considered datasets. Our work, thus, signifies a significant step forward in deep learning-based side-channel attacks, showcasing a high degree of flexibility and efficiency without any presumption of the leakage model.
Lichao Wu, Sébastien Tiran, Guilherme Perin, Stjepan Picek
Side-channel Collision Attacks (SCCA) constitute a subset of non-profiling attacks that exploit information dependency leaked during cryptographic operations. Unlike traditional collision attacks, which seek instances where two different inputs to a cryptographic algorithm yield identical outputs, SCCAs specifically target the internal state, where identical outputs are more likely. In CHES 2023, Staib et al. presented a Deep Learning-based SCCA (DL-SCCA), which enhanced the attack performance while decreasing the required effort for leakage preprocessing. Nevertheless, this method inherits the conventional SCCA's limitations, as it operates on trace segments reflecting the target operation explicitly, leading to issues such as portability and low tolerance to errors.
This paper introduces an end-to-end plaintext-based SCCA to address these challenges. We leverage the bijective relationship between plaintext and secret data to label the leakage measurement with known information, then learn plaintext-based profiling models to depict leakages from varying operations. By comparing the leakage representations produced by the profiling model, an adversary can reveal the key difference. As an end-to-end approach, we propose an error correction scheme to rectify false predictions. Experimental results indicate our approach significantly surpasses DL-SCCA in terms of attack performance (e.g., success rate increased from 53\% to 100\%) and computational complexity (training time reduced from approximately 2 hours to 10 minutes). These findings underscore our method's effectiveness and practicality in real-world attack scenarios.
This paper introduces an end-to-end plaintext-based SCCA to address these challenges. We leverage the bijective relationship between plaintext and secret data to label the leakage measurement with known information, then learn plaintext-based profiling models to depict leakages from varying operations. By comparing the leakage representations produced by the profiling model, an adversary can reveal the key difference. As an end-to-end approach, we propose an error correction scheme to rectify false predictions. Experimental results indicate our approach significantly surpasses DL-SCCA in terms of attack performance (e.g., success rate increased from 53\% to 100\%) and computational complexity (training time reduced from approximately 2 hours to 10 minutes). These findings underscore our method's effectiveness and practicality in real-world attack scenarios.
Sengim Karayalcin, Marina Krcek, Lichao Wu, Stjepan Picek, Guilherme Perin
Profiling side-channel analysis is an essential technique to assess the security of protected cryptographic implementations by subjecting them to the worst-case security analysis. This approach assumes the presence of a highly capable adversary with knowledge of countermeasures and randomness employed by the target device. However, black-box profiling attacks are commonly employed when aiming to emulate real-world scenarios. These attacks leverage deep learning as a prominent alternative since deep neural networks can automatically select points of interest, eliminating the need for secret mask knowledge.
Nevertheless, black-box profiling attacks often result in non-worst-case security evaluations, leading to suboptimal profiling models.
In this study, we propose modifying the conventional black-box threat model by incorporating a new assumption: the adversary possesses a similar implementation that can be used as a white-box reference design. We create an adversarial dataset by extracting features or points of interest from this reference design. These features are then utilized for training a novel conditional generative adversarial network (CGAN) framework, enabling a generative model to extract features from high-order leakages in protected implementation without any assumptions about the masking scheme or secret masks. Our framework empowers attackers to perform efficient black-box profiling attack that achieves (and even surpasses) the performance of the worst-case security assessments.
In this study, we propose modifying the conventional black-box threat model by incorporating a new assumption: the adversary possesses a similar implementation that can be used as a white-box reference design. We create an adversarial dataset by extracting features or points of interest from this reference design. These features are then utilized for training a novel conditional generative adversarial network (CGAN) framework, enabling a generative model to extract features from high-order leakages in protected implementation without any assumptions about the masking scheme or secret masks. Our framework empowers attackers to perform efficient black-box profiling attack that achieves (and even surpasses) the performance of the worst-case security assessments.
Duygu Ozden, Oğuz Yayla
Verifiable timed commitments serve as cryptographic tools that enable the binding of information to specific time intervals. By integrating these commitments into signature schemes, secure and tamper-evident digital signatures can be generated, ensuring the integrity of time-sensitive mechanisms. This article delves into the concept of verifiable timed commitments and explores their efficient applications in digital signature constructions. Specifically, it focuses on two important signature schemes: proxy signatures and multi-signatures. The idea of the timed proxy signature is to enable the delegation of signing rights for a specified period, allowing designated entities to sign messages on behalf of the original signer. On the other hand, multi-signatures allow multiple parties to collectively generate a single signature, ensuring enhanced security and accountability. The article presents an in-depth analysis of the underlying mechanisms, discussing their properties, strengths, and computational complexity. Through this exploration, the article aims to shed light on the potential of verifiable timed commitments and inspire further research in this evolving field of cryptography.
Liam Eagen, Ariel Gabizon
We continue the recent line of work on folding schemes. Building on ideas from ProtoStar [BC23] we construct a folding scheme where the recursive verifier's ``marginal work'', beyond linearly combining witness commitments, consists only of a logarithmic number of field operations and a constant number of hashes. Moreover, our folding scheme performs well when \emph{folding multiple instances at one step}, in which case the marginal number of verifier field operations per instance becomes constant, assuming constant degree gates.
Tung Le, Thang Hoang
Commodity encrypted storage platforms (e.g., IceDrive, pCloud) permit data store and sharing across multiple users while preserving data confidentiality. However, end-to-end encryption may not be sufficient since it only offers confidentiality when the data is at rest or in transit. Meanwhile, sensitive information can be leaked from metadata representing activities during data operations (e.g., query, processing). Recent encrypted search platforms such as DORY (OSDI’20) or DURASIFT (WPES’19) permit multi-user data query functionalities, while protecting metadata privacy. However, they either incur a high processing overhead or offer limited secu- rity/functionality, and require strong trust assumptions.
We propose MAPLE, a new metadata-hiding encrypted search platform that offers query functionalities (search, update) on the shared data across multiple users with complex policy controls. MAPLE protects metadata privacy all the time during query processing, while achieving significantly (asymptotically) lower processing overhead than state-of-the-art platforms. The core technique of MAPLE is the design of oblivious data structures for search index and access control coupled with secure computation techniques to enable efficient query processing with a minimal trust. We fully implemented MAPLE and evaluated its performance on commodity cloud (Amazon EC2) under real settings. Experimental results showed that MAPLE achieved a concrete performance comparable with its counterparts, while offering provably stronger security guarantees and more diverse functionalities.
Sushmit Jahan Rose, Umut Ozkaya, Sabina Yasmin, Suraiya Jabin, Robiul Hasan, Elias Kabir
Data security and secrecy from unwanted applications are the subjects of the science known as cryptography. The advanced encryption standard algorithm is the most used and secure algorithm to encrypt data. The AES algorithm is based on the symmetric algorithm and uses a single key to encrypt and decrypt data. The AES algorithm uses 128 bits length of plain text with 128 bits, 192 bits, or 256 bits key size to encrypt data. Latin script uses ASCII codes, and a single byte represents each alphabet. Therefore, in 128 bits AES encryption, 16 characters can be encrypted each time. The other language script used the Unicode standard to represent their alphabets. In Unicode, at least 2 bytes are required to represent a character. Therefore, eight characters can be encrypted at a time. Also, there is no available S-box for Unicode characters. Therefore, a modified algorithm is proposed for Unicode to encrypt data. To use the AES algorithm in Unicode data, we need to convert the Unicode into character encoding, such as UTF-16. Nevertheless, In UTF-16, some Unicode characters have similar recurrent values. This paper demonstrates a modified AES algorithm to encrypt the Unicode script to reduce time complexity.
Orestis Alpos, Christian Cachin, Simon Holmgaard Kamp, Jesper Buus Nielsen
We present simple and practical protocols for generating randomness as used by asynchronous total-order broadcast. The protocols are secure in a proof-of-stake setting with dynamically changing stake. They can be plugged into existing protocols for asynchronous total-order broadcast and will turn these into asynchronous total-order broadcast with dynamic stake. Our contribution relies on two important techniques. The paper ``Random Oracles in Constantinople: Practical Asynchronous Byzantine Agreement using Cryptography'' [Cachin, Kursawe, and Shoup, PODC 2000] has influenced the design of practical total-order broadcast through its use of threshold cryptography. However, it needs a setup protocol to be efficient. In a proof-of-stake setting with dynamic stake this setup would have to be continually recomputed, making the protocol impractical. The work ``Asynchronous Byzantine Agreement with Subquadratic Communication'' [Blum, Katz, Liu-Zhang, and Loss, TCC 2020] showed how to use an initial setup for broadcast to asymptotically efficiently generate sub-sequent setups. The protocol, however, resorted to fully homomorphic encryption and was therefore not practically efficient. We adopt their approach to the proof-of-stake setting with dynamic stake, apply it to the Constantinople paper, and remove the need for fully homomorphic encryption. This results in simple and practical proof-of-stake protocols. We discuss how to use the new coin-flip protocols together with DAG rider [Keidar et al., PODC 2021] and create a variant which works for dynamic proof of stake. Our method can be employed together with many further asynchronous total-order broadcast protocols.
Kristian Gjøsteen, Thomas Haines, Morten Rotvold Solberg
An interesting approach to achieving verifiability in voting systems is to make use of tracking numbers. This gives voters a simple way of verifying that their ballot was counted: they can simply look up their ballot/tracker pair on a public bulletin board. It is crucial to understand how trackers affect other security properties, in particular privacy. However, existing privacy definitions are not designed to accommodate tracker-based voting systems. Furthermore, the addition of trackers increases the threat of coercion. There does however exist techniques to mitigate the coercion threat. While the term coercion mitigation has been used in the literature when describing voting systems such as Selene, no formal definition of coercion mitigation seems to exist. In this paper we formally define what coercion mitigation means for tracker-based voting systems. We model Selene in our framework and we prove that Selene provides coercion mitigation, in addition to privacy and verifiability.
Agnese Gini, Pierrick Méaux
We investigate the concept of $\mathcal{S}_0$-equivalent class, $n$-variable Boolean functions up to the addition of a symmetric function null in $0_n$ and $1_n$, as a tool to study weightwise perfectly balanced functions.
On the one hand we show that weightwise properties, such as being weightwise perfectly balanced, the weightwise nonlinearity and weightwise algebraic immunity, are invariants of these classes.
On the other hand we analyze the variation of global parameters inside the same class, showing for example that there is always a function with high degree, algebraic immunity, or nonlinearity in the $\mathcal{S}_0$-equivalent class of a function.
Finally, we discuss how these results extend to other equivalence relations and their applications in cryptography.
Marina Krček, Lichao Wu, Guilherme Perin, Stjepan Picek
Convolutional neural networks (CNNs) offer unrivaled performance in profiling side-channel analysis. This claim is corroborated by numerous results where CNNs break targets protected with masking and hiding countermeasures. One hiding countermeasure is commonly investigated in related works - desynchronization (misalignment). The conclusions usually state that CNNs can break desynchronization as they are shift-invariant. This paper investigates that claim in more detail and reveals that the situation is more complex. While CNNs have certain shift-invariance, it is insufficient for commonly encountered scenarios in deep learning-based side-channel analysis.
We propose to use data augmentation to improve the shift-invariance and, in a more powerful version, ensembles of data augmentation. Our results show the proposed techniques work very well and improve the attack significantly, even for an order of magnitude.
Jean-Luc Beuchat, Valon Rexhepi
The Swiss law on electronic identity (LSIE) was rejected on March 7, 2021. Its opponents accused it of involving private companies which could thus collect citizens' data and store them centrally. Six motions with identical wording were tabled on March 10, 2021: they all ask the Swiss Federal Council to set up a state-run system allowing citizens to prove their identity online in complete confidence. They stipulate that only necessary information is collected and stored in a decentralized manner. The Swiss Federal Council has recommended to Parliament to approve these motions on May 26, 2021, and wishes to propose a new e-ID solution responding to citizens' concerns as soon as possible. The Federal Department of Justice and Police has been asked to draw up a first draft presenting several technical solutions and specifying their respective costs. Following the publication of a working document on September 2, 2021, a public consultation was opened. It ended on October 14, 2021, with a public debate organized at the Government House in Bern and broadcasted live on a virtual platform. Self-Sovereign Identity (SSI) is one of the solutions identified during this process. It gives the citizens control of their electronic identity: they hold credentials issued by public administrations and choose the data they wish to disclose when they authenticate with a service (they can for example prove that they are over 18 without specifying their exact date of birth).
We propose here a decentralized and user-centric e-ID system based on SSI principles. Our solution embraces an open-source philosophy, fostering transparency and community involvement. We employ blockchain technology as a design pattern to establish trust and ensure the immutability of identity-related data. By design, our solution ensures the right to be forgotten by exclusively storing the digests of verifiable credentials on the blockchain. To demonstrate the feasibility and effectiveness of our SSI solution, we have developed a proof of concept leveraging the Partisia blockchain.
Banashri Karmakar, Nishat Koti, Arpita Patra, Sikhar Patranabis, Protik Paul, Divya Ravi
Secure multiparty computation (MPC) enables privacy-preserving collaborative computation over sensitive data held by multiple mutually distrusting parties. Unfortunately, in the most natural setting where a majority of the parties are maliciously corrupt (also called the $\textit{dishonest majority}$ setting), traditional MPC protocols incur high overheads and offer weaker security guarantees than are desirable for practical applications. In this paper, we explore the possibility of circumventing these drawbacks and achieving practically efficient dishonest majority MPC protocols with strong security guarantees by assuming an additional semi-honest, non-colluding helper party $\textsf{HP}$. We believe that this is a more realistic alternative to assuming an honest majority, since many real-world applications of MPC involving potentially large numbers of parties (such as secure auctions and dark pools) are typically enabled by a central entity that can be modeled as the $\textsf{HP}$.
In the above model, we are the first to design, implement and benchmark a practically-efficient and general multi-party framework, $\textsf{Asterisk}$, which achieves the strong security guarantee of $\textit{fairness}$ (either all parties learn the output or none do), scales to hundreds of parties, outperforms all existing dishonest majority MPC protocols, and is, in fact, competitive with state-of-the-art honest majority MPC protocols. Our experiments show that $\textsf{Asterisk}$ achieves $900-1200\times$ speedup in preprocessing as compared to the best dishonest majority MPC protocols, and supports $100$-party evaluation of a circuit with $10^6$ multiplication gates in under $2$ minutes. We also implement and benchmark practically efficient and highly scalable instances of two applications, namely privacy-preserving secure auctions and dark pools, using $\textsf{Asterisk}$ as the building block. This showcases the effectiveness of $\textsf{Asterisk}$ in enabling real-world privacy-preserving applications with strong efficiency and security guarantees.
In the above model, we are the first to design, implement and benchmark a practically-efficient and general multi-party framework, $\textsf{Asterisk}$, which achieves the strong security guarantee of $\textit{fairness}$ (either all parties learn the output or none do), scales to hundreds of parties, outperforms all existing dishonest majority MPC protocols, and is, in fact, competitive with state-of-the-art honest majority MPC protocols. Our experiments show that $\textsf{Asterisk}$ achieves $900-1200\times$ speedup in preprocessing as compared to the best dishonest majority MPC protocols, and supports $100$-party evaluation of a circuit with $10^6$ multiplication gates in under $2$ minutes. We also implement and benchmark practically efficient and highly scalable instances of two applications, namely privacy-preserving secure auctions and dark pools, using $\textsf{Asterisk}$ as the building block. This showcases the effectiveness of $\textsf{Asterisk}$ in enabling real-world privacy-preserving applications with strong efficiency and security guarantees.
Mark Zhandry
We give a candidate construction of public key quantum money, and even a strengthened version called quantum lightning, from abelian group actions, which can in turn be constructed from suitable isogenies over elliptic curves. We prove our scheme is secure under new but plausible assumptions on suitable group actions.
Hatice Kübra Güner, Ceyda Mangır, Oğuz Yayla
Protecting secret keys from malicious observers in untrusted environments is a critical security issue. White-box cryptography suggests software protection by hiding the key in the white-box setting. One method for hiding the key in the cipher code is through encoding methods. Unfortunately, encoding methods may be vulnerable to algebraic attacks and side-channel analysis. Another technique to hide the key is (M,Z)-space hardness approach that conceals the key into a large lookup table generated with a reliable small block cipher. In (M,Z)-space-hard algorithms, the key extraction problem in the white-box setting turns into a key recovery problem in the black-box setting. One of the problems for (M,Z)-space-hard algorithms is improving run-time performance. In this study, we aim to improve the run-time performance of the existing white-box implementations. We propose an LS-design based white-box algorithm with better run-rime performance than space-hard SPNbox algorithm. Moreover, an LS-design based table creation method is designed. When we compare the run-time performance of our method with the SPNbox algorithm, we obtain 28% improvement for white-box implementation and 27% for black-box implementation for 128-bit block size. The LS-design based method is also used for 256-bit block size in the white-box setting.
Xiaoyang Dong, Shun Li, Phuong Pham
At CRYPTO 2020, Liu et al. find that many differentials on Gimli are actually incompatible. On the related-key differential of AES, the incompatibilities also exist and are handled in different ad-hoc ways by adding respective constraints into the searching models. However, such an ad-hoc method is insufficient to rule out all the incompatibilities and may still output false positive related-key differentials. At CRYPTO 2022, a new approach combining a Constraint Programming (CP) tool and a triangulation algorithm to search for rebound attacks against AES- like hashing was proposed. In this paper, we combine and extend these techniques to create a uniform related-key differential search model, which can not only generate the related-key differentials on AES and similar ciphers but also immediately verify the existence of at least one key pair fulfilling the differentials. With the innovative automatic tool, we find new related-key differentials on full-round AES-192, AES-256, Kiasu-BC, and round-reduced Deoxys-BC. Based on these findings, full- round limited-birthday chosen-key distinguishing attacks on AES-192, AES-256, and Kiasu-BC are presented, as well as the first chosen-key dis- tinguisher on reduced Deoxys-BC. Furthermore, a limited-birthday dis- tinguisher on 9-round Kiasu-BC with practical complexities is found for the first time.
Jonathan Katz
Protocols for distributed key generation (DKG) in the discrete-logarithm setting have received a tremendous amount of attention in the past few years. Several synchronous DKG protocols have been proposed, but most such protocols are either not fully secure (in the sense of simulatability) or are not robust in that they allow even a single malicious party to prevent successful generation of a key.
In this paper we explore the round complexity of (robust) DKG in the honest-majority setting where robust DKG is feasible. On the negative side, we show the impossibility of one-round (robust) DKG protocols regardless of any prior setup the parties have. On the positive side, we show various two-round---and hence, round-optimal---protocols for robust DKG offering tradeoffs in terms of their efficiency, necessary setup, and required assumptions.
In this paper we explore the round complexity of (robust) DKG in the honest-majority setting where robust DKG is feasible. On the negative side, we show the impossibility of one-round (robust) DKG protocols regardless of any prior setup the parties have. On the positive side, we show various two-round---and hence, round-optimal---protocols for robust DKG offering tradeoffs in terms of their efficiency, necessary setup, and required assumptions.
16 July 2023
Alessandro Budroni, Jesús-Javier Chi-Domínguez, Mukul Kulkarni
Group actions have been used as a foundation in Public-key Cryptography to provide a framework for hard problems and assumptions. In this work we formalize the Lattice Isomorphism Problem (LIP) within the context of cryptographic group actions. Our main result shows that a quadratic number of queries to a randomized oracle outputting LIP instances sharing the same secret is enough for inverting the group action in polynomial time. We use this result to uncover a family of weak isomorphisms and to derive two new hard problems on quadratic forms equivalent to LIP for the case of lattices with trivial automorphism.
Tomoki Moriya
Isogeny-based cryptography is one of the candidates for post-quantum cryptography. In 2023, Kani's theorem breaks some isogeny-based schemes including SIDH, which was considered as a promising post-quantum scheme. Though Kani's theorem damaged isogeny-based cryptography, some researchers try to dig into the applications of Kani's theorem. FESTA is an isogeny-based trapdoor function that is one trial to apply Kani's theorem to cryptography.
In this paper, we provide an adaptive attack for a possible PKE scheme based on FESTA trapdoor functions. Our attack reveals the secret key of the function. Our attack may be used if the recipient of the PKE scheme does not check whether the hidden matrix corresponding to the ciphertext is correct. In other words, the recipient can prevent our attack by checking the correctness of the matrix.
In this paper, we provide an adaptive attack for a possible PKE scheme based on FESTA trapdoor functions. Our attack reveals the secret key of the function. Our attack may be used if the recipient of the PKE scheme does not check whether the hidden matrix corresponding to the ciphertext is correct. In other words, the recipient can prevent our attack by checking the correctness of the matrix.