Elena Dubrova
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Elena Dubrova has a total of 43 co-authors. Here are the top co-authors by number of shared publications:
- Kalle Ngo
KTH Royal Institute of Technology
6 shared publications - Sönke Jendral
KTH Royal Institute of Technology, Stockholm, Sweden
6 shared publications - Ruize Wang
KTH Royal Institute of Technology
5 shared publications - Michail Moraitis
Royal Institute of Technology (KTH)
4 shared publications - Gunnar Carlsson
Development Unit Radio, Ericsson AB, Stockholm, Sweden
3 shared publications - Joel Gärtner
KTH Royal Institute of Technology
3 shared publications
Latest Publications
- Open Access
- 14 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 37-45
Single-Trace SAT-Assisted Side-Channel Attack on MILENAGE
- 14 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 37-45
Open AccessSingle-Trace SAT-Assisted Side-Channel Attack on MILENAGE
- Open Access
- 14 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Understanding the Impact of Compiler Optimization Flags on Side-Channel Leakage: A Case Study of ML-KEM Inverse NTT
- 14 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Open AccessUnderstanding the Impact of Compiler Optimization Flags on Side-Channel Leakage: A Case Study of ML-KEM Inverse NTT
- Open Access
- 26 October 2026
- Published by International Association for Cryptologic Research in IACR Communications in Cryptology
- Vol. 3 (3)
Beyond Microcontrollers: End-to-End Full Decryption Oracle Power Attacks of HQC on FPGAs
The National Institute of Standards and Technology recently selected Hamming Quasi-Cyclic (HQC) as a critical code-based alternative for post-quantum cryptography. Hardware architectures like FPGAs successfully mitigate the severe performance bottlenecks of HQC software deployments. However, their physical security remains largely unexplored. It is often assumed that the massive concurrency and systemic electrical noise of FPGAs provide a natural barrier against side-channel analysis. In this paper, we challenge this assumption. We present the first end-to-end full decryption (FD) style power attack against a parallel FPGA implementation of HQC. Unlike Cortex-M4 software, where leakage in Reed-Muller (RM) decoding manifests as sharp, localized peaks, FPGA signals are inherently weaker and broadly distributed across the second half of the pipelined RM decoding phase. To overcome this challenging signal-to-noise ratio, we refine our methodology. Rather than attempting to predict full RM decoding outputs, we construct a “yes/no” oracle for each individual decoding block. This approach exploits the decoder's parallelism to recover independent chunks of the secret simultaneously. Our practical evaluations demonstrate high efficiency: an adversary can recover the secret key with a 58% success rate using merely 120 FPGA power traces. In this setting, we employ an optimized Soft-ISD solver and strictly bound the post-processing complexity to just 30 minutes on a Mac computer. This work highlights the urgent need to design and integrate robust hardware masking schemes for HQC.
- 26 October 2026
- Published by International Association for Cryptologic Research in IACR Communications in Cryptology
- Vol. 3 (3)
Open AccessBeyond Microcontrollers: End-to-End Full Decryption Oracle Power Attacks of HQC on FPGAs
The National Institute of Standards and Technology recently selected Hamming Quasi-Cyclic (HQC) as a critical code-based alternative for post-quantum cryptography. Hardware architectures like FPGAs successfully mitigate the severe performance bottlenecks of HQC software deployments. However, their physical security remains largely unexplored. It is often assumed that the massive concurrency and systemic electrical noise of FPGAs provide a natural barrier against side-channel analysis. In this paper, we challenge this assumption. We present the first end-to-end full decryption (FD) style power attack against a parallel FPGA implementation of HQC. Unlike Cortex-M4 software, where leakage in Reed-Muller (RM) decoding manifests as sharp, localized peaks, FPGA signals are inherently weaker and broadly distributed across the second half of the pipelined RM decoding phase. To overcome this challenging signal-to-noise ratio, we refine our methodology. Rather than attempting to predict full RM decoding outputs, we construct a “yes/no” oracle for each individual decoding block. This approach exploits the decoder's parallelism to recover independent chunks of the secret simultaneously. Our practical evaluations demonstrate high efficiency: an adversary can recover the secret key with a 58% success rate using merely 120 FPGA power traces. In this setting, we employ an optimized Soft-ISD solver and strictly bound the post-processing complexity to just 30 minutes on a Mac computer. This work highlights the urgent need to design and integrate robust hardware masking schemes for HQC.
- Open Access
- 16 January 2026
- Published by Universitatsbibliothek der Ruhr-Universitat Bochum in IACR Transactions on Cryptographic Hardware and Embedded Systems
- Vol. 2026 (1) , 225-249
Fault Attacks on VOLEitH Signature Schemes
Ongoing efforts to transition to post-quantum public-key cryptosystems have created the need for algorithms with a variety of performance characteristics and security assumptions. Among the candidates in NIST’s post-quantum standardisation process for additional digital signatures are FAEST and SDitH, two Vector Oblivious Linear Evaluation in-the-Head (VOLEitH)-based schemes. The security of FAEST is based on the Advanced Encryption Standard (AES), while SDitH relies on the regular syndrome decoding problem. The VOLEitH paradigm enables competitive performance and signature sizes under conservative security assumptions. However, since it was introduced relatively recently, in 2023, its resistance to physical attacks has not yet been analysed. In this paper, we present the first security analysis of VOLEitH-based signature schemes in the context of fault injection attacks. We demonstrate two practical attacks on the reference implementations of the secondround variants of FAEST and SDitH in ARM Cortex-M4 capable of recovering the secret key from a single signature. These attacks exploit vulnerabilities of components specific to VOLEitH schemes, namely the batch all-but-one vector commitments and the VOLE generation. We also propose countermeasures to mitigate these attacks and enhance the physical security of VOLEitH-based digital signature schemes.
- 16 January 2026
- Published by Universitatsbibliothek der Ruhr-Universitat Bochum in IACR Transactions on Cryptographic Hardware and Embedded Systems
- Vol. 2026 (1) , 225-249
Open AccessFault Attacks on VOLEitH Signature Schemes
Ongoing efforts to transition to post-quantum public-key cryptosystems have created the need for algorithms with a variety of performance characteristics and security assumptions. Among the candidates in NIST’s post-quantum standardisation process for additional digital signatures are FAEST and SDitH, two Vector Oblivious Linear Evaluation in-the-Head (VOLEitH)-based schemes. The security of FAEST is based on the Advanced Encryption Standard (AES), while SDitH relies on the regular syndrome decoding problem. The VOLEitH paradigm enables competitive performance and signature sizes under conservative security assumptions. However, since it was introduced relatively recently, in 2023, its resistance to physical attacks has not yet been analysed. In this paper, we present the first security analysis of VOLEitH-based signature schemes in the context of fault injection attacks. We demonstrate two practical attacks on the reference implementations of the secondround variants of FAEST and SDitH in ARM Cortex-M4 capable of recovering the secret key from a single signature. These attacks exploit vulnerabilities of components specific to VOLEitH schemes, namely the batch all-but-one vector commitments and the VOLE generation. We also propose countermeasures to mitigate these attacks and enhance the physical security of VOLEitH-based digital signature schemes.
- Open Access
- 8 January 2026
- Published by International Association for Cryptologic Research in IACR Communications in Cryptology
- Vol. 2 (4)
Side-Channel Attacks on VOLEitH Signature Schemes
The ongoing transition to post-quantum cryptography has highlighted the need for digital signature schemes offering diverse performance and security trade-offs. Among the candidates in NIST’s ongoing post-quantum signature standardisation process is FAEST, a scheme built upon the Vector Oblivious Linear Evaluation in-the-Head (VOLEitH) paradigm introduced in 2023. VOLEitH enables efficient zero-knowledge proofs with competitive signature sizes under conservative assumptions, allowing FAEST to rely primarily on the one-wayness of the Advanced Encryption Standard (AES). Despite their promising efficiency, VOLEitH-based signature schemes have remained relatively unexplored from a physical security perspective. In this paper, we present the first side-channel security evaluation. Specifically, we demonstrate two single-trace, deep learning-assisted power analysis attacks on the masked implementation of FAEST by Aranha, Degn, Eilath, Nielsen, and Scholl. These attacks exploit leakage from witness bits and VOLE tag computations, recovering the full secret key with success probability above 0.99 from a single signature on an ARM Cortex-M4 processor. We further analyse how the VOLEitH construction enables profiling of VOLE tags without knowledge of the secret key and how even partial leakage of these tags compromises security. Finally, we discuss practical countermeasures to mitigate such leakages and strengthen the physical resilience of VOLEitH-based signature implementations.
- 8 January 2026
- Published by International Association for Cryptologic Research in IACR Communications in Cryptology
- Vol. 2 (4)
Open AccessSide-Channel Attacks on VOLEitH Signature Schemes
The ongoing transition to post-quantum cryptography has highlighted the need for digital signature schemes offering diverse performance and security trade-offs. Among the candidates in NIST’s ongoing post-quantum signature standardisation process is FAEST, a scheme built upon the Vector Oblivious Linear Evaluation in-the-Head (VOLEitH) paradigm introduced in 2023. VOLEitH enables efficient zero-knowledge proofs with competitive signature sizes under conservative assumptions, allowing FAEST to rely primarily on the one-wayness of the Advanced Encryption Standard (AES). Despite their promising efficiency, VOLEitH-based signature schemes have remained relatively unexplored from a physical security perspective. In this paper, we present the first side-channel security evaluation. Specifically, we demonstrate two single-trace, deep learning-assisted power analysis attacks on the masked implementation of FAEST by Aranha, Degn, Eilath, Nielsen, and Scholl. These attacks exploit leakage from witness bits and VOLE tag computations, recovering the full secret key with success probability above 0.99 from a single signature on an ARM Cortex-M4 processor. We further analyse how the VOLEitH construction enables profiling of VOLE tags without knowledge of the secret key and how even partial leakage of these tags compromises security. Finally, we discuss practical countermeasures to mitigate such leakages and strengthen the physical resilience of VOLEitH-based signature implementations.
- Open Access
- 13 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 21-30
Single-trace Side-Channel Attacks on MAYO Exploiting Leaky Modular Multiplication
- 13 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 21-30
Open AccessSingle-trace Side-Channel Attacks on MAYO Exploiting Leaky Modular Multiplication
- Open Access
- 2 April 2025
- Published by Springer Nature in Journal of Cryptographic Engineering
- Vol. 15 (1) , 1-23
A side-channel attack on a masked hardware implementation of CRYSTALS-Kyber
NIST has recently selected CRYSTALS-Kyber as a new public key encryption and key establishment algorithm to be standardized. This makes it important to evaluate the resistance of CRYSTALS-Kyber implementations to side-channel attacks. Software implementations of CRYSTALS-Kyber have already been thoroughly analysed. The discovered vulnerabilities have helped improve subsequently released versions and promoted stronger countermeasures against side-channel attacks. In this paper, we present the first attack on a protected hardware implementation of CRYSTALS-Kyber. We demonstrate a practical message (shared key) recovery attack on the first-order masked FPGA implementation of Kyber-512 by Kamucheka et al. (2022) using power analysis based on the Hamming distance leakage model. The presented attack exploits a vulnerability located in the masked message decoding function executed during the decryption step of decapsulation. The message recovery is performed using a profiled deep learning-assisted method which extracts the message directly, without explicitly retrieving each share. By repeating the same decapsulation multiple times, it is possible to increase the success rate of full shared key recovery to 99%. We also analyse the feasibility of recovering shared keys during encapsulation and propose a countermeasure against the presented attack that is also applicable to FPGA implementations of other cryptographic algorithms.
- 2 April 2025
- Published by Springer Nature in Journal of Cryptographic Engineering
- Vol. 15 (1) , 1-23
Open AccessA side-channel attack on a masked hardware implementation of CRYSTALS-Kyber
NIST has recently selected CRYSTALS-Kyber as a new public key encryption and key establishment algorithm to be standardized. This makes it important to evaluate the resistance of CRYSTALS-Kyber implementations to side-channel attacks. Software implementations of CRYSTALS-Kyber have already been thoroughly analysed. The discovered vulnerabilities have helped improve subsequently released versions and promoted stronger countermeasures against side-channel attacks. In this paper, we present the first attack on a protected hardware implementation of CRYSTALS-Kyber. We demonstrate a practical message (shared key) recovery attack on the first-order masked FPGA implementation of Kyber-512 by Kamucheka et al. (2022) using power analysis based on the Hamming distance leakage model. The presented attack exploits a vulnerability located in the masked message decoding function executed during the decryption step of decapsulation. The message recovery is performed using a profiled deep learning-assisted method which extracts the message directly, without explicitly retrieving each share. By repeating the same decapsulation multiple times, it is possible to increase the success rate of full shared key recovery to 99%. We also analyse the feasibility of recovering shared keys during encapsulation and propose a countermeasure against the presented attack that is also applicable to FPGA implementations of other cryptographic algorithms.
- Open Access
- 13 January 2025
- Published by International Association for Cryptologic Research in IACR Communications in Cryptology
- Vol. 1 (4)
MAYO Key Recovery by Fixing Vinegar Seeds
As the industry prepares for the transition to post-quantum secure public key cryptographic algorithms, vulnerability analysis of their implementations is gaining importance. A theoretically secure cryptographic algorithm should also be able to withstand the challenges of physical attacks in real-world environments. MAYO is a candidate in the ongoing second round of the NIST post-quantum standardization process for selecting additional digital signature schemes. This paper demonstrates three first-order single-execution fault injection attacks on the official MAYO implementation on the ARM Cortex-M4. By using voltage glitching to disrupt the computation of the vinegar seed during the signature generation, we enable the recovery of the secret key directly from the faulty signatures. Our experimental results show that the success rates of the fault attacks in a single execution are 36%, 82%, and 99%, respectively. They emphasize the importance of developing countermeasures against fault attacks prior to the widespread deployment of post-quantum algorithms like MAYO.
- 13 January 2025
- Published by International Association for Cryptologic Research in IACR Communications in Cryptology
- Vol. 1 (4)
Open AccessMAYO Key Recovery by Fixing Vinegar Seeds
As the industry prepares for the transition to post-quantum secure public key cryptographic algorithms, vulnerability analysis of their implementations is gaining importance. A theoretically secure cryptographic algorithm should also be able to withstand the challenges of physical attacks in real-world environments. MAYO is a candidate in the ongoing second round of the NIST post-quantum standardization process for selecting additional digital signature schemes. This paper demonstrates three first-order single-execution fault injection attacks on the official MAYO implementation on the ARM Cortex-M4. By using voltage glitching to disrupt the computation of the vinegar seed during the signature generation, we enable the recovery of the secret key directly from the faulty signatures. Our experimental results show that the success rates of the fault attacks in a single execution are 36%, 82%, and 99%, respectively. They emphasize the importance of developing countermeasures against fault attacks prior to the widespread deployment of post-quantum algorithms like MAYO.
- Open Access
- 7 October 2024
- Published by International Association for Cryptologic Research in IACR Communications in Cryptology
- Vol. 1 (3)
Unpacking Needs Protection
Most of the previous attacks on Dilithium exploit side-channel information which is leaked during the computation of the polynomial multiplication cs1, where s1 is a small-norm secret and c is a verifier's challenge. In this paper, we present a new attack utilizing leakage during secret key unpacking in the signing algorithm. The unpacking is also used in other post-quantum cryptographic algorithms, including Kyber, because inputs and outputs of their API functions are byte arrays. Exploiting leakage during unpacking is more challenging than exploiting leakage during the computation of cs1 since c varies for each signing, while the unpacked secret key remains constant. Therefore, post-processing is required in the latter case to recover a full secret key. We present two variants of post-processing. In the first one, a half of the coefficients of the secret s1 and the error s2 is recovered by profiled deep learning-assisted power analysis and the rest is derived by solving linear equations based on t = As1 + s2, where A and t are parts of the public key. This case assumes knowledge of the least significant bits of t, t0. The second variant uses lattice reduction to derive s1 without the knowledge of t0. However, it needs a larger portion of s1 to be recovered by power analysis. We evaluate both variants on an ARM Cortex-M4 implementation of Dilithium-2. The experiments show that the attack assuming the knowledge of t0 can recover s1 from a single trace captured from a different from profiling device with a non-negligible probability.
- 7 October 2024
- Published by International Association for Cryptologic Research in IACR Communications in Cryptology
- Vol. 1 (3)
Open AccessUnpacking Needs Protection
Most of the previous attacks on Dilithium exploit side-channel information which is leaked during the computation of the polynomial multiplication cs1, where s1 is a small-norm secret and c is a verifier's challenge. In this paper, we present a new attack utilizing leakage during secret key unpacking in the signing algorithm. The unpacking is also used in other post-quantum cryptographic algorithms, including Kyber, because inputs and outputs of their API functions are byte arrays. Exploiting leakage during unpacking is more challenging than exploiting leakage during the computation of cs1 since c varies for each signing, while the unpacked secret key remains constant. Therefore, post-processing is required in the latter case to recover a full secret key. We present two variants of post-processing. In the first one, a half of the coefficients of the secret s1 and the error s2 is recovered by profiled deep learning-assisted power analysis and the rest is derived by solving linear equations based on t = As1 + s2, where A and t are parts of the public key. This case assumes knowledge of the least significant bits of t, t0. The second variant uses lattice reduction to derive s1 without the knowledge of t0. However, it needs a larger portion of s1 to be recovered by power analysis. We evaluate both variants on an ARM Cortex-M4 implementation of Dilithium-2. The experiments show that the attack assuming the knowledge of t0 can recover s1 from a single trace captured from a different from profiling device with a non-negligible probability.
- 25 April 2024
- Published by Springer Nature
- p. 424-439
A Shared Key Recovery Attack on a Masked Implementation of CRYSTALS-Kyber’s Encapsulation Algorithm
- 25 April 2024
- Published by Springer Nature
- p. 424-439
A Shared Key Recovery Attack on a Masked Implementation of CRYSTALS-Kyber’s Encapsulation Algorithm