Skip to main navigation Skip to search Skip to main content

Breaking Masked Kyber: ResNet-Based Masked Kyber Share Recovery Method

  • Yaoling Ding
  • , Haotong Xu
  • , Chong Luo
  • , Annyu Liu*
  • , Zheyu Zhang*
  • , Jing Yu
  • , An Wang
  • *Corresponding author for this work
  • Beijing Institute of Technology
  • CAS - Institute of Computing Technology
  • China Industrial Control Systems Cyber Emergency Response Team

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

As post-quantum cryptography becomes the cornerstone of future security systems, it is crucial to understand its practical resilience under present side-channel analysis. In this paper, we propose an effective side-channel analysis method for the poly_tomsg() function in the decapsulation of the first-order mask implementation of CRYSTALS-Kyber (the NIST-selected post-quantum key encapsulation mechanism). Unlike previous methods that target individual bits, our method targets a masked shared byte. By systematically evaluating multiple neural network architectures (MLP, CNN, ResNet), we find that residual learning has the best robustness under low signal-to-noise ratios. The method achieves up to 98% average accuracy by training a ResNet18 network, and is able to recover m from a single trace with a success rate of over 38%, outperforming classical template attacks without trace alignment. These findings reveal that even well-implemented Boolean masking can be defeated by neural-based profiling attacks, thus emphasizing the need for stronger countermeasures in post-quantum cryptographic implementations.

Original languageEnglish
Title of host publicationInformation Security and Cryptology – ICISC 2025 - 28th International Conference, Revised Selected Papers
EditorsHwajeong Seo
PublisherSpringer Science and Business Media Deutschland GmbH
Pages185-197
Number of pages13
ISBN (Print)9789819580330
DOIs
Publication statusPublished - 2026
Externally publishedYes
Event28th International Conference on Information Security and Cryptology, ICISC 2025 - Seoul, Korea, Republic of
Duration: 19 Nov 202521 Nov 2025

Publication series

NameLecture Notes in Computer Science
Volume16487 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference28th International Conference on Information Security and Cryptology, ICISC 2025
Country/TerritoryKorea, Republic of
CitySeoul
Period19/11/2521/11/25

Keywords

  • CRYSTALS-Kyber
  • Masking
  • Neural networks
  • Profiling attacks
  • Side-channel analysis

Fingerprint

Dive into the research topics of 'Breaking Masked Kyber: ResNet-Based Masked Kyber Share Recovery Method'. Together they form a unique fingerprint.

Cite this