TARU PUBLICATIONS
Journal of Discrete Mathematical Sciences and Cryptography cover

Journal of Discrete Mathematical Sciences and Cryptography

(Since 1998)
Open Access ·Peer-reviewed·ISSN (Online): 2169-0065·ISSN (Print): 0972-0529

CiteScore 2025: 2.7 (Q1)

Powered by:Powered by

Monthly Journal: Publishes theoretical and applied research in all areas of Discrete Mathematical Sciences, Cryptography, Combinatorics, Elliptic Curves and Information Security.

Issues up to 2022 co-published with and available at:Taylor & Francis Online
submissions@tarupublications.com
Taru Online First Current IssueSpecial IssuesBrowse all Issues
Journal of Discrete Mathematical Sciences and Cryptography
Open Access Research Article

AI-augmented cryptanalysis using combinatorics and reinforcement learning

, , * ,

* Corresponding author · click or hover a name for details

pp. 2071–2080Vol. 28Issue 5-BAugust 2025DOI: 10.47974/JDMSC-2424 Crossmark XML
Received:
05 Nov 2024
Published Online:
30 Aug 2025
Article type:
Research Article
Language:
EN
Article no.:
JDMSC-2424
Pages:
2071–2080

Abstract

On one side, the modern day symmetric-key cryptography protects everything from our online banking to critical infrastructures, and the giant key-space of AES-128 alone around 3.4×10³⁸ possibilities makes the brute-force attack impractical, which makes sense for the development of more sophisticated search techniques. Although traditional combinatorial attacks eliminate keys based on algebraic relationships, they have exponential complexity even for low-degree equations and are not applicable to multiple cipher designs. At the same time, advances in machine learning that have been made recently demonstrate that reinforcement learning agents can learn to control search heuristics, cutting solution times by as much as a factor of ten in closely associated optimization problems and greatly enhanced success rates in side-channel key recovery. In this paper, we bridge by incorporating an RL policy in a combinatorial key-search engine, with partial-key candidates and their statistical properties as states, to guide an order of candidates of favorable branches. Results on several cipher instances demonstrate 14–19 percentage-point improvement in recovery rates and close to two-times speed improvement over standard and pureRL baselines, suggesting a promising path towards more intelligent, automated cryptanalysis.

Keywords

AI-augmented cryptanalysisReinforcement learningCombinatorial key-searchBlock-cipher securitySearch-space pruningAutomated key recovery

Subject Classifications

90C27

References

[1] M. Medwed, J. Heyszl, and C. Paar, “Advanced key-space pruning techniques for meet-in-the-middle attacks on lightweight block ciphers,” IACR Transactions on Symmetric Cryptology, vol. 2021, no. 4, pp. 345–367 (2021).
[2] M. Zhang and W. Hsieh, “Deep reinforcement learning–guided beam search for the traveling salesman problem,” IEEE Transactions on Neural Networks and Learning Systems, vol. 34, no. 2, pp. 789–802 (2023).
[3] D. Cagli and S. Guilley, “TripletNet: Triplet-loss profiling for deep-learning side-channel analysis of AES,” IEEE Transactions on Information Forensics and Security, vol. 18, pp. 1234–1245 (2023).
[4] A. Kumar, P. Singh, and S. Rani, “DLFA: A deep-learning framework for fault analysis in AES implementations,” Journal of Cryptographic Engineering, vol. 13, no. 1, pp. 55–68 (2024).
[5] X. Li and J. Zhang, “MILP-based automatic differential trail search for lightweight block ciphers,” IACR Journal of Cryptographic Hardware and Embedded Systems, vol. 2023, no. 4, pp. 112–130 (2023).
[6] R. Patel and S. Singh, “A hybrid particle swarm optimization and simulated annealing approach to block cipher key recovery,” International Journal of Information Security, vol. 21, no. 5, pp. 467–484 (2022).
[7] W. Kool, H. van Hoof, and M. Welling, “Attention, learn to solve routing problems!,” IEEE Transactions on Neural Networks and Learning Systems, vol. 32, no. 4, pp. 1620–1633 (2021).
[8] E. Borges-Trenard, H. Kim, and R. Singh, “Genetic algorithm partitioning attack on reduced-round DES,” Journal of Cryptographic Engineering, vol. 12, no. 3, pp. 221–234 (2022).
[9] R. Mehra, V. Singh, and A. Patel, “A mixed-integer programming approach to analyzing block-cipher resistance under truncated differential attacks,” Journal of Information & Optimization Sciences, vol. 47, no. 2, pp. 201–216 (2025).
[10] S. Roy, M. Banerjee, and T. Chakraborty, “A hybrid heuristic combining genetic algorithms and simulated annealing for cryptanalysis of lightweight block ciphers,” Journal of Interdisciplinary Mathematics, vol. 29, no. 3, pp. 345–360 (2025).
[11] T. Wu, Q. Li, and Y. Zhang, “Reinforcement learning for combinatorial optimization: A survey of recent advances,” Applied Sciences, vol. 13, no. 7, p. 3532 (2023).

Views: 276Downloads: 8Citations: 0