Paper 2024/1408

Multiple-Tweak Differential Attack Against SCARF

Christina Boura, IRIF, Université Paris Cité, Paris, France
Shahram Rasoolzadeh, Ruhr University Bochum, Bochum, Germany
Dhiman Saha, Indian Institute of Technology, Bhilai, India
Yosuke Todo, NTT Social Informatics Laboratories, Tokyo, Japan
Abstract

In this paper, we present the first third-party cryptanalysis of SCARF, a tweakable low-latency block cipher designed to thwart contention-based cache attacks through cache randomization. We focus on multiple-tweak differential attacks, exploiting biases across multiple tweaks. We establish a theoretical framework explaining biases for any number of rounds and verify this framework experimentally. Then, we use these properties to develop a key recovery attack on 7-round SCARF with a time complexity of \(2^{76}\), achieving a 98.9% success rate in recovering the 240-bit secret key. Additionally, we introduce a distinguishing attack on the full 8-round SCARF in a multi-key setting, with a complexity of \(c \times 2^{67.55}\), demonstrating that SCARF does not provide 80-bit security under these conditions. We also explore whether our approach could be extended to the single-key model and discuss the implications of different S-box choices on the attack success.

Metadata
Available format(s)
PDF
Category
Secret-key cryptography
Publication info
A major revision of an IACR publication in ASIACRYPT 2024
Keywords
SCARFcache randomizationlow latencydifferential cryptanalysismultiple-tweak differential attack
Contact author(s)
christina boura @ irif fr
shahram rasoolzadeh @ rub de
dhiman @ iitbhilai ac in
todo yosuke @ gmail com
History
2024-09-11: approved
2024-09-09: received
See all versions
Short URL
https://ia.cr/2024/1408
License
Creative Commons Attribution-ShareAlike
CC BY-SA

BibTeX

@misc{cryptoeprint:2024/1408,
      author = {Christina Boura and Shahram Rasoolzadeh and Dhiman Saha and Yosuke Todo},
      title = {Multiple-Tweak Differential Attack Against {SCARF}},
      howpublished = {Cryptology {ePrint} Archive, Paper 2024/1408},
      year = {2024},
      url = {https://eprint.iacr.org/2024/1408}
}
Note: In order to protect the privacy of readers, eprint.iacr.org does not use cookies or embedded third party content.