Paper 2021/1569

ABBY: Automating the creation of fine-grained leakage models

Omid Bazangani, Alexandre Iooss, Ileana Buhan, and Lejla Batina

Abstract

Side-channel leakage simulators allow testing the resilience of cryptographic implementations to power side-channel attacks without a dedicated setup. The main challenge in their large-scale deployment is the limited support for target devices, a direct consequence of the effort required for reverse engineering microarchitecture implementations. We introduce ABBY, the first solution for the automated creation of fine-grained leakage models. The main innovation of ABBY is the training framework, which can automatically characterize the microarchitecture of the target device and is portable to other platforms. Evaluation of ABBY on real-world crypto implementations exhibits comparable performance to other state-of-the-art leakage simulators.

Metadata
Available format(s)
PDF
Category
Implementation
Publication info
Preprint. MINOR revision.
Keywords
side-channelsleakage simulatorneural networksmicroarchitecture
Contact author(s)
omid bazangani @ gmail com
omid bazangani @ ru nl
alexandre iooss @ crans org
ileana buhan @ ru nl
lejla batina @ ru nl
History
2021-12-06: revised
2021-12-03: received
See all versions
Short URL
https://ia.cr/2021/1569
License
Creative Commons Attribution
CC BY

BibTeX

@misc{cryptoeprint:2021/1569,
      author = {Omid Bazangani and Alexandre Iooss and Ileana Buhan and Lejla Batina},
      title = {ABBY: Automating the creation of fine-grained leakage models},
      howpublished = {Cryptology ePrint Archive, Paper 2021/1569},
      year = {2021},
      note = {\url{https://eprint.iacr.org/2021/1569}},
      url = {https://eprint.iacr.org/2021/1569}
}
Note: In order to protect the privacy of readers, eprint.iacr.org does not use cookies or embedded third party content.