Don't Learn What You Already Know: Grey-Box Modeling for Profiling Side-Channel Analysis against Masking

Masure, Loïc;Cristiani Valence;Lecomte, Maxime;Standaert, François-Xavier
(2023) Transactions on Cryptographic Hardware and Embedded Systems — Vol. 2023, n° 1, p. 32-59 (2023)

Files

TCHES2023_1_0211.pdf
  • Open Access
  • Adobe PDF
  • 874.88 KB
Download

Details

Authors
Abstract
Over the past few years, deep-learning-based attacks have emerged as ade factostandard, thanks to their ability to break implementations of cryptographicprimitives without pre-processing, even against widely used counter-measures suchas hiding and masking. However, the recent works of Bronchain and StandaertatTches2020 questioned the soundness of such tools if used in an uninformedsetting to evaluate implementations protected with higher-order masking. On theopposite, worst-case evaluations may be seen as possibly far from what a real-worldadversary could do, thereby leading to too conservative security bounds. In thispaper, we propose a new threat model that we namescheme-awarebenefiting from atrade-off between uninformed and worst-case models. Our scheme-aware model iscloser to a real-world adversary, in the sense that it does not need to have access tothe random nonces used by masking during the profiling phase like in a worst-casemodel, while it does not need to learn the masking scheme as implicitly done by anuninformed adversary. We show how to combine the power of deep learning withthe prior knowledge of scheme-aware modeling. As a result, we show on simulationsand experiments on public datasets how it sometimes allows to reduce by an orderof magnitude theprofilingcomplexity,i.e., the number of profiling traces needed tosatisfyingly train a model, compared to a fully uninformed adversary.
Affiliations

Citations

  • APA
  • Chicago
  • FWB

Masure, L., Cristiani Valence, Lecomte, M., & Standaert, F.-X. (2023). Don’t Learn What You Already Know: Grey-Box Modeling for Profiling Side-Channel Analysis against Masking. Transactions on Cryptographic Hardware and Embedded Systems, 2023(1), 32-59. https://doi.org/10.46586/tches.v2023.i1.32-59 (Original work published 2023)