Paper 2023/1100

Shift-invariance Robustness of Convolutional Neural Networks in Side-channel Analysis

Marina Krček, Delft University of Technology
Lichao Wu, Radboud University Nijmegen
Guilherme Perin, Leiden University
Stjepan Picek, Radboud University Nijmegen, Delft University of Technology

Convolutional neural networks (CNNs) offer unrivaled performance in profiling side-channel analysis. This claim is corroborated by numerous results where CNNs break targets protected with masking and hiding countermeasures. One hiding countermeasure is commonly investigated in related works - desynchronization (misalignment). The conclusions usually state that CNNs can break desynchronization as they are shift-invariant. This paper investigates that claim in more detail and reveals that the situation is more complex. While CNNs have certain shift-invariance, it is insufficient for commonly encountered scenarios in deep learning-based side-channel analysis. We propose to use data augmentation to improve the shift-invariance and, in a more powerful version, ensembles of data augmentation. Our results show the proposed techniques work very well and improve the attack significantly, even for an order of magnitude.

Available format(s)
Publication info
Side-channel AnalysisDeep learningMisalignmentCountermeasuresShift-invariance
Contact author(s)
m krcek @ tudelft nl
lichao wu @ ru nl
g perin @ liacs leidenuniv nl
stjepan picek @ ru nl
2023-07-17: approved
2023-07-14: received
See all versions
Short URL
Creative Commons Attribution


      author = {Marina Krček and Lichao Wu and Guilherme Perin and Stjepan Picek},
      title = {Shift-invariance Robustness of Convolutional Neural Networks in Side-channel Analysis},
      howpublished = {Cryptology ePrint Archive, Paper 2023/1100},
      year = {2023},
      note = {\url{}},
      url = {}
Note: In order to protect the privacy of readers, does not use cookies or embedded third party content.