OccPoIs: Points of Interest Based on Neural Network’s Key Recovery in Side-Channel Analysis Through Occlusion
Pages 3 - 28
Abstract
Deep neural networks (DNNs) represent a powerful technique for assessing cryptographic security concerning side-channel analysis (SCA) due to their ability to aggregate leakages automatically, rendering attacks more efficient without preprocessing. Despite their effectiveness, DNNs are predominantly black-box algorithms, posing considerable interpretability challenges. In this paper, we propose a novel technique called Key Guessing Occlusion (KGO) that acquires a minimal set of sample points required by the DNN for key recovery, which we call OccPoIs. These OccPoIs provide information about the areas of the traces important to the DNN for retrieving the key, enabling evaluators to know where to refine their cryptographic implementation. After obtaining the OccPoIs, we first explore the leakages found in these OccPoIs to understand what the DNN is learning with first-order Correlation Power Analysis (CPA). We show that KGO obtains relevant sample points that have a high correlation with the given leakage model but also acquires sample points that first-order CPA fails to capture. Furthermore, unlike the first-order CPA in the masking setting, KGO obtains these OccPoIs without knowing the shares or mask. Next, we employ the template attack (TA) using the OccPoIs to investigate if KGO could be used as a feature selection tool. We show that using the OccPoIs with TA can recover the key for all the considered synchronized datasets and is consistent even on datasets protected by first-order masking. Finally, KGO also allows a more efficient attack than other feature selection techniques on the first-order masking dataset called ASCADf.
References
[1]
Alber, M., et al.: iNNvestigate neural networks! J. Mach. Learn. Res. 20(93), 1–8 (2019). http://jmlr.org/papers/v20/18-540.html
[2]
Amazon: Model Explainability with AWS Artificial Intelligence and Machine Learning Solutions (2021). https://docs.aws.amazon.com/whitepapers/latest/model-explainability-aws-ai-ml/interpretability-versus-explainability.html
[3]
Ancona, M., Ceolini, E., Öztireli, C., Gross, M.: Towards better understanding of gradient-based attribution methods for deep neural networks. arXiv preprint arXiv:1711.06104 (2017)
[4]
Bach S, Binder A, Montavon G, Klauschen F, Müller KR, and Samek W On pixel-wise explanations for non-linear classifier decisions by layer-wise relevance propagation PLoS ONE 2015 10 7 e0130140
[5]
Benadjila R, Prouff E, Strullu R, Cagli E, and Dumas C Deep learning for side-channel analysis and introduction to ASCAD database J. Cryptogr. Eng. 2020 10 2 163-188
[6]
Bhasin, S., Danger, J.L., Guilley, S., Najm, Z.: NICV: normalized inter-class variance for detection of side-channel leakage. In: 2014 International Symposium on Electromagnetic Compatibility, Tokyo, pp. 310–313 (2014)
[7]
Burkart N and Huber MF A survey on the explainability of supervised machine learning J. Artif. Intell. Res. 2021 70 245-317
[8]
Cagli E, Dumas C, and Prouff E Fischer W and Homma N Convolutional neural networks with data augmentation against jitter-based countermeasures Cryptographic Hardware and Embedded Systems – CHES 2017 2017 Cham Springer 45-68 10529
[9]
Choudary O and Kuhn MG Francillon A and Rohatgi P Efficient template attacks Smart Card Research and Advanced Applications 2014 Cham Springer 253-270 8419
[10]
Debande, N., Souissi, Y., Nassar, M., Guilley, S., Le, T.H., Danger, J.L.: “Re-synchronization by moments”: an efficient solution to align Side-Channel traces. In: 2011 IEEE International Workshop on Information Forensics and Security, pp. 1–6 (2011)
[11]
Egger M, Schamberger T, Tebelmann L, Lippert F, and Sigl G Balasch J and O’Flynn C A second look at the ASCAD databases Constructive Side-Channel Analysis and Secure Design 2022 Cham Springer 75-99
[12]
Gierlichs B, Lemke-Rust K, and Paar C Goubin L and Matsui M Templates vs. stochastic methods Cryptographic Hardware and Embedded Systems - CHES 2006 2006 Heidelberg Springer 15-29 4249
[13]
Gilpin, L.H., Bau, D., Yuan, B.Z., Bajwa, A., Specter, M., Kagal, L.: Explaining explanations: An overview of interpretability of machine learning. In: 2018 IEEE 5th International Conference on data science and advanced analytics (DSAA), pp. 80–89. IEEE (2018)
[14]
Hettwer B, Gehrer S, and Güneysu T Paterson KG and Stebila D Deep neural network attribution methods for leakage analysis and symmetric key recovery Selected Areas in Cryptography – SAC 2019 2020 Cham Springer 645-666 11959
[15]
Lerman L, Poussier R, Bontempi G, Markowitch O, and Standaert F-X Mangard S and Poschmann AY Template attacks vs. machine learning revisited (and the curse of dimensionality in side-channel analysis) Constructive Side-Channel Analysis and Secure Design 2015 Cham Springer 20-33 9064
[16]
Masure L, Dumas C, and Prouff E Polian I and Stöttinger M Gradient visualization for general characterization in profiling attacks Constructive Side-Channel Analysis and Secure Design 2019 Cham Springer 145-167 11421
[17]
O’Flynn C and Chen ZD Prouff E ChipWhisperer: an open-source platform for hardware embedded security research Constructive Side-Channel Analysis and Secure Design 2014 Cham Springer 243-260 8622
[18]
Perin G, Buhan I, and Picek S Bhasin S and De Santis F Learning when to stop: a mutual information approach to prevent overfitting in profiled side-channel analysis Constructive Side-Channel Analysis and Secure Design 2021 Cham Springer 53-81 12910
[19]
Perin, G., Wu, L., Picek, S.: I know what your layers did: layer-wise explainability of deep learning side-channel analysis. Cryptology ePrint Archive, Paper 2022/1087 (2022). https://eprint.iacr.org/2022/1087
[20]
Picek, S., Heuser, A., Jovic, A., Batina, L.: Asystematic evaluation of profiling through focused feature selection. IEEE Trans. Very Large Scale Integr. (VLSI) Syst. 27(12), 2802–2815 (2019)
[21]
Picek, S., Heuser, A., Jovic, A., Ludwig, S.A., Guilley, S., Jakobovic, D., Mentens, N.: Side-channel analysis and machine learning: a practical perspective. In: 2017 International Joint Conference on Neural Networks (IJCNN), pp. 4095–4102 (2017)
[22]
Reparaz O, Gierlichs B, and Verbauwhede I Prouff E and Schaumont P Selecting time samples for multivariate DPA attacks Cryptographic Hardware and Embedded Systems – CHES 2012 2012 Heidelberg Springer 155-174 7428
[23]
Rijsdijk J, Wu L, Perin G, and Picek S Reinforcement learning for hyperparameter tuning in deep learning-based side-channel analysis IACR Trans. Cryptographic Hardware Embedded Syst. 2021 2021 3 677-707
[24]
Rioja, U., Batina, L., Flores, J.L., Armendariz, I.: Auto-tune POIS: estimation of distribution algorithms for efficient side-channel analysis. Computer Networks 198, 108405 (2021). https://www.sciencedirect.com/science/article/pii/S1389128621003789
[25]
Schamberger T, Egger M, Tebelmann L, et al. Zhou J et al. Hide and seek: using occlusion techniques for side-channel leakage attribution in CNNs Applied Cryptography and Network Security Workshops 2023 Cham Springer 139-158 13907
[26]
Standaert F-X and Archambeau C Oswald E and Rohatgi P Using subspace-based template attacks to compare and combine power and electromagnetic information leakages Cryptographic Hardware and Embedded Systems – CHES 2008 2008 Heidelberg Springer 411-425 5154
[27]
van der Valk D, Picek S, and Bhasin S Bertoni GM and Regazzoni F Kilroy was here: the first step towards explainability of neural networks in profiled side-channel analysis Constructive Side-Channel Analysis and Secure Design 2021 Cham Springer 175-199 12244
[28]
Wouters L, Arribas V, Gierlichs B, and Preneel B Revisiting a methodology for efficient CNN architectures in profiling attacks IACR Trans. Cryptographic Hardware Embedded Syst. 2020 2020 3 147-168
[29]
Wu L, Perin G, and Picek S I choose you: automated hyperparameter tuning for deep learning-based side-channel analysis IACR Cryptol. ePrint Arch. 2020 2020 1293
[30]
Wu, L., Perin, G., Picek, S.: I choose you: automated hyperparameter tuning for deep learning-based side-channel analysis. IEEE Trans. Emerging Topics Comput. 1–12 (2022)
[31]
Wu L, Perin G, and Picek S The best of two worlds: deep learning-assisted template attack IACR Trans. Cryptographic Hardware Embedded Syst. 2022 2022 3 413-437
[32]
Wu, L., Won, Y.S., Jap, D., Perin, G., Bhasin, S., Picek, S.: Ablation analysis for multi-device deep learning-based physical side-channel analysis. IEEE Trans. Depend. Secure Comput. 1–12 (2023)
[33]
Yap T, Benamira A, Bhasin S, and Peyrin T Peek into the black-box: interpretable neural network using sat equations in side-channel analysis IACR Trans. Cryptographic Hardware Embedded Syst. 2023 2023 2 24-53
[34]
Zaid G, Bossuet L, Carbone M, Habrard A, and Venelli A Conditional variational autoencoder based on stochastic attacks IACR Trans. Cryptographic Hardware Embedded Syst. 2023 2023 2 310-357
[35]
Zaid G, Bossuet L, Habrard A, and Venelli A Methodology for efficient CNN architectures in profiling attacks IACR Trans. Cryptographic Hardware Embedded Syst. 2019 2020 1 1-36
[36]
Zeiler, M.D., Fergus, R.: Visualizing and Understanding Convolutional Networks (2013)
[37]
Zheng Y, Zhou Y, Yu Z, Hu C, and Zhang H Hui LCK, Qing SH, Shi E, and Yiu SM How to compare selections of points of interest for side-channel distinguishers in practice? Information and Communications Security 2015 Cham Springer 200-214 8958
Index Terms
- OccPoIs: Points of Interest Based on Neural Network’s Key Recovery in Side-Channel Analysis Through Occlusion
Index terms have been assigned to the content through auto-classification.
Recommendations
Reinforcement Learning-Based Design of Side-Channel Countermeasures
Security, Privacy, and Applied Cryptography EngineeringAbstractDeep learning-based side-channel attacks are capable of breaking targets protected with countermeasures. The constant progress in the last few years makes the attacks more powerful, requiring fewer traces to break a target. Unfortunately, to ...
Beyond the Last Layer: Deep Feature Loss Functions in Side-channel Analysis
ASHES '23: Proceedings of the 2023 Workshop on Attacks and Solutions in Hardware SecurityThis paper provides a novel perspective on improving the efficiency of side-channel analysis by applying two deep feature loss functions: Soft Nearest Neighbor (SoftNN) and Center loss. By leveraging these loss functions during the deep neural networks (...
Side-channel indistinguishability
HASP '13: Proceedings of the 2nd International Workshop on Hardware and Architectural Support for Security and PrivacyWe introduce a masking strategy for hardware that prevents any side-channel attacker from recovering uniquely the secret key of a cryptographic device. In this masking scheme, termed homomorphic, the sensitive data is exclusive-ored with a random value ...
Comments
Information & Contributors
Information
Published In
Dec 2024
341 pages
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.
Publisher
Springer-Verlag
Berlin, Heidelberg
Publication History
Published: 18 December 2024
Author Tags
Qualifiers
- Article
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- 0Total Citations
- 0Total Downloads
- Downloads (Last 12 months)0
- Downloads (Last 6 weeks)0
Reflects downloads up to 19 Feb 2025