Towards Private Deep Learning-Based Side-Channel Analysis Using Homomorphic Encryption

Fabian Schmid; Shibam Mukherjee; Stjepan Picek; Marc Stöttinger; Fabrizio De Santis; Christian Rechberger

doi:10.1007/978-3-031-57543-3_8

Towards Private Deep Learning-Based Side-Channel Analysis Using Homomorphic Encryption

Fabian Schmid^*, Shibam Mukherjee, Stjepan Picek, Marc Stöttinger, Fabrizio De Santis, Christian Rechberger

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference paper › peer-review

Abstract

This work investigates using Homomorphic Encryption (HE) to assist the security evaluation of cryptographic devices without revealing side-channel information. For the first time, we evaluate the feasibility of execution of deep learning-based side-channel analysis on standard server equipment using an adapted HE protocol. By examining accuracy and execution time, it demonstrates the successful application of private SCA on both unprotected and protected cryptographic implementations. This contribution is a first step towards confidential side-channel analysis. Our study is limited to the honest-but-curious trust model, where we could reconstruct the secret of an unprotected AES implementation in seconds and of a masked AES implementation in under 17 min.

Original language	English
Title of host publication	Constructive Side-Channel Analysis and Secure Design - 15th International Workshop, COSADE 2024, Proceedings
Editors	Romain Wacquez, Naofumi Homma
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	133-154
Number of pages	22
ISBN (Print)	9783031575426
DOIs	https://doi.org/10.1007/978-3-031-57543-3_8
Publication status	Published - 2024
Event	15th International Workshop on Constructive Side-Channel Analysis and Secure Design: COSADE 2024 - 880, route de Mimet, Gardanne, France Duration: 8 Apr 2024 → 10 Apr 2024 https://www.cosade.org/cosade24/program.html

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	14595 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	15th International Workshop on Constructive Side-Channel Analysis and Secure Design
Abbreviated title	COSADE 2024
Country/Territory	France
City	Gardanne
Period	8/04/24 → 10/04/24
Internet address	https://www.cosade.org/cosade24/program.html

Keywords

Deep Learning
Homomorphic Encryption
Neural Networks
Private AI
Side-channel Analysis

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-031-57543-3_8

Cite this

Schmid, F., Mukherjee, S., Picek, S., Stöttinger, M., De Santis, F., & Rechberger, C. (2024). Towards Private Deep Learning-Based Side-Channel Analysis Using Homomorphic Encryption. In R. Wacquez, & N. Homma (Eds.), Constructive Side-Channel Analysis and Secure Design - 15th International Workshop, COSADE 2024, Proceedings (pp. 133-154). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14595 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-57543-3_8

Towards Private Deep Learning-Based Side-Channel Analysis Using Homomorphic Encryption. / Schmid, Fabian; Mukherjee, Shibam; Picek, Stjepan et al.
Constructive Side-Channel Analysis and Secure Design - 15th International Workshop, COSADE 2024, Proceedings. ed. / Romain Wacquez; Naofumi Homma. Springer Science and Business Media Deutschland GmbH, 2024. p. 133-154 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14595 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference paper › peer-review

Schmid, F, Mukherjee, S, Picek, S, Stöttinger, M, De Santis, F & Rechberger, C 2024, Towards Private Deep Learning-Based Side-Channel Analysis Using Homomorphic Encryption. in R Wacquez & N Homma (eds), Constructive Side-Channel Analysis and Secure Design - 15th International Workshop, COSADE 2024, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 14595 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 133-154, 15th International Workshop on Constructive Side-Channel Analysis and Secure Design, Gardanne, France, 8/04/24. https://doi.org/10.1007/978-3-031-57543-3_8

Schmid F, Mukherjee S, Picek S, Stöttinger M, De Santis F, Rechberger C. Towards Private Deep Learning-Based Side-Channel Analysis Using Homomorphic Encryption. In Wacquez R, Homma N, editors, Constructive Side-Channel Analysis and Secure Design - 15th International Workshop, COSADE 2024, Proceedings. Springer Science and Business Media Deutschland GmbH. 2024. p. 133-154. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-57543-3_8

Schmid, Fabian ; Mukherjee, Shibam ; Picek, Stjepan et al. / Towards Private Deep Learning-Based Side-Channel Analysis Using Homomorphic Encryption. Constructive Side-Channel Analysis and Secure Design - 15th International Workshop, COSADE 2024, Proceedings. editor / Romain Wacquez ; Naofumi Homma. Springer Science and Business Media Deutschland GmbH, 2024. pp. 133-154 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{e41ea3f48c914c228a2ea03ef30e4854,

title = "Towards Private Deep Learning-Based Side-Channel Analysis Using Homomorphic Encryption",

abstract = "This work investigates using Homomorphic Encryption (HE) to assist the security evaluation of cryptographic devices without revealing side-channel information. For the first time, we evaluate the feasibility of execution of deep learning-based side-channel analysis on standard server equipment using an adapted HE protocol. By examining accuracy and execution time, it demonstrates the successful application of private SCA on both unprotected and protected cryptographic implementations. This contribution is a first step towards confidential side-channel analysis. Our study is limited to the honest-but-curious trust model, where we could reconstruct the secret of an unprotected AES implementation in seconds and of a masked AES implementation in under 17 min.",

keywords = "Deep Learning, Homomorphic Encryption, Neural Networks, Private AI, Side-channel Analysis",

author = "Fabian Schmid and Shibam Mukherjee and Stjepan Picek and Marc St{\"o}ttinger and {De Santis}, Fabrizio and Christian Rechberger",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.; 15th International Workshop on Constructive Side-Channel Analysis and Secure Design : COSADE 2024, COSADE 2024 ; Conference date: 08-04-2024 Through 10-04-2024",

year = "2024",

doi = "10.1007/978-3-031-57543-3_8",

language = "English",

isbn = "9783031575426",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "133--154",

editor = "Romain Wacquez and Naofumi Homma",

booktitle = "Constructive Side-Channel Analysis and Secure Design - 15th International Workshop, COSADE 2024, Proceedings",

address = "Germany",

url = "https://www.cosade.org/cosade24/program.html",

}

TY - GEN

T1 - Towards Private Deep Learning-Based Side-Channel Analysis Using Homomorphic Encryption

AU - Schmid, Fabian

AU - Mukherjee, Shibam

AU - Picek, Stjepan

AU - Stöttinger, Marc

AU - De Santis, Fabrizio

AU - Rechberger, Christian

N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

PY - 2024

Y1 - 2024

N2 - This work investigates using Homomorphic Encryption (HE) to assist the security evaluation of cryptographic devices without revealing side-channel information. For the first time, we evaluate the feasibility of execution of deep learning-based side-channel analysis on standard server equipment using an adapted HE protocol. By examining accuracy and execution time, it demonstrates the successful application of private SCA on both unprotected and protected cryptographic implementations. This contribution is a first step towards confidential side-channel analysis. Our study is limited to the honest-but-curious trust model, where we could reconstruct the secret of an unprotected AES implementation in seconds and of a masked AES implementation in under 17 min.

AB - This work investigates using Homomorphic Encryption (HE) to assist the security evaluation of cryptographic devices without revealing side-channel information. For the first time, we evaluate the feasibility of execution of deep learning-based side-channel analysis on standard server equipment using an adapted HE protocol. By examining accuracy and execution time, it demonstrates the successful application of private SCA on both unprotected and protected cryptographic implementations. This contribution is a first step towards confidential side-channel analysis. Our study is limited to the honest-but-curious trust model, where we could reconstruct the secret of an unprotected AES implementation in seconds and of a masked AES implementation in under 17 min.

KW - Deep Learning

KW - Homomorphic Encryption

KW - Neural Networks

KW - Private AI

KW - Side-channel Analysis

UR - http://www.scopus.com/inward/record.url?scp=85190446093&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-57543-3_8

DO - 10.1007/978-3-031-57543-3_8

M3 - Conference paper

AN - SCOPUS:85190446093

SN - 9783031575426

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 133

EP - 154

BT - Constructive Side-Channel Analysis and Secure Design - 15th International Workshop, COSADE 2024, Proceedings

A2 - Wacquez, Romain

A2 - Homma, Naofumi

PB - Springer Science and Business Media Deutschland GmbH

T2 - 15th International Workshop on Constructive Side-Channel Analysis and Secure Design

Y2 - 8 April 2024 through 10 April 2024

ER -

Towards Private Deep Learning-Based Side-Channel Analysis Using Homomorphic Encryption

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this