Paper published in a journal (Scientific congresses, symposiums and conference proceedings)
OBSCURE: Versatile Software Obfuscation from a Lightweight Secure Element
Mercadier, Darius; Nguyen, Viet Sang; Rivain, Matthieu et al.
2024In IACR Transactions on Cryptographic Hardware and Embedded Systems, 2024 (2), p. 588 - 629
Peer Reviewed verified by ORBi
 

Files


Full Text
TCHES2024_2_21.pdf
Publisher postprint (1.53 MB) Creative Commons License - Attribution
Download

All documents in ORBilu are protected by a user license.

Send to



Details



Keywords :
Obfuscation; Secure Element; White-Box Cryptography; VBB Security
Abstract :
[en] Software obfuscation is a powerful tool to protect the intellectual property or secret keys inside programs. Strong software obfuscation is crucial in the context of untrusted execution environments (e.g., subject to malware infection) or to face potentially malicious users trying to reverse-engineer a sensitive program. Unfortunately, the state-of-the-art of pure software-based obfuscation (including white-box cryptography) is either insecure or infeasible in practice. This work introduces OBSCURE, a versatile framework for practical and cryptographically strong software obfuscation relying on a simple stateless secure element (to be embedded, for example, in a protected hardware chip or a token). Based on the foundational result by Goyal et al. from TCC 2010, our scheme enjoys provable security guarantees, and further focuses on practical aspects, such as efficient execution of the obfuscated programs, while maintaining simplicity of the secure element. In particular, we propose a new rectangular universalization technique, which is also of independent interest. We provide an implementation of OBSCURE taking as input a program source code written in a subset of the C programming language. This ensures usability and a broad range of applications of our framework. We benchmark the obfuscation on simple software programs as well as on cryptographic primitives, hence highlighting the possible use cases of the framework as an alternative to pure software-based white-box implementations.
Disciplines :
Computer science
Author, co-author :
Mercadier, Darius;  Google, Munich, Germany
Nguyen, Viet Sang;  Université Jean Monnet, Saint-Étienne, France
Rivain, Matthieu;  CryptoExperts, Paris, France
UDOVENKO, Aleksei  ;  University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > Cryptolux
External co-authors :
yes
Language :
English
Title :
OBSCURE: Versatile Software Obfuscation from a Lightweight Secure Element
Publication date :
12 March 2024
Event name :
Conference on Cryptographic Hardware and Embedded Systems (CHES)
Event organizer :
International Association for Cryptologic Research (IACR)
Event place :
Halifax, Canada
Event date :
September 4-7, 2024
Audience :
International
Journal title :
IACR Transactions on Cryptographic Hardware and Embedded Systems
eISSN :
2569-2925
Publisher :
Ruhr-University of Bochum
Volume :
2024
Issue :
2
Pages :
588 - 629
Peer reviewed :
Peer Reviewed verified by ORBi
FnR Project :
FNR13641232 - Analysis And Protection Of Lightweight Cryptographic Algorithms, 2019 (01/01/2021-31/12/2023) - Alex Biryukov
Name of the research project :
R-AGR-3748 - C19/IS/13641232/APLICA - BIRYUKOV Alexei
Funders :
ANR - Agence Nationale de la Recherche
FNR - Fonds National de la Recherche
DFG - Deutsche Forschungsgemeinschaft
Funding text :
This work was done while the authors were at CryptoExperts. This work was partially supported by the French ANR-AAPG2019 SWITECH project. The fourth author was partially supported by the Luxembourg National Research Fund’s (FNR) and the German Research Foundation’s (DFG) joint project APLICA (C19/IS/13641232). We are very grateful to Pascal Paillier for generating the discretized neural network used in our benchmark (as reported in Appendix C). We would also like to thank the anonymous reviewers of TCHES for their fruitful comments that helped us improve the paper.
Available on ORBilu :
since 09 October 2024

Statistics


Number of views
101 (4 by Unilu)
Number of downloads
33 (0 by Unilu)

Scopus citations®
 
1
Scopus citations®
without self-citations
1
OpenCitations
 
0
OpenAlex citations
 
1

Bibliography


Similar publications



Contact ORBilu