CEFI: Command Execution Flow Integrity for Embedded Devices

Anni Peng; Dongliang Fang; Wei Zhou; Erik van der Kouwe; Yin Li; Yuqing Zhang

doi:10.1007/978-3-031-35504-2_12

CEFI: Command Execution Flow Integrity for Embedded Devices

Anni Peng
, Dongliang Fang
, Wei Zhou
, Erik van der Kouwe
, Yin Li
, Yuqing Zhang

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

96 Downloads (Pure)

Abstract

As embedded devices are widely used in increasingly complex
settings (e.g., smart homes and industrial control systems), one device
is usually connected with multiple entities, such as mobile apps and the
cloud. Recent research has shown that privilege separation vulnerabilities,
which allow violations of authority between different entities, are
occuring in IoT systems. Because such vulnerabilities can be exploited
without violating static control flow and data flow, existing CFI and
DFI solutions cannot prevent them. We present CEFI , the first method
to enforce integrity of command execution on embedded devices after
deployment. CEFI provides fine-grained Command Execution Flow Integrity
by preventing external commands from being executed on control
flow paths belonging to interaction channels that are not authorized to
perform them. Using minimal manual annotations as a starting point,
CEFI statically determined the legal path set (from the start to the
end point) and instruments the program to verify the legitimacy of the
command execution at runtime by checking whether the calling context
is consistent between the runtime executed path and statically obtained
legal path set. We evaluate our prototype with five real-world firmware
samples, and show that CEFI has an average performance overhead of
just 0.18%, an average memory overhead of 0.19%, and that CEFI can
effectively protect embedded devices against attacks on privilege separation
vulnerabilities even if they do not violate control flow.

Original language	English
Title of host publication	Detection of Intrusions and Malware, and Vulnerability Assessment
Subtitle of host publication	20th International Conference, DIMVA 2023, Hamburg, Germany, July 12–14, 2023, Proceedings
Editors	Daniel Gruss, Federico Maggi, Mathias Fischer, Michele Carminati
Publisher	Springer Nature
Pages	235-255
Number of pages	21
ISBN (Electronic)	9783031355042
ISBN (Print)	9783031355035
DOIs	https://doi.org/10.1007/978-3-031-35504-2_12
Publication status	Published - 2023

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	13959 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Funding

Acknowledgements. We thank our shepherd Roland YAP Hock Chuan and anonymous reviewers for their valuable feedback. This work was supported by the National Natural Science Foundation of China (U1836210), the Key Research and Development Science and Technology of Hainan Province (GHYF2022010), the National Natural Science Foundation of China (No.62202188), and the National Key R&D Program of China (No.2022YFB31033400). Meanwhile, this work was partly done at VU Amsterdam. We thank the support provided by the China Scholarship Council (CSC) and the VUSec Group at VU Amsterdam. Acknowledgments. We would like to thank the anonymous reviewers and the shepherd Moritz Lipp for their very helpful comments and feedback during revision, which have significantly improved the quality and clarity of the work. This research was partially supported by NSF award 2207202. Any opinions, findings, and conclusions or recommendations in this paper are those of the authors and do not necessarily reflect the views of the NSF. Acknowledgments. Code to collect power consumption traces is based on Gras et al. [8]. This work has been partly funded by the ANR-19-CE39-0007 MIAOUS. Experiments presented in this paper were carried out using the Grid’5000 testbed, supported by a scientific interest group hosted by Inria and including CNRS, RENATER and several Universities as well as other organizations (see https://www.grid5000.fr). Acknowledgements. This material is based upon work supported by the National Science Foundation under Grant No. 1814402 and 1814234. Acknowledgements. We thank the reviewers and our shepherd for their helpful comments and suggestions. This work has been supported by ERDF through the EMSIK project and by BMBF through the PeT-HMR project. Acknowledgements. We thank the library maintainers for the smooth disclosure process, and the reviewers and our shepherd for their helpful comments and suggestions. This work has been supported by DFG under grants 427774779 and 439797619, and by BMBF through projects ENCOPIA and PeT-HMR. reviewers for their valuable comments that improved the quality of the paper. We acknowledge the support of the Natural Sciences and Engineering Research Council of Canada (NSERC), funding reference number RGPIN-2018-05919.

Funders	Funder number
ANR-19-CE39-0007 MIAOUS
Key Research and Development Science and Technology of Hainan Province	GHYF2022010, 62202188
VUSec Group at VU Amsterdam
National Science Foundation	1814234, 2207202, 1814402
Natural Sciences and Engineering Research Council of Canada	RGPIN-2018-05919
Deutsche Forschungsgemeinschaft	427774779, 439797619
National Natural Science Foundation of China	U1836210
Bundesministerium für Bildung und Forschung
China Scholarship Council
European Regional Development Fund
National Key Research and Development Program of China	2022YFB31033400

Access to Document

10.1007/978-3-031-35504-2_12

CEFI: Command Execution Flow Integrity for Embedded DevicesFinal published version, 586 KBLicence: Article 25fa Dutch Copyright Act

Persistent URL (handle)

Persistent link

Cite this

Peng, A., Fang, D., Zhou, W., van der Kouwe, E., Li, Y., & Zhang, Y. (2023). CEFI: Command Execution Flow Integrity for Embedded Devices. In D. Gruss, F. Maggi, M. Fischer, & M. Carminati (Eds.), Detection of Intrusions and Malware, and Vulnerability Assessment: 20th International Conference, DIMVA 2023, Hamburg, Germany, July 12–14, 2023, Proceedings (pp. 235-255). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13959 LNCS). Springer Nature. https://doi.org/10.1007/978-3-031-35504-2_12

Peng, Anni ; Fang, Dongliang ; Zhou, Wei et al. / CEFI: Command Execution Flow Integrity for Embedded Devices. Detection of Intrusions and Malware, and Vulnerability Assessment: 20th International Conference, DIMVA 2023, Hamburg, Germany, July 12–14, 2023, Proceedings. editor / Daniel Gruss ; Federico Maggi ; Mathias Fischer ; Michele Carminati. Springer Nature, 2023. pp. 235-255 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{3085685d854042bd8afe1eac7083af36,

title = "CEFI: Command Execution Flow Integrity for Embedded Devices",

abstract = "As embedded devices are widely used in increasingly complex settings (e.g., smart homes and industrial control systems), one device is usually connected with multiple entities, such as mobile apps and the cloud. Recent research has shown that privilege separation vulnerabilities, which allow violations of authority between different entities, are occuring in IoT systems. Because such vulnerabilities can be exploited without violating static control flow and data flow, existing CFI and DFI solutions cannot prevent them. We present CEFI , the first method to enforce integrity of command execution on embedded devices after deployment. CEFI provides fine-grained Command Execution Flow Integrity by preventing external commands from being executed on control flow paths belonging to interaction channels that are not authorized to perform them. Using minimal manual annotations as a starting point, CEFI statically determined the legal path set (from the start to the end point) and instruments the program to verify the legitimacy of the command execution at runtime by checking whether the calling context is consistent between the runtime executed path and statically obtained legal path set. We evaluate our prototype with five real-world firmware samples, and show that CEFI has an average performance overhead of just 0.18\%, an average memory overhead of 0.19\%, and that CEFI can effectively protect embedded devices against attacks on privilege separation vulnerabilities even if they do not violate control flow.",

author = "Anni Peng and Dongliang Fang and Wei Zhou and \{van der Kouwe\}, Erik and Yin Li and Yuqing Zhang",

year = "2023",

doi = "10.1007/978-3-031-35504-2\_12",

language = "English",

isbn = "9783031355035",

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

publisher = "Springer Nature",

pages = "235--255",

editor = "Daniel Gruss and Federico Maggi and Mathias Fischer and Michele Carminati",

booktitle = "Detection of Intrusions and Malware, and Vulnerability Assessment",

}

Peng, A, Fang, D, Zhou, W, van der Kouwe, E, Li, Y & Zhang, Y 2023, CEFI: Command Execution Flow Integrity for Embedded Devices. in D Gruss, F Maggi, M Fischer & M Carminati (eds), Detection of Intrusions and Malware, and Vulnerability Assessment: 20th International Conference, DIMVA 2023, Hamburg, Germany, July 12–14, 2023, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 13959 LNCS, Springer Nature, pp. 235-255. https://doi.org/10.1007/978-3-031-35504-2_12

CEFI: Command Execution Flow Integrity for Embedded Devices. / Peng, Anni; Fang, Dongliang; Zhou, Wei et al.
Detection of Intrusions and Malware, and Vulnerability Assessment: 20th International Conference, DIMVA 2023, Hamburg, Germany, July 12–14, 2023, Proceedings. ed. / Daniel Gruss; Federico Maggi; Mathias Fischer; Michele Carminati. Springer Nature, 2023. p. 235-255 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13959 LNCS).

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

TY - GEN

T1 - CEFI: Command Execution Flow Integrity for Embedded Devices

AU - Peng, Anni

AU - Fang, Dongliang

AU - Zhou, Wei

AU - van der Kouwe, Erik

AU - Li, Yin

AU - Zhang, Yuqing

PY - 2023

Y1 - 2023

N2 - As embedded devices are widely used in increasingly complex settings (e.g., smart homes and industrial control systems), one device is usually connected with multiple entities, such as mobile apps and the cloud. Recent research has shown that privilege separation vulnerabilities, which allow violations of authority between different entities, are occuring in IoT systems. Because such vulnerabilities can be exploited without violating static control flow and data flow, existing CFI and DFI solutions cannot prevent them. We present CEFI , the first method to enforce integrity of command execution on embedded devices after deployment. CEFI provides fine-grained Command Execution Flow Integrity by preventing external commands from being executed on control flow paths belonging to interaction channels that are not authorized to perform them. Using minimal manual annotations as a starting point, CEFI statically determined the legal path set (from the start to the end point) and instruments the program to verify the legitimacy of the command execution at runtime by checking whether the calling context is consistent between the runtime executed path and statically obtained legal path set. We evaluate our prototype with five real-world firmware samples, and show that CEFI has an average performance overhead of just 0.18%, an average memory overhead of 0.19%, and that CEFI can effectively protect embedded devices against attacks on privilege separation vulnerabilities even if they do not violate control flow.

AB - As embedded devices are widely used in increasingly complex settings (e.g., smart homes and industrial control systems), one device is usually connected with multiple entities, such as mobile apps and the cloud. Recent research has shown that privilege separation vulnerabilities, which allow violations of authority between different entities, are occuring in IoT systems. Because such vulnerabilities can be exploited without violating static control flow and data flow, existing CFI and DFI solutions cannot prevent them. We present CEFI , the first method to enforce integrity of command execution on embedded devices after deployment. CEFI provides fine-grained Command Execution Flow Integrity by preventing external commands from being executed on control flow paths belonging to interaction channels that are not authorized to perform them. Using minimal manual annotations as a starting point, CEFI statically determined the legal path set (from the start to the end point) and instruments the program to verify the legitimacy of the command execution at runtime by checking whether the calling context is consistent between the runtime executed path and statically obtained legal path set. We evaluate our prototype with five real-world firmware samples, and show that CEFI has an average performance overhead of just 0.18%, an average memory overhead of 0.19%, and that CEFI can effectively protect embedded devices against attacks on privilege separation vulnerabilities even if they do not violate control flow.

UR - https://www.scopus.com/pages/publications/85164263064

UR - https://www.scopus.com/pages/publications/85164263064#tab=citedBy

U2 - 10.1007/978-3-031-35504-2_12

DO - 10.1007/978-3-031-35504-2_12

M3 - Conference contribution

SN - 9783031355035

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

SP - 235

EP - 255

BT - Detection of Intrusions and Malware, and Vulnerability Assessment

A2 - Gruss, Daniel

A2 - Maggi, Federico

A2 - Fischer, Mathias

A2 - Carminati, Michele

PB - Springer Nature

ER -

Peng A, Fang D, Zhou W, van der Kouwe E, Li Y, Zhang Y. CEFI: Command Execution Flow Integrity for Embedded Devices. In Gruss D, Maggi F, Fischer M, Carminati M, editors, Detection of Intrusions and Malware, and Vulnerability Assessment: 20th International Conference, DIMVA 2023, Hamburg, Germany, July 12–14, 2023, Proceedings. Springer Nature. 2023. p. 235-255. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). Epub 2023 Jul 12. doi: 10.1007/978-3-031-35504-2_12

CEFI: Command Execution Flow Integrity for Embedded Devices

Abstract

Publication series

Funding

Access to Document

Persistent URL (handle)

Other files and links

Fingerprint

Cite this