AARC-FE: Electrical Assembly Authentication with Random Convolution Kernels and Fuzzy Extractors

Spinnler C, Reißland T, Franchi N (2025)

Publication Language: English

Publication Type: Conference contribution, Conference Contribution

Publication year: 2025

Event location: Ghent University

Open Access Link: https://link.springer.com/chapter/10.1007/978-3-032-00624-0_13

Abstract

Identity management of devices in the Internet of Things (IoT) has become an essential part of a secure IoT infrastructure. Enrollment and authentication is performed based on public key infrastructure (PKI) in state-of-the-art deployments. This ensures device authenticity based on certificates. To mitigate the risk of compromised certificates and to strengthen the security, the certificates can be created on the device during enrollment based on physical properties, such as a physical unclonable function (PUF) of the device or a connected secure element.

To add further physical properties to the device identity, we propose AARC-FE: Electrical Assemblie Authentication with Random Convoultion Kernels and Fuzzy Extractors. This is a new approach for hardware fingerprint creation of a serial communication interface. The proposed solution uses characteristics of the analog signal introduced by manufacturing variances, which are made detectable by using random convolutional kernels. The features produced by the convolution are transformed in a two-step approach to suit the requirements of a fuzzy extractor that is used to create strong keys for the observed communication interface.

We create a stochastic model for evaluating the analog domain and transfer this model to a SPICE-based simulation of a common communication bus: I2C. To characterize the approach, metrics for PUFs and authentication systems are applied.

The evaluation shows, that AARC-FE is a feasible approach for key generation of an electical assembly. Depending on the chosen parameters, the authentication system achieves an equal error rate as low as 0.09.

With this approach, it is possible to detect attacks such as device swapping or sniffing attacks on a serial communication bus. The system derives asymmetric keys from the analog values of the communication bus and can thus participate in standard public key authentication schemes.

Authors with CRIS profile

Christian Spinnler Lehrstuhl für Intelligente Technische Elektronik und Systeme Torsten Reißland Lehrstuhl für Intelligente Technische Elektronik und Systeme Norman Franchi Lehrstuhl für Intelligente Technische Elektronik und Systeme

How to cite

APA:

Spinnler, C., Reißland, T., & Franchi, N. (2025). AARC-FE: Electrical Assembly Authentication with Random Convolution Kernels and Fuzzy Extractors. In Proceedings of the The 20th International Conference on Availability, Reliability and Security (ARES 2025). Ghent University, BE.

MLA:

Spinnler, Christian, Torsten Reißland, and Norman Franchi. "AARC-FE: Electrical Assembly Authentication with Random Convolution Kernels and Fuzzy Extractors." Proceedings of the The 20th International Conference on Availability, Reliability and Security (ARES 2025), Ghent University 2025.

BibTeX: Download