Keep Private Networks Private: Secure Channel-PUFs, and Physical Layer Security by Linear Regression Enhanced Channel Profiles

Christoph Lipps, Sachinkumar Bavikatti Mallikarjun, Mathias Strufe, Christopher Heinz, Christoph Grimm, Hans Dieter Schotten

In: Proceedings of the 3rd International Conference on Data Intelligence and Security. International Conference on Data Intelligence and Security (ISDIS-2020) November 10-12 South Padre Island Texas United States Seiten 93-100 ISBN 978-1-7281-9379-3/20 IEEEXplore 11/2020.


In the context of a rapidly changing and increasingly complex (industrial) production landscape, securing the (commu- nication) infrastructure is becoming an ever more important but also more challenging task - accompanied by the application of radio communication. A worthwhile and promising approach to overcome the arising attack vectors, and to keep private networks private, are Physical Layer Security (PhySec) implementations. The paper focuses on the transfer of the IEEE802.11 (WLAN) PhySec - Secret Key Generation (SKG) algorithms to Next Gen- eration Mobile Networks (NGMNs), as they are the driving forces and key enabler of future industrial networks. Based on a real world Long Term Evolution (LTE) testbed, improvements of the SKG algorithms are validated. The paper presents and evaluates significant improvements in the establishment of channel profiles, whereby especially the Bit Disagreement Rate (BDR) can be improved substantially. The combination of the Discrete Cosine Transformation (DCT) and the supervised Machine Learning (ML) algorithm - Linear Regression (LR) – provides outstanding results, which can be used beyond the SKG application. The evaluation also emphasizes the appropriateness of PhySec for securing private networks.


Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence