IICoSeP

Increasing Industrial Communication Security by Physical Layer Security

The ever-increasing networking of different devices, sensors, actuators and many other components to form an Internet of Things (IoT) is further enhanced by novel application scenarios such as Machine-to-Machine (M2M) and Machine-to-Service (M2S) communication. Cyber-physical production systems (CPPS) and industrial automation and control systems (I-ACS) are becoming more flexible and dynamic through increased interconnectivity and embedded systems. Added to this are the increasing demands on information and communication technology for scalability, security and reliability. It is also the enabler to a multitude of new digital services and applications, many of which we cannot even imagine today. Especially in times of limited mobility, it is becoming clear that network security and integrity are essential to guarantee network operation and configure services reliably. However, many of the integrated components do not have sufficient security mechanisms to withstand the attack surface that is also growing as a result. Many of the deployed components are limited in terms of resources, do not have enough memory, energy or computing power to apply established and powerful cryptography.

The project "Increasing Industrial Communication Security by Physical Layer Security" (IICoSeP) addresses the security, integrity and confidentiality of this communication infrastructure by methods of Physical Layer Security (PLS).  Here, existing methods are developed in the areas of wireless and wired PLS. In the wireless domain, PLS methods are used to secure private, corporate, and campus 5G networks, among others. One starting point is the change from the usual information-theory-driven PLS to a signal-processing approach. The focus in wireline PLS is to explore novel Physically Unclonable Functions (PUFs) that are being developed for use in freely available hardware. The goal is to equip Industrie 4.0 environments with freely available PUFs to establish secure communication among IoT devices protected by a shared secret among participants. In both application areas, the use of machine learning to increase performance is being evaluated. The IICoSeP project will develop demonstration environments for mapping cross-technology communication and integrate them into an overall demonstrator. The demonstration environments will further be used for the generation of training data and the adaptation of machine learning methods.

Project Partners: German Research Center for Artificial Intelligence (DFKI), Technical University of Dresden (TUD), KMPC Innovations GmbH