The research of the visiting scientists focuses on device identification using deep-learning-based RF fingerprinting. Instead of identifying devices based on assigned addresses, this approach relies on characteristic radio frequency properties. These properties result from minimal and unavoidable hardware variations that occur during manufacturing or operation, giving each transmitter a unique signature.

The methods developed by the visiting researchers had previously been validated under static conditions. However, their reliability under realistic scenarios involving moving transmitters had not yet been demonstrated. To address this, Long Range Wide Area Network (LoRaWAN) transmitters were mounted on a UAV and flown along various trajectories. The collected measurement data is now being analyzed to evaluate the robustness of the proposed methods.

We would like to thank the visiting researchers from the University of Liverpool as well as the DKS for the successful collaboration.

Overview