Germany-UK: Securing Device Identification Using Hardware Impairments in RF Circuits

Overview

Please visit the UKRI project page for more information.

Abstract

Wireless communications have revolutionised everyday life and industry by enabling ubiquitous connectivity across billions of devices, estimated at 21.5 billion in 2025, and powering transformative applications such as smart homes, connected healthcare, and autonomous driving. Traditional wireless infrastructures primarily rely on cryptographic methods for security; however, these approaches often face significant limitations when applied to low-cost or resource-constrained devices, where affordability and accessibility are critical concerns.

Radio frequency fingerprinting identification (RFFI) is an emerging physical-layer security technique that offers a non-cryptographic approach to securing current and future telecommunication infrastructure. Due to inevitable variations in RF component characteristics introduced during semiconductor manufacturing, each RF circuit exhibits unique hardware impairments. These inherent imperfections can be leveraged for secure device identification. While existing RFFI research has largely focused on improving identification performance, it often overlooks the systematic modelling and deeper understanding of these intrinsic hardware impairments, leading to predominantly empirical algorithm design.

This project brings together leading experts from the University of Liverpool (UK) and Ruhr-Universität Bochum (Germany) to conduct a comprehensive investigation into RFFI, with the aim of gaining a deep understanding of the origins of hardware impairments in RF circuits. A synergistic methodology will be employed, combining impairments modelling, simulation, and experimental validation. The entire RF chain will be modelled, with particular emphasis on power amplifiers, as their pronounced nonlinearity is widely regarded as a primary source of RF impairments. The impairments will be initially modelled in MATLAB, followed by the development of a testbed using software-defined radio (SDR) platforms and power amplifier evaluation modules. Extensive simulations and experimental measurements will be conducted, with results used to iteratively refine the models. This holistic approach marks a significant departure from the empirical methods prevalent in existing research, enabling optimal modelling and the principled design of RFFI algorithms.