HEEDFUL: Leveraging Sequential Transfer Learning for Robust WiFi Device Fingerprinting Amid Hardware Warm-Up Effects

TL;DR

This work investigates how hardware warm-up undermines deep-learning RF fingerprinting and introduces HEEDFUL, a robust framework that leverages sequential transfer learning and explicit impairment estimation to stabilize device identification across warm-up and stable phases. HEEDFUL learns a shared representation by first estimating eight hardware impairments from time-domain I/Q data and then transferring this knowledge to a device classifier, achieving up to 96% accuracy within the initial 6 minutes and strong cross-day and cross-protocol performance. The authors provide extensive experiments on WiFi 802.11b and 802.11n data from 15 Pycom devices, including wired and wireless channels, and release new impairment-annotated RF fingerprinting datasets. Overall, HEEDFUL demonstrates that grounding learning in hardware impairments can dramatically improve robustness and generalization of RF fingerprinting in realistic, dynamic environments, with potential applicability to drones and mobile devices while highlighting directions for explainable, semi-supervised, and open-set extensions.

Abstract

Deep Learning-based RF fingerprinting approaches struggle to perform well in cross-domain scenarios, particularly during hardware warm-up. This often-overlooked vulnerability has been jeopardizing their reliability and their adoption in practical settings. To address this critical gap, in this work, we first dive deep into the anatomy of RF fingerprints, revealing insights into the temporal fingerprinting variations during and post hardware stabilization. Introducing HEEDFUL, a novel framework harnessing sequential transfer learning and targeted impairment estimation, we then address these challenges with remarkable consistency, eliminating blind spots even during challenging warm-up phases. Our evaluation showcases HEEDFUL's efficacy, achieving remarkable classification accuracies of up to 96% during the initial device operation intervals-far surpassing traditional models. Furthermore, cross-day and cross-protocol assessments confirm HEEDFUL's superiority, achieving and maintaining high accuracy during both the stable and initial warm-up phases when tested on WiFi signals. Additionally, we release WiFi type B and N RF fingerprint datasets that, for the first time, incorporate both the time-domain representation and real hardware impairments of the frames. This underscores the importance of leveraging hardware impairment data, enabling a deeper understanding of fingerprints and facilitating the development of more robust RF fingerprinting solutions.

Figures (20)

• Figure 1: I/Q signal behavior observed at different times during the device's initial operation period.
• Figure 2: IoT Testbed consisting of 15 Pycom transmitting devices and a USRP B210 receiving device
• Figure 3: Exp. 1: training & testing on Warm-up captures
• Figure 4: Exp. 2: training & testing on Stable captures
• Figure 5: Exp. 3: training on Stable captures and testing on Warm-up captures