A Multi-Physics Drift Model for Robust Radio Frequency Fingerprinting: From Theory to Application
IEEE Communications Letters · 2026 · 30 · 2530-2534
DOI: 10.1109/lcomm.2026.3710940 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Communications Letters, 2026. Verfasst von Zichuan Yu und 3 weiteren Autor:innen. Die Arbeit nennt 19 Literaturangaben. Gefördert von: Open Fund of the Key Laboratory of Short-Range Radio Equipment Testing and Evaluation of the Ministry of Industry and Information Technology.
Abstract (Originalsprache)
Radio Frequency Fingerprint Identification (RFFI) is severely bottlenecked by physical fingerprint drift. Moving beyond treating this drift as a black-box statistical shift, we propose a comprehensive multi-physics theoretical model that rigorously decouples macro-scale thermodynamics, meso-scale stress relaxation, and micro-scale circuit modulation. As a direct application and empirical validation of this theory, we introduce the Universal Physical Drift Data Augmentation (Uni-PDA) framework. Operating purely via target-free tensor transformations, Uni-PDA injects deterministic physical drift boundaries into a single steady-state dataset. Extensive experiments across diverse architectures demonstrate that this theory-guided synthesis improves cross-drift generalization without target-domain data or additional inference-stage overhead. When integrated with Decoupling-RFFI, Uni-PDA achieves 88.8% drift-domain accuracy while maintaining 93.9% steady-state accuracy, reducing the absolute accuracy drop from 14.1 to 5.1 percentage points.
Förderung
- Open Fund of the Key Laboratory of Short-Range Radio Equipment Testing and Evaluation of the Ministry of Industry and Information Technology
Zitieren
APA
Yu, Z., Ge, H., Tang, X., & Liu, H. (2026). A Multi-Physics Drift Model for Robust Radio Frequency Fingerprinting: From Theory to Application. IEEE Communications Letters, 30, 2530-2534. https://doi.org/10.1109/lcomm.2026.3710940
Vancouver
Yu Z, Ge H, Tang X, Liu H. A Multi-Physics Drift Model for Robust Radio Frequency Fingerprinting: From Theory to Application. IEEE Communications Letters. 2026;30:2530-2534. doi:10.1109/lcomm.2026.3710940
BibTeX
@article{Yu2026,
author = {Yu, Zichuan and Ge, Hongyu and Tang, Xusheng and Liu, Hongyuan},
title = {A Multi-Physics Drift Model for Robust Radio Frequency Fingerprinting: From Theory to Application},
journal = {IEEE Communications Letters},
year = {2026},
volume = {30},
pages = {2530--2534},
doi = {10.1109/lcomm.2026.3710940}
}
EUVIMED