ASKNet: Adaptive Spectral Koopman Network for Cross-Domain Radio Frequency Fingerprinting
IEEE Communications Letters · 2026 · 30 · 2825-2829
DOI: 10.1109/lcomm.2026.3718894 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Communications Letters, 2026. Verfasst von Zhengyang Liu und 2 weiteren Autor:innen. Die Arbeit nennt 21 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Basic Research Program of Jiangsu Province.
Abstract (Originalsprache)
In this letter, we propose the adaptive spectral Koopman network (ASKNet) for cross-domain radio frequency fingerprinting (RFF) under channel and receiver variations. Spectral analysis shows that, in the low-frequency band, such variations mainly reduce to a uniform gain and a linear phase, making low-frequency components relatively robust for RFF. Building on this insight, ASKNet first employs a learnable spectral gating (LSG) module to adaptively localize the signal to the domain-robust low-frequency band, and then applies a spectral Koopman operator (SKO), parameterized as a unitary transformation, to compensate for the residual linear phase. The refined signal is finally fed into a classifier for device identification. Experiments on the WiSig and LoRa datasets demonstrate that ASKNet outperforms representative baselines with significantly fewer parameters and lower computational cost. The code is available at https://github.com/BeechburgPieStar/ASKNet-RFF
Förderung
- National Natural Science Foundation of China
- Basic Research Program of Jiangsu Province
Zitieren
APA
Liu, Z., Wang, Z., & Wang, Y. (2026). ASKNet: Adaptive Spectral Koopman Network for Cross-Domain Radio Frequency Fingerprinting. IEEE Communications Letters, 30, 2825-2829. https://doi.org/10.1109/lcomm.2026.3718894
Vancouver
Liu Z, Wang Z, Wang Y. ASKNet: Adaptive Spectral Koopman Network for Cross-Domain Radio Frequency Fingerprinting. IEEE Communications Letters. 2026;30:2825-2829. doi:10.1109/lcomm.2026.3718894
BibTeX
@article{Liu2026,
author = {Liu, Zhengyang and Wang, Zheyu and Wang, Yu},
title = {ASKNet: Adaptive Spectral Koopman Network for Cross-Domain Radio Frequency Fingerprinting},
journal = {IEEE Communications Letters},
year = {2026},
volume = {30},
pages = {2825--2829},
doi = {10.1109/lcomm.2026.3718894}
}
EUVIMED