Noise-Robust Radio Frequency Fingerprint Identification Using Denoise Diffusion Model

Guolin Yin; Junqing Zhang; Yuan Ding; Simon L. Cotton

Noise-Robust Radio Frequency Fingerprint Identification Using Denoise Diffusion Model

Guolin Yin, Junqing Zhang^*, Yuan Ding, Simon L. Cotton

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Securing Internet of Things (IoT) devices presents increasing challenges due to their limited computational and energy resources. Radio Frequency Fingerprint Identification (RFFI) emerges as a promising authentication technique to identify wireless devices through hardware impairments. RFFI performance under low signal-to-noise ratio (SNR) scenarios is significantly degraded because the minute hardware features can be easily swamped in noise. In this paper, we leveraged the diffusion model to effectively restore the RFF under low SNR scenarios. Specifically, we trained a powerful noise predictor and tailored a noise removal algorithm to effectively reduce the noise level in the received signal and restore the device fingerprints. We used Wi-Fi as a case study and created a testbed involving 6 commercial off-the-shelf Wi-Fi dongles and a USRP N210 software-defined radio (SDR) platform. We conducted experimental evaluations on various SNR scenarios. The experimental results show that the proposed algorithm can improve the classification accuracy by up to 34.9%.

Original language	English
Title of host publication	IEEE Wireless Communications and Networking Conference 2025
Publication status	Accepted/In press - 27 Jan 2025
Event	IEEE Wireless Communications and Networking Conference 2025 - Milan, Italy Duration: 24 Mar 2025 → 27 Mar 2025 https://wcnc2025.ieee-wcnc.org/

Conference

Conference	IEEE Wireless Communications and Networking Conference 2025
Abbreviated title	WCNC 2025
Country/Territory	Italy
City	Milan
Period	24/03/25 → 27/03/25
Internet address	https://wcnc2025.ieee-wcnc.org/

Keywords

Denoising
Diffusion Model
Radio Frequency Fingerprint Identification (RFFI)
Transformer Wi-Fi

Embargoed Document

Download pdf
Accepted author manuscript, 917 KB
Embargo ends: 27/03/25

Cite this

@inproceedings{3114c7efa07e44a3a4217cb81104cfc4,

title = "Noise-Robust Radio Frequency Fingerprint Identification Using Denoise Diffusion Model",

abstract = "Securing Internet of Things (IoT) devices presents increasing challenges due to their limited computational and energy resources. Radio Frequency Fingerprint Identification (RFFI) emerges as a promising authentication technique to identify wireless devices through hardware impairments. RFFI performance under low signal-to-noise ratio (SNR) scenarios is significantly degraded because the minute hardware features can be easily swamped in noise. In this paper, we leveraged the diffusion model to effectively restore the RFF under low SNR scenarios. Specifically, we trained a powerful noise predictor and tailored a noise removal algorithm to effectively reduce the noise level in the received signal and restore the device fingerprints. We used Wi-Fi as a case study and created a testbed involving 6 commercial off-the-shelf Wi-Fi dongles and a USRP N210 software-defined radio (SDR) platform. We conducted experimental evaluations on various SNR scenarios. The experimental results show that the proposed algorithm can improve the classification accuracy by up to 34.9%.",

keywords = "Denoising, Diffusion Model, Radio Frequency Fingerprint Identification (RFFI), Transformer Wi-Fi",

author = "Guolin Yin and Junqing Zhang and Yuan Ding and Cotton, {Simon L.}",

year = "2025",

month = jan,

day = "27",

language = "English",

booktitle = "IEEE Wireless Communications and Networking Conference 2025",

note = "IEEE Wireless Communications and Networking Conference 2025, WCNC 2025 ; Conference date: 24-03-2025 Through 27-03-2025",

url = "https://wcnc2025.ieee-wcnc.org/",

}

TY - GEN

T1 - Noise-Robust Radio Frequency Fingerprint Identification Using Denoise Diffusion Model

AU - Yin, Guolin

AU - Zhang, Junqing

AU - Ding, Yuan

AU - Cotton, Simon L.

PY - 2025/1/27

Y1 - 2025/1/27

N2 - Securing Internet of Things (IoT) devices presents increasing challenges due to their limited computational and energy resources. Radio Frequency Fingerprint Identification (RFFI) emerges as a promising authentication technique to identify wireless devices through hardware impairments. RFFI performance under low signal-to-noise ratio (SNR) scenarios is significantly degraded because the minute hardware features can be easily swamped in noise. In this paper, we leveraged the diffusion model to effectively restore the RFF under low SNR scenarios. Specifically, we trained a powerful noise predictor and tailored a noise removal algorithm to effectively reduce the noise level in the received signal and restore the device fingerprints. We used Wi-Fi as a case study and created a testbed involving 6 commercial off-the-shelf Wi-Fi dongles and a USRP N210 software-defined radio (SDR) platform. We conducted experimental evaluations on various SNR scenarios. The experimental results show that the proposed algorithm can improve the classification accuracy by up to 34.9%.

AB - Securing Internet of Things (IoT) devices presents increasing challenges due to their limited computational and energy resources. Radio Frequency Fingerprint Identification (RFFI) emerges as a promising authentication technique to identify wireless devices through hardware impairments. RFFI performance under low signal-to-noise ratio (SNR) scenarios is significantly degraded because the minute hardware features can be easily swamped in noise. In this paper, we leveraged the diffusion model to effectively restore the RFF under low SNR scenarios. Specifically, we trained a powerful noise predictor and tailored a noise removal algorithm to effectively reduce the noise level in the received signal and restore the device fingerprints. We used Wi-Fi as a case study and created a testbed involving 6 commercial off-the-shelf Wi-Fi dongles and a USRP N210 software-defined radio (SDR) platform. We conducted experimental evaluations on various SNR scenarios. The experimental results show that the proposed algorithm can improve the classification accuracy by up to 34.9%.

KW - Denoising

KW - Diffusion Model

KW - Radio Frequency Fingerprint Identification (RFFI)

KW - Transformer Wi-Fi

M3 - Conference contribution

BT - IEEE Wireless Communications and Networking Conference 2025

T2 - IEEE Wireless Communications and Networking Conference 2025

Y2 - 24 March 2025 through 27 March 2025

ER -