Radio frequency fingerprint recognition based on deep transfer learning

Yuanhui Wu; Li Hao

doi:10.1117/12.2640233

9 May 2022 Radio frequency fingerprint recognition based on deep transfer learning

Yuanhui Wu, Li Hao

Proceedings Volume 12252, International Conference on Biometrics, Microelectronic Sensors, and Artificial Intelligence (BMSAI); 122520K (2022) https://doi.org/10.1117/12.2640233
Event: International Conference on Biometrics, Microelectronic Sensors, and Artificial Intelligence (BMSAI), 2022, Guangzhou, China

Abstract

Radio frequency (RF) fingerprint refers to the inevitable subtle defects in the hardware circuit of radio transmitter, which are reflected in the signal waveform transmitted by it. Therefore, the wireless device can be uniquely identified by analyzing the received signal waveform. Due to the advantages of multidimensional mapping of convolutional neural network (CNN), it can automatically extract signal features and transmitter nonlinear features, which can not be extracted by traditional low-dimensional algorithms. Therefore, it is very important to design a deep neural network with reasonable structure and high recognition accuracy. Taking the in phase and quadrature (I & Q) time domain signals of eight radiation sources of the same model as the input samples, the fine features of IQ time domain signals are extracted by designing six deep learning radiation source individual recognition algorithms with different structures to realize the identification of RF fingerprint. Besides, the change of wireless channel probably weakens the identification robustness of RF fingerprint. Thus, a deep transfer learning algorithm is proposed to cross train and test the I & Q data collected at different times, which solves the problem of poor robustness.

Citation Download Citation

Yuanhui Wu and Li Hao "Radio frequency fingerprint recognition based on deep transfer learning", Proc. SPIE 12252, International Conference on Biometrics, Microelectronic Sensors, and Artificial Intelligence (BMSAI), 122520K (9 May 2022); https://doi.org/10.1117/12.2640233

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available