Gaussian low-pass channel attention convolution network for RF fingerprinting

Radio frequency (RF) fingerprinting is a challenging and important technique for individual identification of wireless devices. Recent work has applied deep learning-based classifiers to ADS-B signals without missing aircraft ID information. However, traditional methods are not very effective in achieving high accuracy for deep learning models to recognize RF signals. In this letter, a Gaussian low-pass channel attention convolution network, which uses a Gaussian low-pass channel attention module (GLCAM) to extract fingerprint features with low frequency. Specifically, in GLCAM, a frequency-convolutional global average pooling module is designed to help the channel attention mechanism learn channel weights in the frequency domain. Experimental results on large-scale real-world ADS-B signal datasets show that the method can achieve an accuracy of 92.08%, which is 6.21% higher than convolutional neural networks.

Automated summary

RF fingerprinting is a challenging and important technique for wireless device identification. Recent research has applied deep learning-based classifiers to ADS-B signals without losing aircraft ID information. However, traditional methods are not effective in achieving high accuracy for deep learning models to recognize RF signals. In this study, a Gaussian low-pass channel attention convolution network is proposed, which uses a Gaussian low-pass channel attention module (GLCAM) to extract low-frequency fingerprint features. Experimental results demonstrate that the method achieves an accuracy of 92.08%, which is 6.21% higher than convolutional neural networks.