Few-Shot Recognition and Classification Framework for Jamming Signal: A CGAN-Based Fusion CNN Approach

TY - JOUR

T1 - Few-Shot Recognition and Classification Framework for Jamming Signal

T2 - A CGAN-Based Fusion CNN Approach

AU - Ding, Xuhui

AU - Lu, Qi

AU - Zhang, Yue

AU - Li, Gaoyang

AU - Gao, Xiaozheng

AU - Ye, Neng

AU - Niyato, Dusit

AU - Yang, Kai

N1 - Publisher Copyright: © 1967-2012 IEEE.

PY - 2026

Y1 - 2026

N2 - Subject to intricate environmental variables in global navigation satellite systems (GNSS), the precise recognition of jamming signals holds paramount significance in the effective implementation of anti-jamming strategies. In light of this imperative, we propose an innovative fusion method based on conditional generative adversarial network (CGAN) and convolutional neural network (CNN), which aims to deal with the difficulty in applying deep learning (DL) algorithms due to the instantaneous nature of jamming signals in practical communication systems. Compared with previous methods, our method embeds jamming category labels to constrain the range of generated signals in the frequency domain by using the CGAN model, which simultaneously captures label-conditioned information while learning the distribution of signal data thus achieving an 8% improvement in accuracy over a semi-supervised GAN baseline even when working with a few-shot dataset. Real-world satellite communication scenarios are simulated by adopting a hardware platform, and we validate the proposed method by using the resulting time-domain waveform data. The experimental results indicate that our method performs exceptionally well, demonstrating significant potential for practical application in GNSS.

AB - Subject to intricate environmental variables in global navigation satellite systems (GNSS), the precise recognition of jamming signals holds paramount significance in the effective implementation of anti-jamming strategies. In light of this imperative, we propose an innovative fusion method based on conditional generative adversarial network (CGAN) and convolutional neural network (CNN), which aims to deal with the difficulty in applying deep learning (DL) algorithms due to the instantaneous nature of jamming signals in practical communication systems. Compared with previous methods, our method embeds jamming category labels to constrain the range of generated signals in the frequency domain by using the CGAN model, which simultaneously captures label-conditioned information while learning the distribution of signal data thus achieving an 8% improvement in accuracy over a semi-supervised GAN baseline even when working with a few-shot dataset. Real-world satellite communication scenarios are simulated by adopting a hardware platform, and we validate the proposed method by using the resulting time-domain waveform data. The experimental results indicate that our method performs exceptionally well, demonstrating significant potential for practical application in GNSS.

KW - communication anti-jamming

KW - convolutional neural network

KW - generative adversarial network

KW - jamming signal classification

KW - neural network algorithm

UR - https://www.scopus.com/pages/publications/105034755931

U2 - 10.1109/TVT.2026.3678290

DO - 10.1109/TVT.2026.3678290

M3 - Article

AN - SCOPUS:105034755931

SN - 0018-9545

JO - IEEE Transactions on Vehicular Technology

JF - IEEE Transactions on Vehicular Technology

ER -