Targeted Adversarial Attacks Against Deep Learning-Based Wideband Spectrum Sensing
IEEE Transactions on Machine Learning in Communications and Networking · 2026 · 4 · 950-965
DOI: 10.1109/tmlcn.2026.3694666 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Machine Learning in Communications and Networking, 2026. Verfasst von Shilian Zheng und 4 weiteren Autor:innen. Die Arbeit nennt 37 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: National Key Research and Development Program of China.
Abstract (Originalsprache)
Deep learning-based wideband spectrum sensing models have achieved impressive performance in subband occupancy detection, but their susceptibility to adversarial attacks raises serious concerns. In this paper, we construct a targeted adversarial attack framework specifically designed for multi-label wideband spectrum sensing tasks. We further propose two optimization-based attack methods. MLS_CW formulates perturbation generation as a constrained optimization problem and solves it via gradient descent. MLS_DF is a greedy iterative method that approximates the solution to an underdetermined linear system through linearization. Both attacks are developed under strict power constraints and are capable of generating subtle yet highly effective perturbations. These attacks further incorporate a threshold-aware loss and selective subband targeting to enhance precision and stealthiness. Extensive experiments across various SIR conditions demonstrate that the proposed attacks significantly degrade the model’s detection performance in both untargeted and targeted scenarios, while maintaining high stealthiness. These results expose the severe vulnerability of DL-based spectrum sensing models and highlight the need for further research into their robustness under adversarial settings.
Förderung
- National Key Research and Development Program of China
Zitieren
APA
Zheng, S., Ye, Z., Zhang, L., Yue, K., & Zhao, Z. (2026). Targeted Adversarial Attacks Against Deep Learning-Based Wideband Spectrum Sensing. IEEE Transactions on Machine Learning in Communications and Networking, 4, 950-965. https://doi.org/10.1109/tmlcn.2026.3694666
Vancouver
Zheng S, Ye Z, Zhang L, Yue K, Zhao Z. Targeted Adversarial Attacks Against Deep Learning-Based Wideband Spectrum Sensing. IEEE Transactions on Machine Learning in Communications and Networking. 2026;4:950-965. doi:10.1109/tmlcn.2026.3694666
BibTeX
@article{Zheng2026,
author = {Zheng, Shilian and Ye, Zhihao and Zhang, Luxin and Yue, Keqiang and Zhao, Zhijin},
title = {Targeted Adversarial Attacks Against Deep Learning-Based Wideband Spectrum Sensing},
journal = {IEEE Transactions on Machine Learning in Communications and Networking},
year = {2026},
volume = {4},
pages = {950--965},
doi = {10.1109/tmlcn.2026.3694666}
}
EUVIMED