Research Article
CAP-GAN: Channel-Aware Perturbation for Secure Wireless Transmission
@INPROCEEDINGS{10.1007/978-3-032-33662-0_14, author={Menghao Chen and Zhisheng Yin and Nan Cheng and Zhijie Zhang and Yusi Zhang}, title={CAP-GAN: Channel-Aware Perturbation for Secure Wireless Transmission}, proceedings={Communications and Networking. 20th EAI International Conference, ChinaCom 2025, Xi’an, China, November 8--9, 2025, Proceedings}, proceedings_a={CHINACOM}, year={2026}, month={10}, keywords={channel-aware adversarial perturbation Automatic Modulation Recognition GAN}, doi={10.1007/978-3-032-33662-0_14} }- Menghao Chen
Zhisheng Yin
Nan Cheng
Zhijie Zhang
Yusi Zhang
Year: 2026
CAP-GAN: Channel-Aware Perturbation for Secure Wireless Transmission
CHINACOM
Springer
DOI: 10.1007/978-3-032-33662-0_14
Abstract
This paper proposes a novel channel-aware adversarial perturbation generation framework, called CAP-GAN (Channel-Aware Perturbation GAN). The framework enables active defense in wireless communication systems with potential eavesdroppers. Unlike traditional receiver-side defenses that passively remove perturbations, CAP-GAN actively injects them into the original transmitted signal. Using the original signal and the channel responses of both Bob and Eve as inputs, the framework generates channel-aware perturbations that preserve reception quality of Bob while deceiving Eve. Based on the GAN framework, CAP-GAN is trained by jointly optimizing a multi-objective loss function. Simulation results verify the effectiveness of the proposed CAP-GAN framework, which provides an efficient active defense strategy for secure wireless transmission.