通信学报2026，Vol.47Issue(3)：15-29,15.DOI:10.11959/j.issn.1000-436x.2026030
面向自动调制分类的域自适应对抗防御方法
Domain adaptive adversarial defense method for automatic modulation classification
摘要
Abstract
Deep learning-based automatic modulation classification(AMC)models are vulnerable to adversarial example attacks,posing severe adversarial security threats in practical scenarios characterized by dynamic channel conditions and limited availability of signal labels.To address this issue,a multi-domain distribution alignment based on domain-adaptive adversarial defense method was proposed.Firstly,a phase rotation data augmentation strategy was used to en-rich the discriminative and domain-invariant features learned by the model.Secondly,a dual-discriminator architecture was constructed to reduce the feature distribution discrepancy between the original and adversarial signals in the target domain and those in the source domain.Thirdly,contrastive learning constraints were introduced in conjunction with high-confidence pseudo-labels,leveraging source domain class anchor to enhance intra-class compactness and inter-class separability in the target domain.Finally,a consistency constraint was employed to reduce the output discrepancy be-tween original and adversarial signals in the target domain.Experimental results on both public and simulated datasets demonstrate that,compared with existing methods,the proposed method exhibits superior domain adaptability and adver-sarial robustness under various adversarial attacks,effectively enhancing the reliability and security of AMC systems in complex electromagnetic environments.关键词
频谱监测/自动调制分类/无监督域自适应/对抗鲁棒性Key words
spectrum monitoring/automatic modulation classification/unsupervised domain adaptation/adversarial robustness分类
信息技术与安全科学引用本文复制引用
杨研蝶,林云,徐路平,张思成,李奎贤,韩宇..面向自动调制分类的域自适应对抗防御方法[J].通信学报,2026,47(3):15-29,15.基金项目
中央高校基本科研业务费专项资金资助项目(No.3072025YY0801) （No.3072025YY0801）
黑龙江省博士后基金资助项目(No.3236340036) （No.3236340036）
国家自然科学基金资助项目(No.62201172) The Fundamental Research Funds for the Central Universities(No.3072025YY0801),Heilongjiang Postdoctoral Fund(No.3236340036),The National Natural Science Foundation of China(No.62201172) （No.62201172）
