太赫兹科学与电子信息学报2026，Vol.24Issue(1)：98-106,9.DOI:10.11805/TKYDA2024572
一种用于特定辐射源识别的无监督域适应方法
An unsupervised domain adaptation method for specific emitter identification
摘要
Abstract
The rapid development of mobile communication technology has generated abundant unlabeled radio source signals.To fully utilize unlabeled data,this paper proposes an Independence Criterion-based Unsupervised Domain Adaptation(ICUDA)method for specific emitter identification.The independence criterion is employed to measure the similarity between the source domain and the target domain,and combined with an improved convolutional neural network to transfer knowledge from the source domain to the target domain,thereby helping improve the classification performance of the target domain that contains only unlabeled data.Under seven transfer scenarios constructed based on Software-Defined Radio(SDR)dataset collected in a laboratory environment,compared with baseline methods and three unsupervised domain adaptation methods,the proposed method achieves the best classification performance in the target domain across all scenarios,with an average recognition accuracy of 84.2%,demonstrating that the proposed method can extract features with good inter-class separability and intra-class compactness on the target domain,effectively reducing the target domain's dependence on high-quality labeled data.关键词
辐射源个体识别/无监督域适应/相似性度量/希尔伯特-施密特独立性准则Key words
Specific Emitter Identification(SEI)/Unsupervised Domain Adaptation(UDA)/similarity measure/Hilbert-Schmidt Independence Criterion(HSIC)分类
信息技术与安全科学引用本文复制引用
吴琼,李志刚,史继博,王谦,查浩然..一种用于特定辐射源识别的无监督域适应方法[J].太赫兹科学与电子信息学报,2026,24(1):98-106,9.基金项目
国家自然科学基金资助项目(62201172) （62201172）
国家重点研发计划基金资助项目(2022YFE0136800) （2022YFE0136800）
