Few-shot electromagnetic signal classification: A data union augmentation method
ZHOU Huaji,BAI Jing,WANG Yiran,JIAO Licheng,Shilian ZHENG,SHEN Weiguo,XU Jie,YANG Xiaoniu,Huaji ZHOU,Jing BAI,Yiran WANG,Licheng JIAO,Weiguo SHEN,Jie XU,Xiaoniu YANG
DOI: https://doi.org/10.1016/j.cja.2021.07.014
IF: 5.7
2021-09-01
Chinese Journal of Aeronautics
Abstract:Deep learning has been fully verified and accepted in the field of electromagnetic signal classification. However, in many specific scenarios, such as radio resource management for aircraft communications, labeled data are difficult to obtain, which makes the best deep learning methods at present seem almost powerless, because these methods need a large amount of labeled data for training. When the training dataset is small, it is highly possible to fall into overfitting, which causes performance degradation of the deep neural network. For few-shot electromagnetic signal classification, data augmentation is one of the most intuitive countermeasures. In this work, a generative adversarial network based on the data augmentation method is proposed to achieve better classification performance for electromagnetic signals. Based on the similarity principle, a screening mechanism is established to obtain high-quality generated signals. Then, a data union augmentation algorithm is designed by introducing spatiotemporally flipped shapes of the signal. To verify the effectiveness of the proposed data augmentation algorithm, experiments are conducted on the RADIOML 2016.04C dataset and real-world ACARS dataset. The experimental results show that the proposed method significantly improves the performance of few-shot electromagnetic signal classification.
engineering, aerospace