A Novel Aggregated Multipath Extreme Gradient Boosting Approach for Radar Emitter Classification
Shiqiang Zhao,Wenhai Wang,Deguo Zeng,Xinwei Chen,Zeyin Zhang,Fuyuan Xu,Xuanyu Mao,Xinggao Liu
DOI: https://doi.org/10.1109/TIE.2021.3055155
IF: 7.7
2022-01-01
IEEE Transactions on Industrial Electronics
Abstract:Radar emitter classification (REC) plays an increasingly significant role in the electronic reconnaissance system. Due to there are many unreliable factors in traditional feature-based methods, this article focuses on improving the accuracy and stability of REC based on multipaths of original radar signals. A novel aggregated multipath extreme gradient boosting (AMP-XGBoost) is therefore put forward. Multi-path, including multi-scale, multi-domain transformations and their concatenations, is developed to exploit more information of the original signal, which contributes to providing more distinguishable features. XGBoost is used to automatically extract the features contained in these paths and complete recognition. Finally, multiple paths are sifted and aggregated according to certain weights to obtain a higher accuracy. Experiments are carried out based on signals from 6 different intra-pulse modulation mode radars with the same radio frequency (RF) range. It is demonstrated that the proposed method has higher REC accuracy than the methods only based on a single scale or domain. In the meantime, experimental results under different low SNRs show that the proposed method has higher stability. Finally, experiments based on the measured aviation radar data proves the superiority of the proposed method.