| 摘要: |
| 针对目前敌我识别辐射源个体识别（Specific Emitter Identification of Identification Friend or Foe,SEI-IFF）研究不足的问题，提出了一种基于多维特征与Transformer网络的SEI-IFF方法。该方法首先从单个脉冲及信号全局等多维度获取如相位、幅度、时间、功率谱密度等信号特征，结合Transformer网络进一步提取不同IFF辐射源个体特征中如前后关联特性的细微特征并最终实现SEI-IFF。试验结果表明，所提方法针对20个目标搭载的IFF辐射源个体的平均识别正确率达95.3%，可较准确地完成SEI-IFF，有助于提升战场SEI-IFF效率。 |
| 关键词: 敌我识别(IFF) 辐射源个体识别(SEI) 多维特征 Transformer |
| DOI：10.20079/j.issn.1001-893x.240601001 |
|
| 基金项目: |
|
| Specific Emitter Identification of Identification Friend or Foe by Combining Multidimensional Features and Transformer |
| CHEN Jiemei |
| (Southwest China Institute of Electronic Technology,Chengdu 610036,China) |
| Abstract: |
| In order to solve the problem of insufficient research on specific emitter identification of identification friend or foe(SEI-IFF),a method based on multidimensional features and Transformer is proposed.Firstly,multidimensional features from individual pulse and global signal,such as phase,amplitude,time,power spectral density are extracted.Combining Transformer,fine features,such as the correlated characteristics of features in different IFF transmitters are extracted and finally SEI-IFF is achieved.Results show that the proposed method achieves an average recognition accuracy of 95.3% for the individual sources carried by 20 targets.It can effectively achieve SEI-IFF and is helpful for improving the efficiency of SEI-IFF in battlefield. |
| Key words: identification of friend or foe(IFF) specific emitter identification(SEI) multidimensional features Transformer |
