Intelligent radar HRRP target recognition based on CNN-BERT model
Abstract Stable and reliable feature extraction is crucial for radar high-resolution range profile (HRRP) target recognition. Owing to the complex structure of HRRP data, existing feature extraction methods fail to achieve satisfactory performance. This study proposes a new deep learning model named...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SpringerOpen
2022-09-01
|
| Series: | EURASIP Journal on Advances in Signal Processing |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s13634-022-00909-9 |
| _version_ | 1828175333264719872 |
|---|---|
| author | Penghui Wang Ting Chen Jun Ding Mian Pan Sanding Tang |
| author_facet | Penghui Wang Ting Chen Jun Ding Mian Pan Sanding Tang |
| author_sort | Penghui Wang |
| collection | DOAJ |
| description | Abstract Stable and reliable feature extraction is crucial for radar high-resolution range profile (HRRP) target recognition. Owing to the complex structure of HRRP data, existing feature extraction methods fail to achieve satisfactory performance. This study proposes a new deep learning model named convolutional neural network–bidirectional encoder representations from transformers (CNN-BERT), using the spatio–temporal structure embedded in HRRP for target recognition. The convolutional token embedding module characterizes the local spatial structure of the target and generates the sequence features by token embedding. The BERT module captures the long-term temporal dependence among range cells within HRRP through the multi-head self-attention mechanism. Furthermore, a novel cost function that simultaneously considers the recognition and rejection ability is designed. Extensive experiments on measured HRRP data reveal the superior performance of the proposed model. |
| first_indexed | 2024-04-12T04:26:55Z |
| format | Article |
| id | doaj.art-9ad5371f82f84778a164dfdce0c2eac7 |
| institution | Directory Open Access Journal |
| issn | 1687-6180 |
| language | English |
| last_indexed | 2024-04-12T04:26:55Z |
| publishDate | 2022-09-01 |
| publisher | SpringerOpen |
| record_format | Article |
| series | EURASIP Journal on Advances in Signal Processing |
| spelling | doaj.art-9ad5371f82f84778a164dfdce0c2eac72022-12-22T03:48:03ZengSpringerOpenEURASIP Journal on Advances in Signal Processing1687-61802022-09-012022112610.1186/s13634-022-00909-9Intelligent radar HRRP target recognition based on CNN-BERT modelPenghui Wang0Ting Chen1Jun Ding2Mian Pan3Sanding Tang4National Laboratory of Radar Signal Processing, Xidian UniversityNational Laboratory of Radar Signal Processing, Xidian UniversityNational Laboratory of Radar Signal Processing, Xidian UniversitySchool of Electronics and Information, Hangzhou Dianzi UniversitySchool of Electronics and Information, Hangzhou Dianzi UniversityAbstract Stable and reliable feature extraction is crucial for radar high-resolution range profile (HRRP) target recognition. Owing to the complex structure of HRRP data, existing feature extraction methods fail to achieve satisfactory performance. This study proposes a new deep learning model named convolutional neural network–bidirectional encoder representations from transformers (CNN-BERT), using the spatio–temporal structure embedded in HRRP for target recognition. The convolutional token embedding module characterizes the local spatial structure of the target and generates the sequence features by token embedding. The BERT module captures the long-term temporal dependence among range cells within HRRP through the multi-head self-attention mechanism. Furthermore, a novel cost function that simultaneously considers the recognition and rejection ability is designed. Extensive experiments on measured HRRP data reveal the superior performance of the proposed model.https://doi.org/10.1186/s13634-022-00909-9High-resolution range profile (HRRP)Convolutional neural network (CNN)Bidirectional encoder representations from transformers (BERT)Attention mechanismIntelligent target recognition |
| spellingShingle | Penghui Wang Ting Chen Jun Ding Mian Pan Sanding Tang Intelligent radar HRRP target recognition based on CNN-BERT model EURASIP Journal on Advances in Signal Processing High-resolution range profile (HRRP) Convolutional neural network (CNN) Bidirectional encoder representations from transformers (BERT) Attention mechanism Intelligent target recognition |
| title | Intelligent radar HRRP target recognition based on CNN-BERT model |
| title_full | Intelligent radar HRRP target recognition based on CNN-BERT model |
| title_fullStr | Intelligent radar HRRP target recognition based on CNN-BERT model |
| title_full_unstemmed | Intelligent radar HRRP target recognition based on CNN-BERT model |
| title_short | Intelligent radar HRRP target recognition based on CNN-BERT model |
| title_sort | intelligent radar hrrp target recognition based on cnn bert model |
| topic | High-resolution range profile (HRRP) Convolutional neural network (CNN) Bidirectional encoder representations from transformers (BERT) Attention mechanism Intelligent target recognition |
| url | https://doi.org/10.1186/s13634-022-00909-9 |
| work_keys_str_mv | AT penghuiwang intelligentradarhrrptargetrecognitionbasedoncnnbertmodel AT tingchen intelligentradarhrrptargetrecognitionbasedoncnnbertmodel AT junding intelligentradarhrrptargetrecognitionbasedoncnnbertmodel AT mianpan intelligentradarhrrptargetrecognitionbasedoncnnbertmodel AT sandingtang intelligentradarhrrptargetrecognitionbasedoncnnbertmodel |