A Few-Shot Learning for Predicting Radar Receiver Interference Response Based on Distillation Meta-Learning
Establishing a behavioral model for radar equipment and ensuring the electromagnetic compatibility of radar systems in unmanned ship formations are essential for coordinated operations. Traditional radar receiver modeling methods based on supervised learning require complex models and large datasets...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10806651/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841536178179801088 |
|---|---|
| author | Lingyun Zhang Hui Tan Mingliang Huang |
| author_facet | Lingyun Zhang Hui Tan Mingliang Huang |
| author_sort | Lingyun Zhang |
| collection | DOAJ |
| description | Establishing a behavioral model for radar equipment and ensuring the electromagnetic compatibility of radar systems in unmanned ship formations are essential for coordinated operations. Traditional radar receiver modeling methods based on supervised learning require complex models and large datasets, which are costly and time-consuming. In the complex electromagnetic environment of unmanned ship formations, these methods often prove ineffective quickly. Therefore, this paper proposes a few-shot modeling method based on distillation meta-learning, integrating meta-learning and distillation learning. Firstly, model-agnostic meta-learning (MAML) divides the pre-training process into two stages, using interference data of various modulation types to train a general model. Then, distillation learning transfers knowledge from the complex pre-trained model to a simplified student model. This approach compresses model parameters while maintaining prediction accuracy, facilitating easier deployment in practical equipment. Results based on simulation data demonstrate that our proposed method can effectively predict the receiver’s interference response with minimal gradient steps and a small amount of training data. |
| format | Article |
| id | doaj-art-654f14dd37bc467c916cf93423818792 |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-654f14dd37bc467c916cf934238187922025-01-15T00:01:39ZengIEEEIEEE Access2169-35362024-01-011219507619508410.1109/ACCESS.2024.351972710806651A Few-Shot Learning for Predicting Radar Receiver Interference Response Based on Distillation Meta-LearningLingyun Zhang0https://orcid.org/0009-0001-1829-0806Hui Tan1Mingliang Huang2National Key Laboratory of Electromagnetic Effect and Security on Marine Equipment, China Ship Development and Design Center, Wuhan, ChinaNational Key Laboratory of Electromagnetic Effect and Security on Marine Equipment, China Ship Development and Design Center, Wuhan, ChinaNational Key Laboratory of Electromagnetic Effect and Security on Marine Equipment, China Ship Development and Design Center, Wuhan, ChinaEstablishing a behavioral model for radar equipment and ensuring the electromagnetic compatibility of radar systems in unmanned ship formations are essential for coordinated operations. Traditional radar receiver modeling methods based on supervised learning require complex models and large datasets, which are costly and time-consuming. In the complex electromagnetic environment of unmanned ship formations, these methods often prove ineffective quickly. Therefore, this paper proposes a few-shot modeling method based on distillation meta-learning, integrating meta-learning and distillation learning. Firstly, model-agnostic meta-learning (MAML) divides the pre-training process into two stages, using interference data of various modulation types to train a general model. Then, distillation learning transfers knowledge from the complex pre-trained model to a simplified student model. This approach compresses model parameters while maintaining prediction accuracy, facilitating easier deployment in practical equipment. Results based on simulation data demonstrate that our proposed method can effectively predict the receiver’s interference response with minimal gradient steps and a small amount of training data.https://ieeexplore.ieee.org/document/10806651/Radar receiverelectromagnetic compatibilityinterference responsenonlinear circuit modelingfew-shot learningknowledge distillation |
| spellingShingle | Lingyun Zhang Hui Tan Mingliang Huang A Few-Shot Learning for Predicting Radar Receiver Interference Response Based on Distillation Meta-Learning IEEE Access Radar receiver electromagnetic compatibility interference response nonlinear circuit modeling few-shot learning knowledge distillation |
| title | A Few-Shot Learning for Predicting Radar Receiver Interference Response Based on Distillation Meta-Learning |
| title_full | A Few-Shot Learning for Predicting Radar Receiver Interference Response Based on Distillation Meta-Learning |
| title_fullStr | A Few-Shot Learning for Predicting Radar Receiver Interference Response Based on Distillation Meta-Learning |
| title_full_unstemmed | A Few-Shot Learning for Predicting Radar Receiver Interference Response Based on Distillation Meta-Learning |
| title_short | A Few-Shot Learning for Predicting Radar Receiver Interference Response Based on Distillation Meta-Learning |
| title_sort | few shot learning for predicting radar receiver interference response based on distillation meta learning |
| topic | Radar receiver electromagnetic compatibility interference response nonlinear circuit modeling few-shot learning knowledge distillation |
| url | https://ieeexplore.ieee.org/document/10806651/ |
| work_keys_str_mv | AT lingyunzhang afewshotlearningforpredictingradarreceiverinterferenceresponsebasedondistillationmetalearning AT huitan afewshotlearningforpredictingradarreceiverinterferenceresponsebasedondistillationmetalearning AT minglianghuang afewshotlearningforpredictingradarreceiverinterferenceresponsebasedondistillationmetalearning AT lingyunzhang fewshotlearningforpredictingradarreceiverinterferenceresponsebasedondistillationmetalearning AT huitan fewshotlearningforpredictingradarreceiverinterferenceresponsebasedondistillationmetalearning AT minglianghuang fewshotlearningforpredictingradarreceiverinterferenceresponsebasedondistillationmetalearning |