LPI Radar Waveform Recognition Based on Multi-Resolution Deep Feature Fusion
Deep neural networks are used as effective methods for the Low Probability of Intercept (LPI) radar waveform recognition. However, existing models' performance degrades seriously at low Signal-to-Noise Ratios (SNRs) because the effective features extracted by the networks are insufficient under...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2021-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/9350641/ |
id |
doaj-0acff32a6b1e49c29435afaac4e6f4ef |
---|---|
record_format |
Article |
spelling |
doaj-0acff32a6b1e49c29435afaac4e6f4ef2021-03-30T15:25:04ZengIEEEIEEE Access2169-35362021-01-019261382614610.1109/ACCESS.2021.30583059350641LPI Radar Waveform Recognition Based on Multi-Resolution Deep Feature FusionXue Ni0https://orcid.org/0000-0002-7524-3360Huali Wang1https://orcid.org/0000-0003-2084-6110Fan Meng2https://orcid.org/0000-0002-4818-7009Jing Hu3https://orcid.org/0000-0001-5567-6911Changkai Tong4https://orcid.org/0000-0001-8705-093XCollege of Communications Engineering, Army Engineering University of PLA, Nanjing, ChinaCollege of Communications Engineering, Army Engineering University of PLA, Nanjing, ChinaNanjing Marine Radar Institute, Nanjing, ChinaCollege of Communications Engineering, Army Engineering University of PLA, Nanjing, ChinaCollege of Communications Engineering, Army Engineering University of PLA, Nanjing, ChinaDeep neural networks are used as effective methods for the Low Probability of Intercept (LPI) radar waveform recognition. However, existing models' performance degrades seriously at low Signal-to-Noise Ratios (SNRs) because the effective features extracted by the networks are insufficient under noise jamming. In this paper, we propose a multi-resolution deep feature fusion method for LPI radar waveform recognition. First, we apply the enhanced Fourier-based Synchrosqueezing Transform (FSST), which shows good performance at low SNRs, to convert radar signals into time-frequency images. Then, we construct a multi-resolution deep convolutional network to extract more deep features from each resolution channel. Next, we explore an interactive feature fusion strategy for deep feature fusion. By some down-sampling or up-sampling blocks, different resolution features are fused to generate new features. Finally, we apply a fusion algorithm to the fully connected layer to achieve classification fusion for better performance. Simulation experiments on twelve kinds of LPI radar waveforms show that the overall recognition accuracy of our method can reach 95.2% at the SNR of -8 dB. It is proved that our approach does indeed improve the recognition accuracy effectively at low SNRs.https://ieeexplore.ieee.org/document/9350641/Radar waveform recognitionmulti-resolutionfeature fusionFSSTconvolutional neural network |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Xue Ni Huali Wang Fan Meng Jing Hu Changkai Tong |
spellingShingle |
Xue Ni Huali Wang Fan Meng Jing Hu Changkai Tong LPI Radar Waveform Recognition Based on Multi-Resolution Deep Feature Fusion IEEE Access Radar waveform recognition multi-resolution feature fusion FSST convolutional neural network |
author_facet |
Xue Ni Huali Wang Fan Meng Jing Hu Changkai Tong |
author_sort |
Xue Ni |
title |
LPI Radar Waveform Recognition Based on Multi-Resolution Deep Feature Fusion |
title_short |
LPI Radar Waveform Recognition Based on Multi-Resolution Deep Feature Fusion |
title_full |
LPI Radar Waveform Recognition Based on Multi-Resolution Deep Feature Fusion |
title_fullStr |
LPI Radar Waveform Recognition Based on Multi-Resolution Deep Feature Fusion |
title_full_unstemmed |
LPI Radar Waveform Recognition Based on Multi-Resolution Deep Feature Fusion |
title_sort |
lpi radar waveform recognition based on multi-resolution deep feature fusion |
publisher |
IEEE |
series |
IEEE Access |
issn |
2169-3536 |
publishDate |
2021-01-01 |
description |
Deep neural networks are used as effective methods for the Low Probability of Intercept (LPI) radar waveform recognition. However, existing models' performance degrades seriously at low Signal-to-Noise Ratios (SNRs) because the effective features extracted by the networks are insufficient under noise jamming. In this paper, we propose a multi-resolution deep feature fusion method for LPI radar waveform recognition. First, we apply the enhanced Fourier-based Synchrosqueezing Transform (FSST), which shows good performance at low SNRs, to convert radar signals into time-frequency images. Then, we construct a multi-resolution deep convolutional network to extract more deep features from each resolution channel. Next, we explore an interactive feature fusion strategy for deep feature fusion. By some down-sampling or up-sampling blocks, different resolution features are fused to generate new features. Finally, we apply a fusion algorithm to the fully connected layer to achieve classification fusion for better performance. Simulation experiments on twelve kinds of LPI radar waveforms show that the overall recognition accuracy of our method can reach 95.2% at the SNR of -8 dB. It is proved that our approach does indeed improve the recognition accuracy effectively at low SNRs. |
topic |
Radar waveform recognition multi-resolution feature fusion FSST convolutional neural network |
url |
https://ieeexplore.ieee.org/document/9350641/ |
work_keys_str_mv |
AT xueni lpiradarwaveformrecognitionbasedonmultiresolutiondeepfeaturefusion AT hualiwang lpiradarwaveformrecognitionbasedonmultiresolutiondeepfeaturefusion AT fanmeng lpiradarwaveformrecognitionbasedonmultiresolutiondeepfeaturefusion AT jinghu lpiradarwaveformrecognitionbasedonmultiresolutiondeepfeaturefusion AT changkaitong lpiradarwaveformrecognitionbasedonmultiresolutiondeepfeaturefusion |
_version_ |
1724179486656692224 |