LPI Radar Waveform Recognition Based on Multi-Resolution Deep Feature Fusion

被引:22
作者
Ni, Xue [1 ]
Wang, Huali [1 ]
Meng, Fan [2 ]
Hu, Jing [1 ]
Tong, Changkai [1 ]
机构
[1] Army Engn Univ PLA, Coll Commun Engn, Nanjing 210007, Peoples R China
[2] Nanjing Marine Radar Inst, Nanjing 210007, Peoples R China
关键词
Feature extraction; Radar; Radar imaging; Image resolution; Signal resolution; Signal to noise ratio; Time-frequency analysis; Radar waveform recognition; multi-resolution; feature fusion; FSST; convolutional neural network; NETWORKS;
D O I
10.1109/ACCESS.2021.3058305
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
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.
引用
收藏
页码:26138 / 26146
页数:9
相关论文
共 24 条
[1]  
Anjaneyulu L., 2009, International Journal of Recent Trends in Engineering, V1, P176
[2]  
Cao P., 2018, PROC 10 INT C WIRELE, P1
[3]   Separability of ternary codes for sparse designs of error-correcting output codes [J].
Escalera, Sergio ;
Pujol, Oriol ;
Radeva, Petia .
PATTERN RECOGNITION LETTERS, 2009, 30 (03) :285-297
[4]   Fusion Image Based Radar Signal Feature Extraction and Modulation Recognition [J].
Gao, Lipeng ;
Zhang, Xiaoli ;
Gao, Jingpeng ;
You, Shixun .
IEEE ACCESS, 2019, 7 :13135-13148
[5]   An Improved Sensor Fault Diagnosis Scheme Based on TA-LSSVM and ECOC-SVM [J].
Gu, Xiaodan ;
Deng, Fang ;
Gao, Xin ;
Zhou, Rui .
JOURNAL OF SYSTEMS SCIENCE & COMPLEXITY, 2018, 31 (02) :372-384
[6]   LPI Radar Waveform Recognition Based on Deep Convolutional Neural Network Transfer Learning [J].
Guo, Qiang ;
Yu, Xin ;
Ruan, Guoqing .
SYMMETRY-BASEL, 2019, 11 (04)
[7]  
Kong G., 2019, CAMSAP, P664
[8]   Automatic LPI Radar Wave form Recognition Using CNN [J].
Kong, Seung-Hyun ;
Kim, Minjun ;
Linh Manh Hoang ;
Kim, Eunhui .
IEEE ACCESS, 2018, 6 :4207-4219
[9]  
LeCun Y, 2010, IEEE INT SYMP CIRC S, P253, DOI 10.1109/ISCAS.2010.5537907
[10]   Feature Pyramid Networks for Object Detection [J].
Lin, Tsung-Yi ;
Dollar, Piotr ;
Girshick, Ross ;
He, Kaiming ;
Hariharan, Bharath ;
Belongie, Serge .
30TH IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2017), 2017, :936-944