An operating modes identification of spaceborne synthetic aperture radar (SAR) model based on one-dimensional convolutional neural network was proposed according to the timing characteristics of SAR signals, in order to solve the limitation of the recognition accuracy and the timeliness of traditional spaceborne SAR operating modes inversion methods. The impulse peak amplitude of the SAR signal was taken as input, more subtle and representative features of the original signal were learned by using adaptive learning and pattern recognition ability of the convolutional neural network, the human interference factors of traditional methods were avoided, and the effective identification of the operating modes of spaceborne SAR was finally realized. The one-dimensional convolutional neural network structure was designed referring to the existing convolutional neural network with good performance, and the better parameters were adjusted and set to train a model with good recognition performance according to the feedback of the accuracy and the loss value in the training process of the network. Contrast experiments based on simulation data demonstrate that the model has higher recognition accuracy than traditional spaceborne SAR operating modes inversion methods and has excellent robustness and noise anti-noise ability under different types of signals and different detection conditions.
Jun HE,Ya-sheng ZHANG,Can-bin YIN. Operating modes identification of spaceborne SAR based on deep learning. Journal of ZheJiang University (Engineering Science), 2022, 56(8): 1676-1684.
Fig.1Schematic diagram of operating modes of spaceborne SAR
Fig.2Schematic diagram of one-dimensional convolutional neural network structure
Fig.3Peak amplitude of pulse signal in different operating modes
仿真参数
数值
仿真参数
数值
发射功率
5 000 W
合成孔径时间
20 s
天线增益
45 dB
最小侧视角
20°
波束方位宽度
0.2°
最大侧视角
50°
波束距离宽度
1°
最大斜视角
30°
信号带宽
30 MHz
发射脉冲宽度
5 us
载波频率
10 GHz
侦收采样率
60 MHz
Tab.1Parameters of spaceborne SAR
Fig.4Curve diagram of training accuracy
Fig.5Curve diagram of training loss
Fig.6Confusion matrixes of different recognition methods
方法
Re/%
Acc/%
条带 模式
聚束 模式
滑动 聚束
斜视 扫描
侧式 扫描
文献[5]
79.41
84.24
75.41
56.54
55.03
70.46
LSTM
89.22
92.39
95.63
89.53
96.04
92.59
GRU
90.20
95.11
92.35
91.62
97.63
93.23
1D-CNN
91.18
94.02
92.90
93.19
100.00
94.09
Tab.2Comparison of recognition accuracy by different classification methods
方法
Nt
tt
LSTM
80 755 205
7 h 16 min 22 s
GRU
80 229 381
5 h 53 min 56 s
1D-CNN
4 262 597
1 h 10 min 46 s
Tab.3Comparison of training by different deep learning models
Fig.7Confusion matrix of different types of signals
信号类型
Re/%
Acc/%
条带 模式
聚束 模式
滑动 聚束
斜视 扫描
侧式 扫描
主瓣信号
90.60
93.60
90.24
90.15
100
92.82
旁瓣信号
90.24
90.58
91.30
90.90
94.53
91.50
Tab.4Classification accuracy of different types of signals
Fig.8Division of deployment areas of ground receiving station
Fig.9Confusion matrix of different regions of receiving
接收区域
Re/%
Acc/%
条带 模式
聚束 模式
滑动 聚束
斜视 扫描
侧式 扫描
近区
90.00
94.44
94.51
91.21
94.62
92.97
中区
90.22
94.57
90.53
89.13
93.98
91.63
远区
88.18
90.32
89.25
86.02
96.74
90.09
Tab.5Classification accuracy of different regions of receiving
[1]
王振力, 钟海 国外先进星载SAR卫星的发展现状及应用[J]. 国防科技, 2016, 37 (1): 19- 24 WANG Zhen-li, ZHONG Hai The nowaday development and application of oversea advanced spaceborne SAR[J]. National Defense Science and Technology, 2016, 37 (1): 19- 24
[2]
陈颖颖, 吴彦鸿, 贾鑫 对不同工作模式星载合成孔径雷达的侦察研究[J]. 计算机工程与应用, 2013, 49 (12): 223- 227,241 CHEN Ying-ying, WU Yan-hong, JIA Xin Surveillance study of spaceborne synthetic aperture radar in different working modes[J]. Computer Engineering and Applications, 2013, 49 (12): 223- 227,241
doi: 10.3778/j.issn.1002-8331.1110-0569
[3]
唐小明, 立春升, 孙兵 基于遗传算法的星载SAR工作模式反演方法[J]. 空间电子技术, 2013, 10 (2): 90- 94 TANG Xiao-ming, LI Chun-sheng, SUN Bing An operation inversion method of spaceborne SAR based on genetic algorithm[J]. Space Electronic Technology, 2013, 10 (2): 90- 94
doi: 10.3969/j.issn.1674-7135.2013.02.017
[4]
JORDI I, ARTHUR V, MARCELA A, et al Improved early crop type identification by joint use of high temporal resolution SAR and optical image time series[J]. Remote Sensing, 2016, 8 (5): 362- 382
doi: 10.3390/rs8050362
[5]
夏周越, 钟华, 陈维 侦察模式下星载合成孔径雷达工作模式鉴别[J]. 杭州电子科技大学学报, 2017, 37 (4): 36- 40 XIA Zhou-yue, ZHONG Hua, CHEN Wei Identification of the spaceborne SAR operating modes under reconnaissance mode[J]. Journal of Hangzhou Dianzi University, 2017, 37 (4): 36- 40
[6]
刘寒艳, 宋红军, 程增菊 条带模式、聚束模式和滑动聚束模式的比较[J]. 中国科学院研究生院学报, 2011, 28 (3): 410- 417 LIU Han-yan, SONG Hong-jun, CHENG Zeng-ju Comparative study on stripmap mode, spotlight mode, and sliding spotlight mode[J]. Journal of the Graduate School of the Chinese Academy of Sciences, 2011, 28 (3): 410- 417
[7]
胡新宇, 张铁军, 王昀 低截获概率雷达信号侦察技术[J]. 航天电子对抗, 2020, 36 (5): 40- 43 HU Xin-yu, ZHANG Tie-jun, WANG Jun Low probability of intercept radar signal reconnaissance technology[J]. Aerospace Electronic Warfare, 2020, 36 (5): 40- 43
doi: 10.3969/j.issn.1673-2421.2020.05.010
[8]
陈颖颖, 贾鑫, 吴彦鸿 对聚束和滑动聚束模式下星载合成孔径雷达的旁瓣侦察比较研究[J]. 科学技术与工程, 2012, 12 (8): 1785- 1789 CHEN Ying-ying, JIA Xin, WU Yan-hong The sidelobe surveillance study of spaceborne synthetic aperture radar in spotlight and sliding spotlight mode[J]. Science Technology and Engineering, 2012, 12 (8): 1785- 1789
doi: 10.3969/j.issn.1671-1815.2012.08.014
[9]
周飞燕, 金林鹏, 董军 卷积神经网络研究综述[J]. 计算机学报, 2017, 40 (6): 1229- 1251 ZHOU Fei-yan, JIN Lin-peng, DONG Jun Review of convolutional neural network[J]. Chinese Journal of Computers, 2017, 40 (6): 1229- 1251
doi: 10.11897/SP.J.1016.2017.01229
[10]
李红光, 郭英, 眭萍, 等 基于时频特征的卷积神经网络跳频调制识别[J]. 浙江大学学报: 工学版, 2020, 54 (10): 1945- 1954 LI Hong-guang, GUO Ying, SUI Ping, et al Frequency hopping modulation recognition of convolutional neural network based on time-frequency characteristics[J]. Journal of Zhejiang University: Engineering Science, 2020, 54 (10): 1945- 1954
[11]
IOFFE S, SZEGEDY C. Batch normalization: accelerating deep network training by reducing internal covariate shift [C] // Proceedings of the 32nd International Conference on Machine Learning. Lille: ICML, 2015: 448-456.
[12]
HAO X, ZHANG G Deep learning[J]. Encyclopedia with Semantic Computing and Robotic Intelligence, 2017, 1 (1): 1630018
[13]
LI Z W, LIU F, YANG W J, et al. A survey of convolutional neural networks: analysis, applications, and prospects [EB/OL]. [2021-06-10]. https://ieeexplore.ieee.org/document/9451544.
[14]
金列俊, 詹建明, 陈俊华, 等 基于一维卷积神经网络的钻杆故障诊断[J]. 浙江大学学报: 工学版, 2020, 54 (3): 467- 474 JIN Lie-jun, ZHAN Jian-ming, CHEN Jun-hua, et al Drill pipe fault diagnosis method based on one-dimensional convolutional neural network[J]. Journal of Zhejiang University: Engineering Science, 2020, 54 (3): 467- 474
[15]
LJABER O, AVCI O, KIRANYAZ S, et al Real-time vibration-based structural damage detection using one-dimensional convolutional neural networks[J]. Journal of Sound and Vibration, 2017, 388: 154- 170
doi: 10.1016/j.jsv.2016.10.043
[16]
汪海晋, 尹宗宇, 柯臻铮, 等 基于一维卷积神经网络的螺旋铣刀具磨损监测[J]. 浙江大学学报: 工学版, 2020, 54 (5): 931- 939 WANG Hai-jin, YIN Zong-yu, KE Zhen-zheng, et al Wear monitoring of helical milling tool based on one-dimensional convolutional neural network[J]. Journal of Zhejiang University: Engineering Science, 2020, 54 (5): 931- 939
[17]
OLKHOVSKIY M, MÜLLEROVÁ E, MARTÍNEK P Impulse signals classification using one dimensional convolutional neural network[J]. Journal of Electrical Engineering, 2020, 71 (6): 397- 405
doi: 10.2478/jee-2020-0054
[18]
ZHANG M, LIU Z, LI L, et al Enhanced efficiency BPSK demodulator based on one-dimensional convolutional neural network[J]. IEEE Access, 2018, 6: 26939- 26948
doi: 10.1109/ACCESS.2018.2834144
[19]
LIN M, CHEN Q, YAN S C. Network in network [C] // Proceedings of the International Conference on Learning Representations. Banff: ICLR, 2014: 1-10.
[20]
陈杰, 杨威, 王鹏波, 等 多方位角观测星载SAR技术研究[J]. 雷达学报, 2020, 9 (2): 205- 220 CHEN Jie, YANG Wei, WANG Peng-bo, et al Review of novel azimuthal multi-angle observation spaceborne SAR technique[J]. Journal of Radars, 2020, 9 (2): 205- 220
doi: 10.12000/JR20015
[21]
王岩. 深度神经网络的归一化技术研究[D]. 南京: 南京邮电大学, 2019. WANG Yan. Analysis of normalization for deep neural network [D]. Nanjing: Nanjing University of Posts and Telecommunications, 2019.
