基于深度分割的端到端雷达信号分选

doi:10.12305/j.issn.1001-506X.2023.05.11

Abstract

Abstract:

Aiming at the problem that the traditional radar signal sorting method relies heavily on preset parameters and prior information and the problem of inflexible structure, an end-to-end radar signal sorting method based on deep segmentation is proposed. Firstly, collected pulse description words (PDWs) are mapped into pulse sequence images, while retaining the index of the pixel to the pulse. Secondly, the trained depth segmentation U-Net model segments the pulse sequence image to obtain the pixel classification results. Finally, all pulses are searched and sorted according to the pixel index and the pixel classification results. The whole process is carried out in an end-to-end manner. Experiments show that this method can strengthen the classification ability to sort parameters of radars such as frequency agile radar, pulse group agile radar, and other unknown electromagnetic environments with severe time-frequency domain aliasing and severe pulse loss.

Key words: radar signal sorting, end-to-end, pulse description word (PDW), U-Net model

CLC Number:

Tao CHEN, Fuyue LIU, Jinxin LI, Yu LEI. End-to-end radar signal sorting based on deep segmentation[J]. Systems Engineering and Electronics, 2023, 45(5): 1351-1358.

Figures/Tables 13

Fig.1

Table 1

Table 2

Fig.2

Table 3

Fig.3

Table 4

Fig.4

Fig.5

Fig.6

Fig.7

Fig.8

Fig.9

References 23

1	张怡霄, 郭文普, 康凯, 等. 基于数据场联合PRI变换与聚类的雷达信号分选[J]. 系统工程与电子技术, 2019, 41 (7): 1509- 1515.
	ZHANG Y X , GUO W P , KANG K , et al. Radar signal sorting based on data field combined PRI transform and clustering[J]. Systems Engineering and Electronics, 2019, 41 (7): 1509- 1515.
2	KANUNGO T , MOUNT D M , NETANYAHU N S , et al. An efficient k-means clustering algorithm: analysis and implementation[J]. IEEE Trans. on Pattern Analysis and Machine Intelligence, 2002, 24 (7): 881- 892. doi: 10.1109/TPAMI.2002.1017616
3	HUANG J Z , NG M K , RONG H Q , et al. Automated variable weighting in k-means type clustering[J]. IEEE Trans. on Pattern Analysis and Machine Intelligence, 2005, 27 (5): 657- 668. doi: 10.1109/TPAMI.2005.95
4	GADDAM S R , PHOHA V V , BALAGANI K S . K-Means+ID3: A novel method for supervi-sed anomaly detection by cascading K-Means clustering and ID3 decision tree learning methods[J]. IEEE Trans. on Knowledge and Data Engineering, 2007, 19 (3): 345- 354. doi: 10.1109/TKDE.2007.44
5	SIMITI A, ELOUEDI Z. DBSCAN-GM: an improved clustering method based on gaussian means and DBSCAN techniques[C]//Proc. of the IEEE 16th International Conference on Intelligent Engineering Systems, 2012: 573-578.
6	HOU J , GAO H J , LI X L . DSets-DBSCAN: A parameter-free clustering algorithm[J]. IEEE Trans. on Image Processing, 2016, 25 (7): 3182- 3193. doi: 10.1109/TIP.2016.2559803
7	GIESS M , AHMED S , TANG B . Sorting radar signal from symmetry clustering perspective[J]. Journal of Systems Engineering and Electronics, 2017, 28 (4): 690- 696. doi: 10.21629/JSEE.2017.04.08
8	MARDIA H K . New techniques for the deinter-leaving of repetitive sequences[J]. IEE Proceedings F (Radar and Signal Processing), 1989, 136 (4): 149- 154. doi: 10.1049/ip-f-2.1989.0025
9	MILOJEVIC D J , POPOVIC B M . Improved algorithm for the deinterleaving of radar pulses[J]. IEE Proceedings F (Radar and Signal Processing), 1992, 139 (1): 98- 104. doi: 10.1049/ip-f-2.1992.0012
10	LIU Y C , ZHANG Q Y . Improved method for deinterleaving radar signals and estimating PRI values[J]. IET Radar, Sonar & Navigation, 2018, 12 (5): 506- 514.
11	NELSON D. Special purpose correlation functions for improved signal detection and parameter estimation[C]//Proc. of the IEEE International Conference on Acoustics, Speech, and Signal Processing, 1993, 4: 73-76.
12	NISHIGUCHI K , KOBAYASHI M . Improved algorithm for estimating pulse repetition intervals[J]. IEEE Trans. on Aerospace and Electronic Systems, 2000, 36 (2): 407- 421. doi: 10.1109/7.845217
13	MAHADAVI A, PEZESHK A M. A fast enhanced algorithm of PRI transform[C]//Proc. of the 6th International Symposium on Parallel Computing in Electrical Engineering, 2011: 179-184.
14	王俊岭, 黄琰璟. 基于序列关联的参差信号分选算法[J]. 电子与信息学报, 2021, 43 (4): 1145- 1153.
	WANG J L , HUANG Y J . Stagger pulse repetition interval pulse train deinterleaving algorithm based on sequence association[J]. Journal of Electronics & Information Technology, 2021, 43 (4): 1145- 1153.
15	李雅朦, 武艳伟, 裴晓帅, 等. 基于深度自编码器的PRI调制方式识别的一种新方法[J]. 航天电子对抗, 2021, 37 (2): 52- 57. doi: 10.3969/j.issn.1673-2421.2021.02.013
	LI Y M , WU Y W , PEI X S , et al. A new method of PRI modulation recognition based on data reduction and deep learning[J]. Aerospace Electronic Warfare, 2021, 37 (2): 52- 57. doi: 10.3969/j.issn.1673-2421.2021.02.013
16	孟磊, 曲卫, 马爽, 等. 基于LSTM的雷达脉冲重复间隔调制模式识别[J]. 现代雷达, 2021, 43 (1): 50- 57.
	MENG L , QU W , MA S , et al. Radar PRI modulation pattern recognition method based on LSTM[J]. Modern Radar, 2021, 43 (1): 50- 57.
17	LI X Q , LIU Z M , HUANG Z . Attention-based radar PRI modulation recognition with recurrent neural networks[J]. IEEE Access, 2020, 8, 57426- 57436. doi: 10.1109/ACCESS.2020.2982654
18	LIU Z M , PHILIP S Y . Classification, denoiseng, and deinterleaving of pulse streams with recurrent neural networks[J]. IEEE Trans. on Aerospace and Electronic Systems, 2018, 55 (4): 1624- 1639.
19	姜在阳, 孙思月, 李华旺, 等. 一种基于JANET模型的雷达信号分选方法[J]. 中国科学院大学学报, 2021, 38 (6): 825- 831.
	JIANG Z Y , SUN S Y , LI H W , et al. A method for deinterleaving based on JANET[J]. Journal of University of Chinese Academy of Sciences, 2021, 38 (6): 825- 831.
20	LI X Q , LIU Z M , HUANG Z . Denoising of radar pulse streams with autoencoders[J]. IEEE Communications Letters, 2020, 24 (4): 797- 801. doi: 10.1109/LCOMM.2020.2967365
21	隋金坪, 刘振, 刘丽, 等. 雷达辐射源信号分选研究进展[J]. 雷达学报, 2022, 11 (3): 418- 433.
	SUI J P , LIU Z , LIU L , et al. Progress in radar emitter signal deinterleaving[J]. Journal of Radars, 2022, 11 (3): 418- 433.
22	LONG J, SHELHAMER E, DARRELL T. Fully convolutional networks for semantic segmentation[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2015: 3431-3440.
23	RONNEBERGER Q, FISCHER P, BROX T. U-Net: convolutional networks for biomedical image segmentation[C]//Proc. of International Conference on Medical Image Computing and Computer-Assisted Intervention, 2015: 234-241.

像素点	PDW₁	PDW₂	…	PDW_m	…	PDW_M
1	{TOA, CF, PW}_{1, 1}	{TOA, CF, PW}_{1, 2}	…	{TOA, CF, PW}_{1, m}	…	最多M个
2	{TOA, CF, PW}_{2, 1}	{TOA, CF, PW}_{2, 1}	…	{TOA, CF, PW}_{2, m}	…	最多M个
⋮	⋮	⋮	⋱	⋮	⋱	⋮
k	{TOA, CF, PW}_{k, 1}	{TOA, CF, PW}_{k, 2}	…	{TOA, CF, PW}_{k, m}	…	最多M个
⋮	⋮	⋮	⋱	⋮	⋱	⋮
K	{TOA, CF, PW}_{K, 1}	{TOA, CF, PW}_{K, 2}	…	{TOA, CF, PW}_{K, m}	…	最多M个

网络层级	反卷积个数	特征拼接	卷积个数	激活函数	归一化函数	池化次数
1	0	否	2	ReLU	BN	0
2	0	否	2	ReLU	BN	1
3	0	否	2	ReLU	BN	1
4	0	否	2	ReLU	BN	1
5	0	否	2	ReLU	BN	1
6	1	是	2	ReLU	BN	0
7	1	是	2	ReLU	BN	0
8	1	是	2	ReLU	BN	0
9	1	是	2	ReLU	BN	0
10	0	否	1	Sigmoid	无	0

优化函数	初始学习率	批量处理样本数	训练迭代次数
Adam	0.001	32	10

雷达类型	重复周期/μs	载频/MHz	脉宽/μs	跳频间隔/MHz	跳频点数
常规	[50, 1 000]	[1 000, 2 000]	[1, 200]	-	-
重频抖动(抖动率[10%, 30%])	[50, 1 000]	[1 000, 2 000]	[1, 200]	-	-
重频参差(4参差)	子周期[100, 200]	[1 000, 2 000]	[1, 200]	-	-
捷变频	[50, 1 000]	起始频率[1 000, 1 500]	[1, 200]	[15, 25]	[20, 32]
脉组捷变频	[50, 1 000]	起始频率[1 000, 1 500]	[1, 200]	[20, 50]	[5, 100]

[1]	Chunming TIAN, An YANG, Le YE, Jianxing LI, Yuchen HE. End-to-end antenna optimization based on Bayesian optimization algorithm [J]. Systems Engineering and Electronics, 2021, 43(12): 3413-3419.
[2]	ZHANG Yixiao, GUO Wenpu, KANG kai, YAO Yunlong, ZHANG Linke, ZHANG Wei. Radar signal sorting method based on data field combined PRI transform and clustering [J]. Systems Engineering and Electronics, 2019, 41(7): 1509-1515.

End-to-end radar signal sorting based on deep segmentation

RichHTML

PDF (PC)

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 13

References 23

Related Articles 2

Recommended Articles

Metrics

Comments