基于改进1DCNN+TCN的雷达辐射源快速识别方法

doi:10.12305/j.issn.1001-506X.2022.02.14

摘要/Abstract

摘要：

为了解决传统雷达辐射源识别方式识别速度慢、在低信噪比时很难准确识别等问题, 结合深度学习提出了一种基于改进一维卷积神经网络(one-dimensional convolutional neural network, 1DCNN)和时间卷积网络(temporal convolutional network, TCN)的雷达辐射源快速识别模型。在1DCNN的基础上加入了批归一化层, 并在全连接层前加入注意力机制; 同时在原有TCN的基础上进行改进, 使用Leaky ReLU激活函数代替ReLU函数; 将改进后的TCN与1DCNN相连接。仿真实验结果分析表明, 该模型不仅能够迅速识别出辐射源信号, 识别准确率也较高, 能够有效平衡模型识别速度和识别精度。

关键词: 辐射源信号快速识别, 时间序列, 时间卷积网络, 一维卷积神经网络, 参数化线性修正单元, 注意力机制

Abstract:

In order to solve the problems of low recognition speed and that it is difficult to accurately identify radar emitter in low signal-to-noise ratios (SNRs), a fast radar emitter recognition model based on improved one-dimensional convolution neural network (1DCNN) and temporal convolution network (TCN) is proposed. In this paper, a batch normalization layer is added to the 1DCNN, and the attention mechanism is added before the full connection layer; at the same time, it is improved on the basis of the original TCN, using the Leaky ReLU activation function to replace the ReLU function; and the improved TCN is connected with 1DCNN. Through the analysis of simulation results, the model can not only identify emitter signals quickly, but also have a high accuracy rate of identification, which can effectively balance the recognition speed and model recognition accuracy.

Key words: rapid identification of emitter signals, time series, temporal convolution network (TCN), one dimensional convolution network (1DCNN), parametric linear correction element, attention mechanism

中图分类号:

TN971.1

金涛, 王晓峰, 田润澜, 张歆东. 基于改进1DCNN+TCN的雷达辐射源快速识别方法[J]. 系统工程与电子技术, 2022, 44(2): 463-469.

Tao JIN, Xiaofeng WANG, Runlan TIAN, Xindong ZHANG. Rapid recognition method of radar emitter based on improved 1DCNN+TCN[J]. Systems Engineering and Electronics, 2022, 44(2): 463-469.

图/表 14

图1

图2

图3

图4

表1

图5

图6

表2

表3

表4

图7

表5

图8

图9

参考文献 31

1	WANG S Q , BAI J , HUANG X Y , et al. Analysis of radar emitter signal sorting and recognition model structure[J]. Procedia Computer Science, 2019, 154, 500- 503. doi: 10.1016/j.procs.2019.06.076
2	黄颖坤, 金炜东, 葛鹏, 等. 基于多尺度信息熵的雷达辐射源信号识别[J]. 电子与信息学报, 2019, 41 (5): 1084- 1091.
	HUANG Y K , JIN W D , GE P , et al. Radar emitter signal recognition based on multi-scale information entropy[J]. Journal of Electronics and Information, 2019, 41 (5): 1084- 1091.
3	谭龙, 潘继飞, 杨丽, 等. 基于脉冲角度特征的雷达辐射源识别及仿真实现[J]. 舰船电子工程, 2016, 36 (8): 97- 102.
	TAN L , PAN J F , YANG L , et al. Radar emitter recognition and simulation based on pulse angle feature[J]. Ship Electronic Engineering, 2016, 36 (8): 97- 102.
4	MA J , HUANG G , ZUO W F , et al. Robust radar waveform recognition algorithm based on random projections and sparse classification[J]. IET Radar, Sonar & Navigation, 2014, 8 (40): 290- 296.
5	WANG Y H , ZHANG S C , ZHANG Y W , et al. A cooperative spectrum sensing method based on empirical mode decomposition and information geometry in complex electromagnetic environment[J]. Complexity, 2019, 28 (5): 1709- 1921.
6	孟凡杰, 唐宏, 王义哲, 等. 基于时频图像纹理特征的雷达辐射源信号识别[J]. 弹箭与制导学报, 2017, 37 (3): 152- 156.
	MENG F J , TANG H , WANG Y Z , et al. Radar radiation source signal recognition based on texture characteristics of time-frequency images[J]. Journal of Missile and Guidance, 2017, 37 (3): 152- 156.
7	曹晓航, 汪立新, 束学渊. 基于小波不变矩的雷达辐射源信号识别[J]. 计算机工程与应用, 2020, 56 (19): 269- 272.
	CAO X H , WANG L X , SHU X Y . Radar emitter signal recognition based on wavelet moment invariants[J]. Computer Engineering and applications, 2020, 56 (19): 269- 272.
8	ZHANG Z , GEIGER J , POHJALAINEN J , et al. Deep learning for environmentally robust speech recognition: an overview of recent developments[J]. ACM Trans.on Intelligent Systems and Technology, 2018, 9 (5): 921- 928.
9	LI J H, ZHAO R L, HU H, et al. Improving RNN transducer modeling for end-to-end speech recognition[C]//Proc. of the IEEE Automatic Speech Recognition and Understanding Work-shop, 2019: 114-121.
10	ELBIR A M , MISHRA K V , ELDAR Y C . Cognitive radar antenna selection via deep learning[J]. IET Radar, Sonar & Navigation, 2019, 13 (6): 871- 880.
11	LI X Y , HE Y , JING X J . A survey of deep learning-based human activity recognition in radar[J]. Remote Sensing, 2019, 11 (9): 1068. doi: 10.3390/rs11091068
12	HE K S, ZHANG X, REN S, et al. Deep learning for image recognition[C]//Proc. of the Computer Pattern Recognition, 2016: 770-778.
13	LI Y , PENG C D , CHEN Y P , et al. A deep learning method for change detection in synthetic aperture radar images[J]. IEEE Trans.on Geoscience and Remote Sensing, 2019, 57 (8): 5751- 5763. doi: 10.1109/TGRS.2019.2901945
14	GURBUZ S Z , AMIN M G . Radar-based human-motion recognition with deep learning: promising applications for indoor monitoring[J]. IEEE Signal Processing Magazine, 2019, 36 (4): 16- 28. doi: 10.1109/MSP.2018.2890128
15	CUI K W , HU C , WANG R S , et al. Deep-learning-based extraction of the animal migration patterns from weather radar images[J]. Science China Information Sciences, 2020, 63 (4): 3817- 3829.
16	ZHANG M , LIU L T , DIAO M . LPI radar waveform recognition based on time-frequency distribution[J]. Sensors, 2016, 16, 1682- 1684. doi: 10.3390/s16101682
17	LI Y B , GE J , LIN Y , et al. Radar emitter signal recognition based on multi-scale wavelet entropy and feature weighting[J]. Journal of Central South University, 2014, 21 (11): 4254- 4260. doi: 10.1007/s11771-014-2422-5
18	KONG S H , KIM M , HOANG L M . Automatic LPI radar waveform recognition using CNN[J]. IEEE Access, 2018, 6, 4207- 4219. doi: 10.1109/ACCESS.2017.2788942
19	WAN J , YU X , GUO Q . LPI radar wavefor m recognition based on CNN and TPOT[J]. Symmetry, 2019, 11 (5): 725- 731. doi: 10.3390/sym11050725
20	CAIN L, CLARK J, PAULS E, et al. Convolutional neural networks for radar emitter classification[C]//Proc. of the IEEE 8th Annual Computing and Communication Workshop and Conference, 2018: 79-83.
21	BAI S, KOLTER J Z, KOLTUN V. An empirical evaluation of generic convolutional and recurrent networks for sequence mo-deling[R]. Pittsburgh: Carnegie Mellon University, 2018: 7-9.
22	HE K M, ZHANG X Y, REN S Q, et al. Identity mappings in deep residual networks[C]//Proc. of the 14th European Conference on Computer Vision, 2016: 630-645
23	OORD A, DIELEMAN S, ZEN H, et al. Wave net: a generative model for raw audi[EB/OL]. [2019-06-20]. https://arxiv.org/abs/1609.03499v1.
24	YU F, KOLTUN V. Multi-scale context aggregation by dilated convolutions[EB/OL]. [2019-06-20]. https://arxiv.org/abs/1511.07122.
25	NAIR V, HINTON G E. Rectified linear units improve restric-ted Boltzmann machines[C]//Proc. of the 27th IEEE International Conference on Machine Learning, 2010: 807-814.
26	MAAS A L , HANNUN A Y , NG A Y . Rectifier nonlinearities improve neural network acoustic models[J]. Machine Learning, 2013, 30 (1): 3- 8.
27	CHEN L, ZHANG H W, XIAO J, et al. SCACNN: spatial and channel-wise attention in convolutional networks for image captioning[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 6298-6306.
28	IOFFE S, SZEGEDY C. Batch normalization: accelerating deep network training by reducing internal covariate shift[C]//Proc. of the International Conference on Machine Learning, 2015: 448-456.
29	ZHANG Y D , GOVINDARAJ V V , TANG C , et al. High performance multiple sclerosis classification by data augmentation and AlexNet transfer learning model[J]. Journal of Medical Imaging and Health Informatics, 2019, 9 (9): 2012- 2021. doi: 10.1166/jmihi.2019.2692
30	WEN L, LI X, LI X Y, et al. A new transfer learning based on VGG-19 network for fault diagnosis[C]//Proc. of the IEEE 23rd International Conference on Computer Supported Cooperative Work in Design, 2019: 205-209.
31	AYYACHAMY S, ALEX V, KHENED M, et al. Medical image retrieval using Resnet-18[C]//Proc. of the Medical Imaging: Imaging Informatics for Healthcare, Research, and Applications, 2019: 1095410.

信号	主要参数	取值
BPSK	巴克码位数	{7, 11, 13}
多相码	码位数	{36, 64}
FMCW	采样频率/kHz 调制周期/ms 调制带宽/kHz	{15, 17} {50, 25, 35} {0.25, 0.35, 0.5}
Costas	频率序列/kHz	{[4 7 1 6 5 2 3], [2 6 3 8 7 5 1]}

网络模型	训练时间/s	准确率	损失
TCN	139	0.955 0	0.146 9
TCN+注意力	241	0.967 3	0.124 5

网络模型	训练时间/s	准确率	损失
TCN	139	0.955 0	0.146 9
TCN+注意力	241	0.967 3	0.124 5
CTA	147	0.967 6	0.111 3

对比项	ReLU	ELUS	Leaky ReLU
损失	0.121 4	0.117 1	0.111 3
准确率	0.964 9	0.965 6	0.967 6

网络模型	训练时间/s	训练轮数	准确率	损失
VGG19	1 363	6	0.948 7	0.152 5
ResNet18	1 490	7	0.948 0	0.153 8
AlexNet	474	5	0.917 1	0.261 3
CTA	147	11	0.967 6	0.111 3
TCN	139	8	0.955 0	0.146 9