轻量化的ML-SNet雷达复合干扰识别算法

doi:10.12305/j.issn.1001-506X.2025.02.09

摘要/Abstract

摘要：

针对复杂电磁环境下雷达复合干扰识别困难和网络模型复杂度高的问题, 将多标签分类与改进的ShuffleNet V2相结合, 提出一种轻量化的多标签ShuffleNet(multi-labeling ShuffleNet, ML-SNet)雷达复合干扰识别算法。首先, 使用轻量化的ShuffleNet V2作为主干网络, 引入SimAM(similarity-based attention module)注意力机制, 提高网络特征提取能力。其次, 使用漏斗激活线性整流函数(funnel activation rectified linear unit, FReLU)代替线性整流单元(rectified linear unit, ReLU)激活函数, 减少特征图的信息损失。最后, 使用多标签分类算法对网络输出进行分类, 得到识别结果。实验结果表明, 在干噪比范围为-10~10 dB的情况下, 所提算法对15类雷达复合干扰的平均识别率为97.9%。与其他网络相比, 所提算法具有较低的计算复杂度, 而且识别性能表现最佳。

关键词: 复合干扰识别, 多标签分类, 轻量化, 计算复杂度

Abstract:

Aiming at the problems of difficult radar compound jamming recognition and high complexity of network model in complex electromagnetic environments, a lightweight multi-labeling ShuffleNet (ML-SNet) radar compound jamming recognition algorithm is proposed by combining multi-label classification with the improved ShuffleNet V2. Firstly, the lightweight ShuffleNet V2 is used as the backbone network, and the similarity-based attention module (SimAM) attention mechanism is introduced to improve the network feature extraction capability. Secondly, the funnel activation rectified linear unit (FReLU) activation function is used instead of the rectified linear unit (ReLU) activation function to reduce the information loss of the feature map. Finally, the recognition results are obtained by classifying the network output using a multi-label classification algorithm. Experimental results indicate that the proposed algorithm achieves an average recognition rate of 97.9% for 15 classes of radar compound jamming with the jamming-to-noise ratio of -10-10 dB. The proposed algorithm has a lower computational complexity and the best performance in terms of recognition performance compared to other networks.

Key words: compound jamming recognition, multi-label classification, lightweight, computational complexity

中图分类号:

TN974

郭立民, 黄文青, 陈前, 王佳宾. 轻量化的ML-SNet雷达复合干扰识别算法[J]. 系统工程与电子技术, 2025, 47(2): 418-427.

Limin GUO, Wenqing HUANG, Qian CHEN, Jiabin WANG. Lightweight algorithm of ML-SNet radar compound jamming recognition[J]. Systems Engineering and Electronics, 2025, 47(2): 418-427.

图/表 12

图1

图2

图3

图4

图5

表1

图6

图7

表2

图8

表3

表4

参考文献 30

1	GENG Z , YAN H J , ZHANG J D , et al. Deep-learning for radar: a survey[J]. IEEE Access, 2021, 9, 141800- 141818. doi: 10.1109/ACCESS.2021.3119561
2	LI K , BO J , WANG P H , et al. Radar active antagonism through deep reinforcement learning: a way to address the challenge of mainlobe jamming[J]. Signal Processing, 2021, 186, 108130. doi: 10.1016/j.sigpro.2021.108130
3	TAN M , WANG C Y , XUE B , et al. A novel deceptive jamming approach against frequency diverse array radar[J]. IEEE Sensors Journal, 2020, 21 (6): 8323- 8332.
4	LYU Q Z , QUAN Y H , FENG W , et al. Radar deception jamming recognition based on weighted ensemble CNN with transfer learning[J]. IEEE Trans.on Geoscience and Remote Sensing, 2021, 60, 5107511.
5	ZHAO Y J , TIAN B , WANG C Y , et al. Research on main-lobe deceptive jamming against FDA-MIMO radar[J]. IET Radar, Sonar & Navigation, 2021, 15 (6): 641- 654.
6	LUO Z , CAO Y , YEO T S , et al. Few-shot radar jamming re-cognition network via time-frequency self-attention and global knowledge distillation[J]. IEEE Trans.on Geoscience and Remote Sensing, 2023, 61, 5105612.
7	SHI C W , ZHANG Q , LIN T , et al. Recognition of micro-motion jamming based on complex-valued convolutional neural network[J]. Sensors, 2023, 23 (3): 1118. doi: 10.3390/s23031118
8	ZHOU H P , WANG L , MA M , et al. Compound radar jamming recognition based on signal source separation[J]. Signal Processing, 2024, 214, 109246. doi: 10.1016/j.sigpro.2023.109246
9	SHAO G P , CHEN Y , WEI Y . Deep fusion for radar jamming signal classification based on CNN[J]. IEEE Access, 2020, 8, 117236- 117244. doi: 10.1109/ACCESS.2020.3004188
10	LIN A , MA Z Y , HUANG Z , et al. Unknown radar waveform recognition based on transferred deep learning[J]. IEEE Access, 2020, 8, 184793- 184807. doi: 10.1109/ACCESS.2020.3029192
11	ZHAO S S , ZHOU Y , ZHANG L R , et al. Discrimination between radar targets and deception jamming in distributed multiple-radar architectures[J]. IET Radar, Sonar & Navigation, 2017, 11 (7): 1124- 1131.
12	GRECO M , GINI F , FAFINA A . Radar detection and classification of jamming signals belonging to a cone class[J]. IEEE Trans.on Signal Procession, 2008, 56 (6): 1984- 1993.
13	SU D T, GAO M G. Research on jamming recognition techno-logy based on characteristic parameters[C]//Proc. of the IEEE 5th International Conference on Signal and Image Processing, 2020: 303-307.
14	朱峰, 蒋倩倩, 林川, 等. 基于支持向量机的典型宽带电磁干扰源识别[J]. 系统工程与电子技术, 2021, 43 (9): 2400- 2406.
	ZHU F , JIANG Q Q , LIN C , et al. Typical wideband EMI identification based on support vector machine[J]. Systems Engineering and Electronics, 2021, 43 (9): 2400- 2406.
15	GAO M, LI H T, JIAO B, et al. Simulation research on classification and identification of typical active jamming against LFM radar[C]//Proc. of the 17th International Conference on Signal Processing Systems, 2019: 214-221.
16	YAO Z K , TANG C B , WANG C , et al. Cooperative jamming resource allocation model and algorithm for netted radar[J]. Electronics Letters, 2022, 58 (22): 834- 836. doi: 10.1049/ell2.12611
17	LING Q , HUANG P H , WANG D H , et al. Range deception jamming performance evaluation for moving targets in a ground-based radar network[J]. Electronics, 2023, 12 (7): 1614. doi: 10.3390/electronics12071614
18	FENG L W , LIU S T , XU H Z . Multifunctional radar cognitive jamming decision based on dueling double deep Q-network[J]. IEEE Access, 2022, 10, 112150- 112157. doi: 10.1109/ACCESS.2022.3214842
19	QU Q Z , WEI S J , LIU S , et al. JRNet: jamming recognition networks for radar compound suppression jamming signals[J]. IEEE Trans.on Vehicular Technology, 2020, 69 (12): 15035- 15045. doi: 10.1109/TVT.2020.3032197
20	SHAO G Q , CHEN Y S , WEI Y S . Deep fusion for radar jamming signal classification based on CNN[J]. IEEE Access, 2020, 8, 117236- 117244. doi: 10.1109/ACCESS.2020.3004188
21	张顺生, 陈爽, 陈晓莹, 等. 面向小样本的多模态雷达有源欺骗干扰识别方法[J]. 雷达学报, 2023, 12 (4): 882- 891.
	ZHANG S S , CHEN S , CHEN X Y , et al. Active deception jamming recognition method in multimodal radar based on small samples[J]. Journal of Radars, 2023, 12 (4): 882- 891.
22	ZHANG J X , ZHOU Z , LIU C , et al. Radar compound jamming cognition based on a deep object detection network[J]. IEEE Trans.on Aerospace and Electronic Systems, 2022, 59 (3): 3251- 3263.
23	FU R R. Compound jamming signal recognition based on neural networks[C]//Proc. of the International Conference on Instrumentation & Measurement, Computer, Communication and Control, 2016: 737-740.
24	XIAO Y H , ZHOU J Y , YU Y Z , et al. Active jamming recognition based on bilinear EfficientNet and attention mechanism[J]. IET Radar, Sonar & Navigation, 2021, 15 (9): 957- 968.
25	ZHANG S J , YANG H , YUAN W , et al. Edge device detection of tea leaves with one bud and two leaves based on ShuffleNetV2-YOLOV5-Lite-E[J]. Agronomy, 2023, 13 (2): 557. doi: 10.3390/agronomy13020557
26	CAI Z Y , QIAO X Y , ZHANG J W , et al. RepVGG-SimAM: an efficient bad image classification method based on RepVGG with simple parameter-free attention module[J]. Applied Sciences, 2023, 13 (21): 11925. doi: 10.3390/app132111925
27	QIU S, XU X M, CAI B L. FReLU: flexible rectified linear units for improving convolutional neural networks[C]//Proc. of the 24th International Conference on Pattern Recognition, 2018: 1223-1228.
28	程玉胜, 赵大卫, 王一宾, 等. 非平衡化标签补全核极限学习机多标签学习[J]. 电子学报, 2019, 47 (3): 719- 725.
	CHENG Y S , ZHAO D W , WANG Y B , et al. Multi-label learning of kernel extreme learning machine with non-equilibrium label completion[J]. Journal of Electronics, 2019, 47 (3): 719- 725.
29	ZHU M T , LI Y J , PAN Z S , et al. Automatic modulation re-cognition of compound signals using a deep multi-label classifier: a case study with radar jamming signals[J]. Signal Processing, 2020, 169, 107393. doi: 10.1016/j.sigpro.2019.107393
30	邵正途, 许登荣, 徐文利, 等. 基于LSTM和残差网络的雷达有源干扰识别[J]. 系统工程与电子技术, 2023, 45 (2): 416- 423. doi: 10.12305/j.issn.1001-506X.2023.02.12
	SHAO Z T , XU D R , XU W L , et al. Radar active jamming recognition based on LSTM and residual network[J]. Systems Engineering and Electronics, 2023, 45 (2): 416- 423. doi: 10.12305/j.issn.1001-506X.2023.02.12

干扰类型	参数设置	训练集	测试集
CI	切片宽度/μs	2	2~3
CI	重复转发次数/次	4	4~5
ISDJ	切片宽度/μs	2	2~2.5
ISDJ	切片次数/次	4	4~5
COMB	频移个数/个	5	4~5
SMSP	调频斜率	5k	4k~5k
SMSP	转发次数/次	5	4~5
NP	带宽/MHz	10~20	8~22
NC	带宽/MHz	10~20	8~22
复合干扰	与相应的干扰分量相同	与相应的干扰分量相同	与相应的干扰分量相同

模型	准确率/%	参数量(×10⁶)	运算量(×10⁹)
1	93.1	1.51	0.13
2	96.4	1.51	0.13
3	95.3	1.71	0.13
4	96.1	1.51	0.13
5	95.5	0.94	0.09
6	97.9	1.07	0.09

干扰类型	MobileNet V2	EfficientNet	ResNet34	ShuffleNet V2	文献[24]	ML-SNet
CI+ISDJ	97.7	95.5	100.0	92.0	98.9	95.9
CI+COMB	95.5	86.4	79.5	96.6	96.6	100.0
CI+SMSP	88.6	95.5	77.3	88.6	95.5	100.0
CI+NP	94.3	87.5	84.1	96.6	92.0	99.0
CI+NC	95.5	98.9	94.3	79.5	97.7	93.1
ISDJ+COMB	97.7	96.6	96.6	79.5	97.7	94.8
ISDJ+SMSP	95.5	92.0	97.7	97.7	98.9	98.9
ISDJ+NP	89.8	96.6	63.6	96.6	93.2	97.0
ISDJ+NC	84.1	89.8	61.4	97.9	96.6	98.9
COMB+SMSP	84.1	88.6	70.5	96.6	95.5	100.0
COMB+NP	97.7	98.9	95.5	100.0	98.9	100.0
COMB+NC	99.0	98.9	97.7	85.2	98.9	99.0
SMSP+NP	95.5	99.0	81.8	98.9	98.9	99.0
SMSP+NC	95.5	95.5	70.5	81.8	93.2	99.0
NP+NC	92.0	94.3	88.6	97.7	98.9	94.8

模型	空间复杂度(×10⁶)	时间复杂度
模型	空间复杂度(×10⁶)	每轮训练时间/s	平均识别时间/ms
ML-SNet	1.07	24.62	7.38
MobileNet V2	3.26	59.66	8.24
EfficientNet	4.12	103.17	9.41
ResNet34	36.80	52.63	8.88
ShuffleNet V2	1.51	33.87	7.47
文献[24]	5.33	113.26	9.92

[1]	孙先涛, 江汪洋, 陈文杰, 陈伟海, 智亚丽. 基于感兴趣区域的物体抓取位姿检测[J]. 系统工程与电子技术, 2024, 46(6): 1867-1877.
[2]	刘绍华, 杜康, 佘春东, 杨傲. 基于CenterNet的多教师联合知识蒸馏[J]. 系统工程与电子技术, 2024, 46(4): 1174-1184.
[3]	黄思佳, 宋纯锋, 李璇. 基于可变尺度先验框的声呐图像目标检测[J]. 系统工程与电子技术, 2024, 46(3): 771-778.
[4]	崔天舒, 王栋, 黄振. 面向空间认知通信的轻量化网络自动调制分类方法[J]. 系统工程与电子技术, 2023, 45(7): 2220-2226.
[5]	宋爽, 张悦, 张琳娜, 岑翼刚, 李浥东. 基于深度学习的轻量化目标检测算法[J]. 系统工程与电子技术, 2022, 44(9): 2716-2725.
[6]	赖德迪, 罗智徽, 马应龙. 基于共现分析的分类器链标签序列优化方法[J]. 系统工程与电子技术, 2021, 43(9): 2526-2534.
[7]	丁嘉辉, 汤建龙, 于正洋. 轻量化的增量式集成学习算法设计[J]. 系统工程与电子技术, 2021, 43(4): 861-867.
[8]	焦亚萌，黄建国，侯云山. 基于蚁群算法的最大似然方位估计快速算法[J]. Journal of Systems Engineering and Electronics, 2011, 33(8): 1718-1721.
[9]	和洁, 冯大政, 吕晖, 向聪. 机载雷达三维空时自适应相关域降维算法[J]. Journal of Systems Engineering and Electronics, 2011, 33(2): 286-289.
[10]	冯道旺, 李腾, 黄知涛. 平方根二阶EKF及其在目标运动分析中的应用[J]. Journal of Systems Engineering and Electronics, 2009, 31(9): 2101-2105.
[11]	喻火根, 朱立东. 一种新的跳频宽带瑞利衰落信道模型[J]. Journal of Systems Engineering and Electronics, 2009, 31(6): 1295-1298.
[12]	肖秦琨, 高晓光, 高嵩, 王海芸. DBN结构学习度量分解性能分析[J]. Journal of Systems Engineering and Electronics, 2009, 31(4): 938-946.
[13]	金勇1,2，李捷1，黄建国2. 基于MetropolisHastings抽样短采样宽带信号方位估计AML算法[J]. Journal of Systems Engineering and Electronics, 2009, 31(12): 2809-2812.