基于双网络协同的多通道雷达前视成像方法

周卓洁; 李悦丽; 刘可; 朱巧鹏; 肖志飞; 代大海

doi:10.12000/JR25121

基于双网络协同的多通道雷达前视成像方法

DOI: 10.12000/JR25121 CSTR: 32380.14.JR25121

国防科技大学电子科学学院长沙 410073

基金项目: 国家部委基金

详细信息

作者简介:
周卓洁，硕士生，研究方向为雷达智能感知与处理

李悦丽，博士，教授，研究方向为机载雷达合成孔径成像、前视成像以及射频干扰抑制技术

刘　可，硕士，研究方向为雷达智能感知与处理

朱巧鹏，硕士生，研究方向为MIMO雷达前视成像技术

肖志飞，硕士生，研究方向为雷达智能感知与处理

代大海，博士，教授，主要研究方向为极化雷达成像、雷达信号处理与目标识别以及合成孔径雷达对抗

通讯作者:
李悦丽 liyueli4uwb@nudt.edu.cn

责任主编：陈洪猛 Corresponding Editor: CHEN Hongmeng
中图分类号: TN957.52
计量
- 文章访问数:
- HTML全文浏览量:
- PDF下载量:
- 被引次数: 0
出版历程
- 收稿日期: 2025-07-04
- 修回日期: 2025-10-26

Multichannel Radar Forward-looking Imaging Method Based on Dual-network Cooperation

College of Electronic Science and Technology, National University of Defense Technology, Changsha 410073, China

Funds: The National Ministries Foundation

More Information

Corresponding author: LI Yueli, liyueli4uwb@nudt.edu.cn

摘要

摘要: 针对雷达正前视方向多普勒梯度消失导致多目标分辨困难以及前视图像模糊的问题，该文提出一种基于双网络协同的多通道雷达前视成像方法，构建了一个分层级联的端到端处理框架：首先，设计轻量化目标数量估计网络(NEN)，基于回波协方差矩阵特征预测主瓣内目标数量；其次，根据目标数量动态选择预训练的角度估计网络(AEN)模型，实现高精度的目标方位角估计；最后，将目标数量与角度估计值作为先验信息，结合迭代自适应算法完成目标强度估计和二维投影成像。仿真和实测实验结果表明：相比于传统超分辨算法，所提方法在正前视区域能够更有效实现对强弱点目标参数的同时估计和精确重构，在目标数量估计上的准确率达到86.75%，角度估计均方根误差在双目标场景下低于0.2°，有效提高了前视成像质量。
- 多通道雷达 /
- 多目标分辨 /
- 前视成像 /
- 双网络协同 /
- 参数估计
Abstract: The challenge of distinguishing multiple targets and mitigating image blurry caused by Doppler gradient disappearance in the forward-looking direction of moving platforms is addressed through a multichannel radar forward-looking imaging method based on dual-network collaboration. The proposed method establishes a hierarchical, cascaded, end-to-end processing framework. First, a target number estimation network predicts the number of targets within the main lobe by analyzing the characteristics of the echo covariance matrix. Then, according to the estimated target count, a pretrained angle estimation network model is dynamically selected to determine the azimuth angles of the targets. Finally, target intensity estimation and two-dimensional projection imaging are performed in combination with an improved iterative adaptive algorithm. Simulation and experimental results demonstrate that, compared with conventional super-resolution algorithms, the proposed method achieves more effective simultaneous estimation and accurate reconstruction of parameters for both strong and weak targets in the forward-looking region. Specifically, it attains 86.75% accuracy in target number estimation, while the root mean square error of angle estimation remains below 0.2° in two-target scenarios, significantly enhancing the quality of forward-looking imaging.
- Multi-channel radar /
- Distinguish multiple targets /
- Forward-looking imaging /
- Dual-network collaboration /
- Parameter estimation

HTML全文

图 1 空载平台多通道雷达前视成像几何示意图

Figure 1. Airborne multi-channel radar forward-looking imaging geometry

下载: 全尺寸图片幻灯片

图 2 NEN网络结构

Figure 2. Network structure of NEN

下载: 全尺寸图片幻灯片

图 3 一维SE网络结构

Figure 3. Network structure of 1D SE

下载: 全尺寸图片幻灯片

图 4 AEN网络结构

Figure 4. Network structure of AEN

下载: 全尺寸图片幻灯片

图 5 NEN-AEN级联网络成像算法流程

Figure 5. The imaging procedure of the NEN-AEN cascaded networks

下载: 全尺寸图片幻灯片

图 6 不同方法目标个数估计性能对比

Figure 6. Performance comparison of target number estimation by different methods

下载: 全尺寸图片幻灯片

图 7 不同目标个数角度估计RMSE随信噪比变化曲线

Figure 7. The variation curves of RMSE with signal-to-noise ratio estimated from different angles of the number of targets

下载: 全尺寸图片幻灯片

图 8 相同功率的点阵目标前视成像效果对比

Figure 8. Comparison of forward looking imaging performance of lattice targets with the same power distribution

下载: 全尺寸图片幻灯片

图 9 不同功率分布点阵目标前视成像效果对比

Figure 9. Comparison of forward looking imaging performance of lattice targets with different power distribution

下载: 全尺寸图片幻灯片

图 10 不同信噪比条件下SSIM, PSNR和MSE指标结果

Figure 10. The results of SSIM, PSNR and MSE metrics varying with SNR

下载: 全尺寸图片幻灯片

图 11 仿真场景前视成像效果对比

Figure 11. Comparison of experimental results in forward looking imaging

下载: 全尺寸图片幻灯片

图 12 实验场景和V型阵坐标示意图

Figure 12. The experimental scenario and V-shaped array coordinates

下载: 全尺寸图片幻灯片

图 13 实测场景前视成像效果对比

Figure 13. Comparison of experimental results in forward looking imaging

下载: 全尺寸图片幻灯片

表 1 前视扫描成像实验仿真参数

Table 1. Simulation parameters of the forward-looking scanning radar

参数	数值	参数	数值
通道数	4	信号脉宽	1 μs
通道相位中心间隔/波长	2	脉冲重复频率PRF	2000 Hz
雷达中心频率	18 GHz	方位向主瓣宽度	5°
平台飞行速度	100 m/s	波束扫描范围	–15°～15°
信号带宽	50 MHz	天线扫描速度	30 °/s
波位脉冲数	32	脉压后信噪比	20 dB

下载: 导出CSV

表 2 AEN对比试验点目标设置仿真参数

Table 2. Simulation parameters for point targets in the AEN comparison experiment

目标个数	方位角(°)	散射强度(dB)
K=1	–0.5	10
K=2	[–0.5, 1.5]	[5, 15]
K=3	[–1.5, 0.5, 2.0]	[1, 10, 19]

下载: 导出CSV

表 3 V型阵成像场景仿真定量评估结果

Table 3. Quantitative evaluation results of V-shaped array imaging scene simulation

方法	强度RMSE	角度RMSE
IAA算法	0.534	0.354
迭代超分辨算法	0.548	0.264
FIIB算法	0.471	0.318
NEN-AEN级联网络	0.105	0.212

下载: 导出CSV

参考文献(46)

[1]	MAO Deqing, ZHANG Yongchao, PEI Jifang, et al. Forward-looking geometric configuration optimization design for spaceborne-airborne multistatic synthetic aperture radar[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2021, 14: 8033–8047. doi: 10.1109/JSTARS.2021.3103802.
[2]	CHEN Hongmeng, LI Yachao, GAO Wenquan, et al. Bayesian forward-looking superresolution imaging using Doppler deconvolution in expanded beam space for high-speed platform[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 5105113. doi: 10.1109/TGRS.2021.3107717.
[3]	吴迪, 朱岱寅, 朱兆达. 机载雷达单脉冲前视成像算法[J]. 中国图象图形学报, 2010, 15(3): 462–469. doi: 10.11834/jig.20100317. WU Di, ZHU Daiyin, and ZHU Zhaoda. Research on nomopulse forward-looking imaging algorithm for airborne radar[J]. Journal of Image and Graphics, 2010, 15(3): 462–469. doi: 10.11834/jig.20100317.
[4]	吴迪, 朱岱寅, 田斌, 等. 单脉冲成像算法性能分析[J]. 航空学报, 2012, 33(10): 1905–1914. WU Di, ZHU Daiyin, TIAN Bin, et al. Performance evaluation for monopulse imaging algorithm[J]. Acta Aeronautica et Astronautica Sinica, 2012, 33(10): 1905–1914.
[5]	LIU Mingjie, WU Di, and REN Lingyun. Monopulse imaging technology based on super-resolution in Doppler domain[C]. The 2nd International Conference on Consumer Electronics and Computer Engineering, Guangzhou, China, 2022: 186–190. doi: 10.1109/ICCECE54139.2022.9712743.
[6]	CHENG Cheng, ZHOU Xiaodong, GAO Min, et al. Research on monopulse forward-looking high-resolution imaging algorithm based on adaptive iteration[J]. Defence Technology, 2020, 16(1): 158–171. doi: 10.1016/j.dt.2019.06.008.
[7]	李悦丽, 梁甸农, 黄晓涛. 一种单脉冲雷达多通道解卷积前视成像方法[J]. 信号处理, 2007, 23(5): 699–703. doi: 10.3969/j.issn.1003-0530.2007.05.013. LI Yueli, LIANG Diannong, and HUANG Xiaotao. A multi-channel deconvolution based on forword-looking imaging method in monopulse radar[J]. Signal Processing, 2007, 23(5): 699–703. doi: 10.3969/j.issn.1003-0530.2007.05.013.
[8]	刘可, 李悦丽, 戴永鹏, 等. 基于快速迭代插值多普勒频率估计的单脉冲前视成像技术[J]. 雷达学报, 2023, 12(6): 1138–1154. doi: 10.12000/JR23145. LIU Ke, LI Yueli, DAI Yongpeng, et al. Monopulse forward-looking imaging based on Doppler estimation using fast iterative interpolated beamforming algorithm[J]. Journal of Radars, 2023, 12(6): 1138–1154. doi: 10.12000/JR23145.
[9]	LI Zhongyu, LI Shanchuan, LIU Zhutian, et al. Bistatic forward-looking SAR MP-DPCA method for space-time extension clutter suppression[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 58(9): 6565–6579. doi: 10.1109/TGRS.2020.2977982.
[10]	YANG Jianyu, GU Xueyu, LI Wenchao, et al. Configuration design of bistatic forward-looking SAR driven by spatial resolution metrics[J]. IEEE Geoscience and Remote Sensing Letters, 2025, 22: 4001905. doi: 10.1109/LGRS.2024.3521450.
[11]	刘裕洲, 蔡天倚, 李亚超, 等. 联合距离方位二维NCS的星弹双基前视SAR成像算法[J]. 雷达学报, 2023, 12(6): 1202–1214. doi: 10.12000/JR23144. LIU Yuzhou, CAI Tianyi, LI Yachao, et al. A range and azimuth combined two-dimensional NCS algorithm for spaceborne-missile bistatic forward-looking SAR[J]. Journal of Radars, 2023, 12(6): 1202–1214. doi: 10.12000/JR23144.
[12]	杨泽慧, 聂炜航, 程高峰, 等. 全变分约束的解卷积常规波束形成方位谱估计算法[J]. 声学学报, 2025, 50(1): 68–76. doi: 10.12395/0371-0025.2023173. YANG Zehui, NIE Weihang, CHENG Gaofeng, et al. Total variation constrained deconvolved conventional beamforming algorithm for azimuthal spectral estimation[J]. Acta Acustica, 2025, 50(1): 68–76. doi: 10.12395/0371-0025.2023173.
[13]	TANG Junkui, RAN Lei, LIU Zheng, et al. A weighted low-rank and sparse constraint-based multichannel radar forward-looking imaging method[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2025, 18: 12973–12987. doi: 10.1109/JSTARS.2025.3568783.
[14]	TUO Xingyu, ZHANG Yin, HUANG Yulin, et al. A fast sparse azimuth super-resolution imaging method of real aperture radar based on iterative reweighted least squares with linear sketching[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2021, 14: 2928–2941. doi: 10.1109/JSTARS.2021.3061430.
[15]	李悦丽, 马萌恩, 赵崇辉, 等. 基于单脉冲雷达和差通道多普勒估计的前视成像[J]. 雷达学报, 2021, 10(1): 131–142. doi: 10.12000/JR20111. LI Yueli, MA Mengen, ZHAO Chonghui, et al. Forward-looking imaging via Doppler estimates of sum-difference measurements in scanning monopulse radar[J]. Journal of Radars, 2021, 10(1): 131–142. doi: 10.12000/JR20111.
[16]	DAI Shengli and WIESBECK W. High resolution imaging for forward looking SAR with multiple receiving antennas[C]. IEEE 2000 International Geoscience and Remote Sensing Symposium. Taking the Pulse of the Planet: The Role of Remote Sensing in Managing the Environment. Proceedings, Honolulu, USA, 2000: 1433–1435. doi: 10.1109/IGARSS.2000.858373.
[17]	DAI Shengli and WIESBECK W. The imaging mode of forward looking SAR with two receiving antennas[C]. 1999 International Geoscience and Remote Sensing Symposium, Hamburg, Germany, 1999: 1776–1778. doi: 10.1109/IGARSS.1999.772092.
[18]	王鑫硕, 卢景月, 孟智超, 等. 前视多通道SAR成像及阵列姿态误差补偿[J]. 雷达学报, 2023, 12(6): 1155–1165. doi: 10.12000/JR23073. WANG Xinshuo, LU Jingyue, MENG Zhichao, et al. Forward-looking multi-channel synthetic aperture radar imaging and array attitude error compensation[J]. Journal of Radars, 2023, 12(6): 1155–1165. doi: 10.12000/JR23073.
[19]	CHEN Rui, LI Wenchao, ZHANG Yongchao, et al. Forward looking imaging of airborne multichannel radar based on modified IAA[C]. 2022 IEEE International Geoscience and Remote Sensing Symposium, Kuala Lumpur, Malaysia, 2022: 2987–2990. doi: 10.1109/IGARSS46834.2022.9884918.
[20]	王宁, 贺鹏超, 卢景月, 等. 基于DOA估计的前视多通道SAR成像方法[J]. 系统工程与电子技术, 2023, 45(8): 2471–2478. doi: 10.12305/j.issn.1001-506X.2023.08.21. WANG Ning, HE Pengchao, LU Jingyue, et al. DOA estimation based imaging method for multi-channel forward-looking SAR[J]. Systems Engineering and Electronics, 2023, 45(8): 2471–2478. doi: 10.12305/j.issn.1001-506X.2023.08.21.
[21]	杨洋, 李悦丽. 单脉冲前视成像多目标分辨算法[J]. 信号处理, 2016, 32(9): 1055–1064. doi: 10.16798/j.issn.1003-0530.2016.09.07. YANG Yang and LI Yueli. Multi-targets discrimination algorithm in monopulse forward-looking imaging[J]. Journal of Signal Processing, 2016, 32(9): 1055–1064. doi: 10.16798/j.issn.1003-0530.2016.09.07.
[22]	陈洪猛, 鲁耀兵, 刘京, 等. 一种用于月面着陆的知识辅助单脉冲前视成像方法[J]. 载人航天, 2019, 25(1): 31–36. doi: 10.3969/j.issn.1674-5825.2019.01.005. CHEN Hongmeng, LU Yaobing, LIU Jing, et al. A knowledge aided monopulse forward-looking imaging algorithm for lunar landing[J]. Manned Spaceflight, 2019, 25(1): 31–36. doi: 10.3969/j.issn.1674-5825.2019.01.005.
[23]	陈洪猛, 李明, 王泽玉, 等. 基于多帧数据联合处理的机载单通道雷达贝叶斯前视成像[J]. 电子与信息学报, 2015, 37(10): 2328–2334. doi: 10.11999/JEIT150153. CHEN Hongmeng, LI Ming, WANG Zeyu, et al. Bayesian forward-looking imaging for airborne single-channel radar based on combined multiple frames data[J]. Journal of Electronics & Information Technology, 2015, 37(10): 2328–2334. doi: 10.11999/JEIT150153.
[24]	陈洪猛, 余继周, 张文杰, 等. 基于概率模型驱动的机载贝叶斯前视超分辨多目标成像方法[J]. 雷达学报, 2023, 12(6): 1125–1137. doi: 10.12000/JR23080. CHEN Hongmeng, YU Jizhou, ZHANG Wenjie, et al. Probability model-driven airborne Bayesian forward-looking super-resolution imaging for multitarget scenario[J]. Journal of Radars, 2023, 12(6): 1125–1137. doi: 10.12000/JR23080.
[25]	张永超, 孙震宇, 蔡晓春, 等. 实孔径雷达无超参数全变差正则化角超分辨方法[J]. 雷达学报(中英文), 待出版. doi: 10.12000/JR25011. ZHANG Yongchao, SUN Zhenyu, CAI Xiaochun, et al. A hyperparameter-free total variation regularization method for real aperture radar angular super-resolution[J]. Journal of Radars, in press. doi: 10.12000/JR25011.
[26]	ZHANG Jie, WU Di, ZHU Daiyin, et al. An airborne/missile-borne array radar forward-looking imaging algorithm based on super-resolution method[C]. 2017 10th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics, Shanghai, China, 2017: 1–5. doi: 10.1109/CISP-BMEI.2017.8302131.
[27]	XI Rongyan, ZHENG Chundi, HUANG Tianyao, et al. Joint range and angle estimation for wideband forward-looking imaging radar[J]. IEEE Sensors Journal, 2022, 22(1): 446–460. doi: 10.1109/JSEN.2021.3126206.
[28]	ZHOU Zhuojie, LI Yueli, and LIU Ke. A Doppler estimation algorithm for radar imaging based on low-rank embedding theory[C]. 2024 IEEE 17th International Conference on Signal Processing, Suzhou, China, 2024: 821–826. doi: 10.1109/ICSP62129.2024.10846578.
[29]	任凌云, 吴迪, 朱岱寅, 等. 基于机载多通道雷达迭代超分辨估计的前视成像[J]. 雷达学报, 2023, 12(6): 1166–1178. doi: 10.12000/JR23085. REN Lingyun, WU Di, ZHU Daiyin, et al. Forward-looking imaging via iterative super-resolution estimation in airborne multi-channel radar[J]. Journal of Radars, 2023, 12(6): 1166–1178. doi: 10.12000/JR23085.
[30]	LIU Ke, LI Yueli, XU Zhou, et al. Airborne multi-channel forward-looking radar super-resolution imaging using improved fast iterative interpolated beamforming algorithm[J]. Remote Sensing, 2024, 16(22): 4121. doi: 10.3390/rs16224121.
[31]	罗迎, 倪嘉成, 张群. 基于“数据驱动+智能学习”的合成孔径雷达学习成像[J]. 雷达学报, 2020, 9(1): 107–122. doi: 10.12000/JR19103. LUO Ying, NI Jiacheng, and ZHANG Qun. Synthetic aperture radar learning-imaging method based on data-driven technique and artificial intelligence[J]. Journal of Radars, 2020, 9(1): 107–122. doi: 10.12000/JR19103.
[32]	ZHOU Jie, LIU Yongxiang, PENG Bowen, et al. MaDiNet: Mamba diffusion network for SAR target detection[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2025, early access. doi: 10.1109/TCSVT.2025.3574657.
[33]	WU Xiaohuan, YANG Xu, JIA Xiaoyuan, et al. A gridless DOA estimation method based on convolutional neural network with toeplitz prior[J]. IEEE Signal Processing Letters, 2022, 29: 1247–1251. doi: 10.1109/LSP.2022.3176211.
[34]	张群, 张宏伟, 倪嘉成, 等. 合成孔径雷达深度学习成像研究综述[J]. 信号处理, 2023, 39(9): 1521–1551. doi: 10.16798/j.issn.1003-0530.2023.09.001. ZHANG Qun, ZHANG Hongwei, NI Jiacheng, et al. A survey of synthetic aperture radar imaging methods based on deep learning[J]. Journal of Signal Processing, 2023, 39(9): 1521–1551. doi: 10.16798/j.issn.1003-0530.2023.09.001.
[35]	吴明华, 饶彬, 王伟. 基于深度残差网络的雷达目标数量估计方法[J]. 太赫兹科学与电子信息学报, 2022, 20(3): 213–217. doi: 10.11805/TKYDA2021354. WU Minghua, RAO Bin, and WANG Wei. Radar target number estimation method based on deep residual network[J]. Journal of Terahertz Science and Electronic Information Technology, 2022, 20(3): 213–217. doi: 10.11805/TKYDA2021354.
[36]	孙晓翰, 李凉海, 张彬. 基于CNN-LSTM神经网络的前视成像算法[J]. 遥测遥控, 2024, 45(2): 29–36. doi: 10.12347/j.ycyk.20231225001. SUN Xiaohan, LI Lianghai, and ZHANG Bin. Forward-looking imaging algorithm based on CNN-LSTM neural network[J]. Journal of Telemetry, Tracking and Command, 2024, 45(2): 29–36. doi: 10.12347/j.ycyk.20231225001.
[37]	ZHANG Ye, YANG Qi, ZENG Yang, et al. High-quality interferometric inverse synthetic aperture radar imaging using deep convolutional networks[J]. Microwave and Optical Technology Letters, 2020, 62(9): 3060–3065. doi: 10.1002/mop.32411.
[38]	任凌云. 机载多通道雷达超分辨前视成像技术研究[D]. [硕士论文], 南京航空航天大学, 2023: 26-27. doi: 10.27239/d.cnki.gnhhu.2023.000880. REN Lingyun. Research on Airborne Multi-Channel Radar Super-Resolution Forward Looking Imaging[D], [Master dissertation], Nanjing University of Aeronautics and Astronautics, 2023: 26-27. doi: 10.27239/d.cnki.gnhhu.2023.000880.
[39]	STOICA P and NEHORAI A. MUSIC, maximum likelihood, and Cramer-Rao bound[J]. IEEE Transactions on Acoustics, Speech, and Signal Processing, 1989, 37(5): 720–741. doi: 10.1109/29.17564.
[40]	ROY III R H and KAILATH T. ESPRIT-estimation osignal parameters via rotational invariance techniques[J]. IEEE Transactions on Acoustics, Speech, and Signal Processing, 1989, 37(7): 984–995. doi: 10.1109/29.32276.
[41]	HU Jie, SHEN Li, ALBANIE S, et al. Squeeze-and-excitation networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020, 42(8): 2011–2023. doi: 10.1109/TPAMI.2019.2913372.
[42]	潘耀雄. 基于深度神经网络的机载雷达前视成像技术研究[D]. [硕士论文], 南京航空航天大学, 2023: 120–137. doi: 10.27239/d.cnki.gnhhu.2023.003682. PAN Yaoxiong. Research on forward-looking imaging technology of airborne radar based on deep neural network[D]. [Master dissertation], Nanjing University of Aeronautics and Astronautics, 2023: 120–137. doi: 10.27239/d.cnki.gnhhu.2023.003682.
[43]	WAX M and KAILATH T. Detection of signals by information theoretic criteria[J]. IEEE Transactions on Acoustics, Speech, and Signal Processing, 1985, 33(2): 387–392. doi: 10.1109/TASSP.1985.1164557.
[44]	KUNDU D and MITRA A. Estimating the number of signals of the damped exponential models[J]. Computational Statistics & Data Analysis, 2001, 36(2): 245–256. doi: 10.1016/S0167-9473(00)00036-0.
[45]	KASE Y, NISHIMURA T, OHGANE T, et al. DOA estimation of two targets with deep learning[C]. 2018 15th Workshop on Positioning, Navigation and Communications, Bremen, Germany, 2018: 1–5. doi: 10.1109/WPNC.2018.8555814.
[46]	ROBERTS W, STOICA P, LI Jian, et al. Iterative adaptive approaches to MIMO radar imaging[J]. IEEE Journal of Selected Topics in Signal Processing, 2010, 4(1): 5–20. doi: 10.1109/JSTSP.2009.2038964.