Deep Learning(DL)model has been widely used in the field of Synthetic Aperture Radar Automatic Target Recognition(SAR-ATR)and has achieved excellent performance.However,the black-box nature of DL models has been the f...Deep Learning(DL)model has been widely used in the field of Synthetic Aperture Radar Automatic Target Recognition(SAR-ATR)and has achieved excellent performance.However,the black-box nature of DL models has been the focus of criticism,especially in the application of SARATR,which is closely associated with the national defense and security domain.To address these issues,a new interpretable recognition model Physics-Guided BagNet(PGBN)is proposed in this article.The model adopts an interpretable convolutional neural network framework and uses time–frequency analysis to extract physical scattering features in SAR images.Based on the physical scattering features,an unsupervised segmentation method is proposed to distinguish targets from the background in SAR images.On the basis of the segmentation result,a structure is designed,which constrains the model's spatial attention to focus more on the targets themselves rather than the background,thereby making the model's decision-making more in line with physical principles.In contrast to previous interpretable research methods,this model combines interpretable structure with physical interpretability,further reducing the model's risk of error recognition.Experiments on the MSTAR dataset verify that the PGBN model exhibits excellent interpretability and recognition performance,and comparative experiments with heatmaps indicate that the physical feature guidance module presented in this article can constrain the model to focus more on the target itself rather than the background.展开更多
Target recognition is a significant part of a Ballistic Missile Defense System(BMDS).However,most existing ballistic target recognition methods overlook the impact of data representation on recognition outcomes.This p...Target recognition is a significant part of a Ballistic Missile Defense System(BMDS).However,most existing ballistic target recognition methods overlook the impact of data representation on recognition outcomes.This paper focuses on systematically investigating the influences of three novel data representations in the Range-Doppler(RD)domain.Initially,the Radar Cross Section(RCS)and micro-Doppler(m-D)characteristics of a cone-shaped ballistic target are analyzed.Then,three different data representations are proposed:RD data,RD sequence tensor data,and RD trajectory data.To accommodate various data inputs,deep-learning models are designed,including a two-Dimensional Residual Dense Network(2D RDN),a three-Dimensional Residual Dense Network-Gated Recurrent Unit(3D RDN-GRU),and a Dynamic Trajectory Recognition Network(DTRN).Finally,an Electromagnetic(EM)computation dataset is collected to verify the performances of the networks.A broad range of experimental results demonstrates the effectiveness of the proposed framework.Moreover,several key parameters of the proposed networks and datasets are extensively studied in this research.展开更多
The micro-Doppler effect of moving targets may suffer from aliasing and Doppler migration due to the translational motion, which affects the application in real-time target identification. A new compensating method fo...The micro-Doppler effect of moving targets may suffer from aliasing and Doppler migration due to the translational motion, which affects the application in real-time target identification. A new compensating method for the rotationally symmetric target is proposed and demonstrated. By utilizing the micro-Doppler symmetry cancellation effect, the method can accurately compensate the translation effect, meanwhile, it behaves well in noise restraint. Computer simulations verify the high accuracy and efficiency of this method.展开更多
A reconfigurable metasurface based on optical control provides a control paradigm for integrating multiple functions at the same aperture,which effectively expands the freedom of control.However,the traditional light ...A reconfigurable metasurface based on optical control provides a control paradigm for integrating multiple functions at the same aperture,which effectively expands the freedom of control.However,the traditional light control method requires the light source to directly illuminate the photosensitive device,which forces the metasurface to be placed only according to the light emitter position,and even to need to be integrated on the light emitter,limiting the application scenarios of light-controlled reconfigurable metasurfaces.In this work,a light control method based on optical fiber is proposed,which guides and controls the light propagation path through optical fiber.The metasurface can be flexibly deployed,breaking through the limitation of physical space.As a verification,photoresistors are embedded in the metasurface,and the active device is directly excited by the light source as a driving signal to realize the switching of a polarization conversion function.The experimental results show that the optical-fiber-controlled metasurface can achieve linear-to-linear polarization conversion in the light environment and linear-to-circular polarization conversion in the dark environment.This work paves a new way,to our knowledge,to achieve a light-controlled metasurface,which enriches the family of intelligent metasurfaces and has great potential in many fields.展开更多
Programmable digital coding metasurfaces(PDCMs)can manipulate electromagnetic waves with high degrees of freedom,significantly enriching metasurface designs.However,most PDCMs are limited to the control of a single po...Programmable digital coding metasurfaces(PDCMs)can manipulate electromagnetic waves with high degrees of freedom,significantly enriching metasurface designs.However,most PDCMs are limited to the control of a single polarization,which cannot meet the requirements of the high integration of intelligent components.To further improve the practicability and flexibility of metasurfaces,we propose an integrated paradigm for spin-decoupling PDCMs based on light emitting diode arrays that fully embed the photoresistor as a part of the meta-atom to independently manipulate the wavefront in different polarizations.As a proof of concept,PDCMs were simulated,fabricated,and measured to verify the feasibility and effectiveness of the proposed method.The functions of scattering and vortices are verified at different polarizations,demonstrating that the metasurface can tailor the EM functions in six channels.This study can improve the integration of intelligent control metasurfaces and lay a solid foundation for their development.展开更多
Coding metasurfaces can manipulate electromagnetic wave in real time with high degree of freedom,the fascinating properties of which enrich the metasurface design with a wide range of application prospects.However,mos...Coding metasurfaces can manipulate electromagnetic wave in real time with high degree of freedom,the fascinating properties of which enrich the metasurface design with a wide range of application prospects.However,most of the coding metasurfaces are designed based on external excitation framework with the wired electrical or wireless light control devices,thus inevitably causing the interference with electromagnetic wave transmission and increasing the complexity of the metasurface design.In this work,a simplistic framework of single-pixel-programmable metasurfaces integrated with a capsuled LED array is proposed to dynamically control electromagnetic wave.The framework fully embeds the photoresistor in the meta-atom,controlling the LED array to directly illuminate the photoresistor to modulate the phase response.With this manner,the complex biasing network is transformed to the universal LED array,which means the physical control framework can be transformed to a software framework,and thus the functions of the metasurface can be freely manipulated by encoding the capsuled LED array avoiding mutual coupling of adjacent meta-atoms in real time.All the results verify that the far-field scattering pattern can be customized with this singlepixel-programmable metasurface.Encouragingly,this work provides a universal framework for coding metasurface design,which lays the foundation for metasurface intelligent perception and adaptive modulation.展开更多
基金co-supported by the National Natural Science Foundation of China(No.62001507)the Youth Talent Lifting Project of the China Association for Science and Technology(No.2021-JCJQ-QT-018)+1 种基金the Program of the Youth Innovation Team of Shaanxi Universitiesthe Natural Science Basic Research Plan in Shaanxi Province of China(No.2023-JC-YB-491)。
文摘Deep Learning(DL)model has been widely used in the field of Synthetic Aperture Radar Automatic Target Recognition(SAR-ATR)and has achieved excellent performance.However,the black-box nature of DL models has been the focus of criticism,especially in the application of SARATR,which is closely associated with the national defense and security domain.To address these issues,a new interpretable recognition model Physics-Guided BagNet(PGBN)is proposed in this article.The model adopts an interpretable convolutional neural network framework and uses time–frequency analysis to extract physical scattering features in SAR images.Based on the physical scattering features,an unsupervised segmentation method is proposed to distinguish targets from the background in SAR images.On the basis of the segmentation result,a structure is designed,which constrains the model's spatial attention to focus more on the targets themselves rather than the background,thereby making the model's decision-making more in line with physical principles.In contrast to previous interpretable research methods,this model combines interpretable structure with physical interpretability,further reducing the model's risk of error recognition.Experiments on the MSTAR dataset verify that the PGBN model exhibits excellent interpretability and recognition performance,and comparative experiments with heatmaps indicate that the physical feature guidance module presented in this article can constrain the model to focus more on the target itself rather than the background.
基金supported by the Natural Science Basic Research Plan in Shaanxi Province of China(No.2023-JCYB-491).
文摘Target recognition is a significant part of a Ballistic Missile Defense System(BMDS).However,most existing ballistic target recognition methods overlook the impact of data representation on recognition outcomes.This paper focuses on systematically investigating the influences of three novel data representations in the Range-Doppler(RD)domain.Initially,the Radar Cross Section(RCS)and micro-Doppler(m-D)characteristics of a cone-shaped ballistic target are analyzed.Then,three different data representations are proposed:RD data,RD sequence tensor data,and RD trajectory data.To accommodate various data inputs,deep-learning models are designed,including a two-Dimensional Residual Dense Network(2D RDN),a three-Dimensional Residual Dense Network-Gated Recurrent Unit(3D RDN-GRU),and a Dynamic Trajectory Recognition Network(DTRN).Finally,an Electromagnetic(EM)computation dataset is collected to verify the performances of the networks.A broad range of experimental results demonstrates the effectiveness of the proposed framework.Moreover,several key parameters of the proposed networks and datasets are extensively studied in this research.
文摘The micro-Doppler effect of moving targets may suffer from aliasing and Doppler migration due to the translational motion, which affects the application in real-time target identification. A new compensating method for the rotationally symmetric target is proposed and demonstrated. By utilizing the micro-Doppler symmetry cancellation effect, the method can accurately compensate the translation effect, meanwhile, it behaves well in noise restraint. Computer simulations verify the high accuracy and efficiency of this method.
基金National Key Research and Development Program of China(2022YFB3806200)National Natural Science Foundation of China(62201609,62401614,62401617)。
文摘A reconfigurable metasurface based on optical control provides a control paradigm for integrating multiple functions at the same aperture,which effectively expands the freedom of control.However,the traditional light control method requires the light source to directly illuminate the photosensitive device,which forces the metasurface to be placed only according to the light emitter position,and even to need to be integrated on the light emitter,limiting the application scenarios of light-controlled reconfigurable metasurfaces.In this work,a light control method based on optical fiber is proposed,which guides and controls the light propagation path through optical fiber.The metasurface can be flexibly deployed,breaking through the limitation of physical space.As a verification,photoresistors are embedded in the metasurface,and the active device is directly excited by the light source as a driving signal to realize the switching of a polarization conversion function.The experimental results show that the optical-fiber-controlled metasurface can achieve linear-to-linear polarization conversion in the light environment and linear-to-circular polarization conversion in the dark environment.This work paves a new way,to our knowledge,to achieve a light-controlled metasurface,which enriches the family of intelligent metasurfaces and has great potential in many fields.
基金supported in part by the National Key Research and Development Program of China under Grant 2022YFB3806200the National Natural Science Foundation of China under Grants 62101588 and 62201609.
文摘Programmable digital coding metasurfaces(PDCMs)can manipulate electromagnetic waves with high degrees of freedom,significantly enriching metasurface designs.However,most PDCMs are limited to the control of a single polarization,which cannot meet the requirements of the high integration of intelligent components.To further improve the practicability and flexibility of metasurfaces,we propose an integrated paradigm for spin-decoupling PDCMs based on light emitting diode arrays that fully embed the photoresistor as a part of the meta-atom to independently manipulate the wavefront in different polarizations.As a proof of concept,PDCMs were simulated,fabricated,and measured to verify the feasibility and effectiveness of the proposed method.The functions of scattering and vortices are verified at different polarizations,demonstrating that the metasurface can tailor the EM functions in six channels.This study can improve the integration of intelligent control metasurfaces and lay a solid foundation for their development.
基金National Key Research and Development Program of China(2022YFB3806200)National Natural Science Foundation of China(12004437,61971435,62101588,62201609,62001504)Natural Science Foundation of Shaanxi Province(2022JM-352,2022JQ-630)。
文摘Coding metasurfaces can manipulate electromagnetic wave in real time with high degree of freedom,the fascinating properties of which enrich the metasurface design with a wide range of application prospects.However,most of the coding metasurfaces are designed based on external excitation framework with the wired electrical or wireless light control devices,thus inevitably causing the interference with electromagnetic wave transmission and increasing the complexity of the metasurface design.In this work,a simplistic framework of single-pixel-programmable metasurfaces integrated with a capsuled LED array is proposed to dynamically control electromagnetic wave.The framework fully embeds the photoresistor in the meta-atom,controlling the LED array to directly illuminate the photoresistor to modulate the phase response.With this manner,the complex biasing network is transformed to the universal LED array,which means the physical control framework can be transformed to a software framework,and thus the functions of the metasurface can be freely manipulated by encoding the capsuled LED array avoiding mutual coupling of adjacent meta-atoms in real time.All the results verify that the far-field scattering pattern can be customized with this singlepixel-programmable metasurface.Encouragingly,this work provides a universal framework for coding metasurface design,which lays the foundation for metasurface intelligent perception and adaptive modulation.
