Pub Date : 2025-11-18DOI: 10.1109/TRS.2025.3634278
Samuel R. Shebert;Benjamin H. Kirk;R. Michael Buehrer
In congested or contested spectrum, radar is costly to operate due to high power usage, low spectral efficiency, and low covertness compared to passive sensors. For this reason, this work proposes a multi-mode radar sensing strategy, in which the sensors can choose between a monostatic radar mode and a passive electronic support measure (ESM) spectrum sensing mode. In ESM mode, a target is localized with a network of multi-mode sensors, which creates opportunities to reduce radar measurements. Radar and ESM measurements are rigorously compared using the Cramér–Rao bound to quantify the localization error of each mode. The best mode for each sensor is chosen using a restless contextual combinatorial multi-armed bandit (RCC-MAB) online learning algorithm. The RCC-MAB increases the flexibility of the network by adapting to the target in real-time based on recent radar and ESM measurements. Two variants, the $epsilon $ -Greedy and covert RCC-MABs, were created to fulfill different tracking objectives. The $epsilon $ -Greedy RCC-MAB variant seeks to minimize the tracking error by selecting the best sensing modes based on the quality of previous measurements and the current context of the tracking filter. The covert RCC-MAB variant significantly reduces radar usage to stay covert or minimize access to a shared spectrum by only exploring radar measurements when the tracking error approaches a predefined maximum error. The $epsilon $ -Greedy RCC-MAB consistently achieved the lowest tracking error of the tested mode controllers, 58% and 16% lower than a single-mode radar and ESM network, respectively, when the radio emissions of opportunity (REO) were available during 50% of measurement opportunities. In the same scenario, the covert RCC-MAB had 55% lower tracking error than single-mode radar, while using 82% less radar than the $epsilon $ -Greedy RCC-MAB.
{"title":"Multi-Mode Radar Network Control With Restless Contextual Combinatorial Multi-Armed Bandits","authors":"Samuel R. Shebert;Benjamin H. Kirk;R. Michael Buehrer","doi":"10.1109/TRS.2025.3634278","DOIUrl":"https://doi.org/10.1109/TRS.2025.3634278","url":null,"abstract":"In congested or contested spectrum, radar is costly to operate due to high power usage, low spectral efficiency, and low covertness compared to passive sensors. For this reason, this work proposes a multi-mode radar sensing strategy, in which the sensors can choose between a monostatic radar mode and a passive electronic support measure (ESM) spectrum sensing mode. In ESM mode, a target is localized with a network of multi-mode sensors, which creates opportunities to reduce radar measurements. Radar and ESM measurements are rigorously compared using the Cramér–Rao bound to quantify the localization error of each mode. The best mode for each sensor is chosen using a restless contextual combinatorial multi-armed bandit (RCC-MAB) online learning algorithm. The RCC-MAB increases the flexibility of the network by adapting to the target in real-time based on recent radar and ESM measurements. Two variants, the <inline-formula> <tex-math>$epsilon $ </tex-math></inline-formula>-Greedy and covert RCC-MABs, were created to fulfill different tracking objectives. The <inline-formula> <tex-math>$epsilon $ </tex-math></inline-formula>-Greedy RCC-MAB variant seeks to minimize the tracking error by selecting the best sensing modes based on the quality of previous measurements and the current context of the tracking filter. The covert RCC-MAB variant significantly reduces radar usage to stay covert or minimize access to a shared spectrum by only exploring radar measurements when the tracking error approaches a predefined maximum error. The <inline-formula> <tex-math>$epsilon $ </tex-math></inline-formula>-Greedy RCC-MAB consistently achieved the lowest tracking error of the tested mode controllers, 58% and 16% lower than a single-mode radar and ESM network, respectively, when the radio emissions of opportunity (REO) were available during 50% of measurement opportunities. In the same scenario, the covert RCC-MAB had 55% lower tracking error than single-mode radar, while using 82% less radar than the <inline-formula> <tex-math>$epsilon $ </tex-math></inline-formula>-Greedy RCC-MAB.","PeriodicalId":100645,"journal":{"name":"IEEE Transactions on Radar Systems","volume":"4 ","pages":"113-128"},"PeriodicalIF":0.0,"publicationDate":"2025-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145778341","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Oil spills can cause serious pollution to the marine environment. Synthetic aperture radar (SAR), as an all-day-all-weather active microwave sensor, can provide a powerful solution for oil spill detection. However, due to the limitations of system characteristics, the data–information imbalance problem exists in research based on polarimetric SAR. To address the above problems, a polarization feature generative mapping model (PF-GMM) for oil spill detection tasks is proposed in this article. PF-GMM maps dual-polarization features (DPFs) to qual-polarization features (QPFs) through a generative adversarial approach. To select DPFs that can cover qual-polarization information, an interpretable analysis module (IAM) was designed. The IAM analyzed the feature contribution and the interaction between DPFs and QPFs to reveal the significance of each DPF in model optimization and physical level, so as to achieve the optimal selection of DPFs. Based on the selected domain feature group (DFG), a dual-pol-SAR oil spill dataset (DPSOS) was constructed to evaluate the performance of the selected features and make up for the lack of a SAR oil spill detection dataset. Experimental results show that the DFG can effectively achieve oil spill segmentation in different scenarios, and, to a certain extent, achieve oil spill detection performance similar to that of QPFs. PF-GMM simultaneously ensures the superiority of the selected features in engineering applications and physical meanings. Its results can cover qual-polarization information to the greatest extent, make up for the defects and deficiencies of dual-polarization data, and have strong engineering guidance value.
{"title":"From Dual to Qual: A Feature-Analysis-Oriented Interpretable Polarization Feature Generative Mapping Model for SAR Oil Spill Detection","authors":"Lingxiao Cheng;Ying Li;Bingxin Liu;Yuanheng Sun;Weimin Huang","doi":"10.1109/TRS.2025.3633309","DOIUrl":"https://doi.org/10.1109/TRS.2025.3633309","url":null,"abstract":"Oil spills can cause serious pollution to the marine environment. Synthetic aperture radar (SAR), as an all-day-all-weather active microwave sensor, can provide a powerful solution for oil spill detection. However, due to the limitations of system characteristics, the data–information imbalance problem exists in research based on polarimetric SAR. To address the above problems, a polarization feature generative mapping model (PF-GMM) for oil spill detection tasks is proposed in this article. PF-GMM maps dual-polarization features (DPFs) to qual-polarization features (QPFs) through a generative adversarial approach. To select DPFs that can cover qual-polarization information, an interpretable analysis module (IAM) was designed. The IAM analyzed the feature contribution and the interaction between DPFs and QPFs to reveal the significance of each DPF in model optimization and physical level, so as to achieve the optimal selection of DPFs. Based on the selected domain feature group (DFG), a dual-pol-SAR oil spill dataset (DPSOS) was constructed to evaluate the performance of the selected features and make up for the lack of a SAR oil spill detection dataset. Experimental results show that the DFG can effectively achieve oil spill segmentation in different scenarios, and, to a certain extent, achieve oil spill detection performance similar to that of QPFs. PF-GMM simultaneously ensures the superiority of the selected features in engineering applications and physical meanings. Its results can cover qual-polarization information to the greatest extent, make up for the defects and deficiencies of dual-polarization data, and have strong engineering guidance value.","PeriodicalId":100645,"journal":{"name":"IEEE Transactions on Radar Systems","volume":"4 ","pages":"1-14"},"PeriodicalIF":0.0,"publicationDate":"2025-11-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145729505","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Deep learning-based synthetic aperture radar (SAR) ship detection methods have emerged as the leading techniques due to their strong feature extraction and generalization capabilities across various scenes and conditions. However, they still face challenges in distinguishing ships from complex backgrounds, especially in cases involving small or offshore vessels, dense inshore regions, or ships with textures and grayscale similar to their surroundings. To address these challenges, this article introduces CSFF-MGDH, a novel SAR ship detector that integrates adaptive feature learning and a mutually guided decoupled head (MGDH) into the YOLOX framework. First, deformable convolution (DCN) is incorporated into the backbone to overcome the limitations of standard square convolution in handling large ship deformations caused by severe noise in remote sensing images. Second, a cross-stage feature fusion module (CSFFM) is introduced to fuse features from adjacent layers, mitigating receptive field discrepancies in multilayer feature maps caused by DCN and reducing noise through local self-supervised interaction. Finally, a MGDH is designed to guide the regression branch using classification features, improving single-category object detection. Extensive experiments on the SAR ship detection dataset (SSDD) and HRSID dataset demonstrate that the proposed method substantially outperforms the baseline methods in detection accuracy.
{"title":"CSFF-MGDH: Cross-Stage Feature Fusion and Decoupled Head With Mutual Guidance for SAR Ship Detection","authors":"Yixin Qiao;Xiaoxiao Yin;Xinyuan Zhou;Shiyong Lan;Wenwu Wang;Haohan Chen","doi":"10.1109/TRS.2025.3632813","DOIUrl":"https://doi.org/10.1109/TRS.2025.3632813","url":null,"abstract":"Deep learning-based synthetic aperture radar (SAR) ship detection methods have emerged as the leading techniques due to their strong feature extraction and generalization capabilities across various scenes and conditions. However, they still face challenges in distinguishing ships from complex backgrounds, especially in cases involving small or offshore vessels, dense inshore regions, or ships with textures and grayscale similar to their surroundings. To address these challenges, this article introduces CSFF-MGDH, a novel SAR ship detector that integrates adaptive feature learning and a mutually guided decoupled head (MGDH) into the YOLOX framework. First, deformable convolution (DCN) is incorporated into the backbone to overcome the limitations of standard square convolution in handling large ship deformations caused by severe noise in remote sensing images. Second, a cross-stage feature fusion module (CSFFM) is introduced to fuse features from adjacent layers, mitigating receptive field discrepancies in multilayer feature maps caused by DCN and reducing noise through local self-supervised interaction. Finally, a MGDH is designed to guide the regression branch using classification features, improving single-category object detection. Extensive experiments on the SAR ship detection dataset (SSDD) and HRSID dataset demonstrate that the proposed method substantially outperforms the baseline methods in detection accuracy.","PeriodicalId":100645,"journal":{"name":"IEEE Transactions on Radar Systems","volume":"4 ","pages":"76-85"},"PeriodicalIF":0.0,"publicationDate":"2025-11-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145778343","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-12DOI: 10.1109/TRS.2025.3630966
Gourab Ghatak
We introduce a novel metric for stochastic geometry-based analysis of automotive radar networks called target tracking probability. Unlike the well-investigated detection probability (often termed the success or coverage probability in stochastic geometry), the tracking probability characterizes the event of successive successful target detection with a sequence of radar pulses. From a theoretical standpoint, this work adds to the rich repertoire of statistical metrics in stochastic geometry-based wireless network analysis. To optimize the target tracking probability in high interference scenarios, we study a block medium access control (MAC) protocol for the automotive radars to share a common channel and recommend the optimal MAC parameter for a given vehicle and street density. Importantly, we show that the optimal MAC parameter that maximizes the detection probability may not be the one that maximizes the tracking probability. Our research reveals how the tracking event can be naturally mapped to the quality-of-service (QoS) requirements of latency and reliability for different vehicular technology use cases. This can enable use-case-specific adaptive selection of radar parameters for optimal target tracking.
{"title":"Statistics of Successive Successful Target Detection in Automotive Radar Networks","authors":"Gourab Ghatak","doi":"10.1109/TRS.2025.3630966","DOIUrl":"https://doi.org/10.1109/TRS.2025.3630966","url":null,"abstract":"We introduce a novel metric for stochastic geometry-based analysis of automotive radar networks called target tracking probability. Unlike the well-investigated detection probability (often termed the success or coverage probability in stochastic geometry), the tracking probability characterizes the event of successive successful target detection with a sequence of radar pulses. From a theoretical standpoint, this work adds to the rich repertoire of statistical metrics in stochastic geometry-based wireless network analysis. To optimize the target tracking probability in high interference scenarios, we study a block medium access control (MAC) protocol for the automotive radars to share a common channel and recommend the optimal MAC parameter for a given vehicle and street density. Importantly, we show that the optimal MAC parameter that maximizes the detection probability may not be the one that maximizes the tracking probability. Our research reveals how the tracking event can be naturally mapped to the quality-of-service (QoS) requirements of latency and reliability for different vehicular technology use cases. This can enable use-case-specific adaptive selection of radar parameters for optimal target tracking.","PeriodicalId":100645,"journal":{"name":"IEEE Transactions on Radar Systems","volume":"3 ","pages":"1449-1462"},"PeriodicalIF":0.0,"publicationDate":"2025-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145560634","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-11DOI: 10.1109/TRS.2025.3631021
Tao Zhang;Nishang Xie;Sinong Quan;Wei Wang;Feiming Wei;Wenxian Yu
In the past few years, polarimetric synthetic aperture radar (PolSAR) as an advanced technology has been widely exploited to Earth observation, among which ship detection is an active research topic. Taking the sub-look decomposition technology as the basis, this article proposes a new ship detection method, abbreviated to amplitude-based ship detection metric (ASM). In brief, two single-look complex (SLC) images <inline-formula> <tex-math>$I_{1}$ </tex-math></inline-formula> and <inline-formula> <tex-math>$I_{2}$ </tex-math></inline-formula> are first obtained from the original PolSAR image <inline-formula> <tex-math>$O$ </tex-math></inline-formula> for forming the data group {<inline-formula> <tex-math>$I_{1}$ </tex-math></inline-formula>, <inline-formula> <tex-math>$O$ </tex-math></inline-formula>, and <inline-formula> <tex-math>$I_{2}$ </tex-math></inline-formula>}. Then, the <inline-formula> <tex-math>$H/A/alpha $ </tex-math></inline-formula> decomposition is performed on {<inline-formula> <tex-math>$I_{1}$ </tex-math></inline-formula>, <inline-formula> <tex-math>$O$ </tex-math></inline-formula>, and <inline-formula> <tex-math>$I_{2}$ </tex-math></inline-formula>} so as to yield the <inline-formula> <tex-math>$H/alpha $ </tex-math></inline-formula> plane group {<inline-formula> <tex-math>$P_{1}$ </tex-math></inline-formula>, <inline-formula> <tex-math>$P_{0}$ </tex-math></inline-formula>, and <inline-formula> <tex-math>$P_{2}$ </tex-math></inline-formula>}, which is subsequently used to suppress sea clutter and generate another filtered data group {<inline-formula> <tex-math>$F_{1}$ </tex-math></inline-formula>, <inline-formula> <tex-math>$F_{0}$ </tex-math></inline-formula>, and <inline-formula> <tex-math>$F_{2}$ </tex-math></inline-formula>} that, respectively, corresponds to <inline-formula> <tex-math>$I_{1}$ </tex-math></inline-formula>, <inline-formula> <tex-math>$O$ </tex-math></inline-formula>, and <inline-formula> <tex-math>$I_{2}$ </tex-math></inline-formula>. Thereafter, a new <inline-formula> <tex-math>$3 times 3$ </tex-math></inline-formula> spatial–spectral coherence difference matrix [<inline-formula> <tex-math>$ST$ </tex-math></inline-formula>] is further constructed by {<inline-formula> <tex-math>$F_{1}$ </tex-math></inline-formula>, <inline-formula> <tex-math>$F_{0}$ </tex-math></inline-formula>, and <inline-formula> <tex-math>$F_{2}$ </tex-math></inline-formula>}, wherein the spatial information and spectral information are simultaneously used. Therefore, [<inline-formula> <tex-math>$ST$ </tex-math></inline-formula>] can effectively highlight ships from sea clutter. To verify this point, an ASM is finally built by multiplying the amplitude values of the terms <inline-formula> <tex-math>$text {ST}_{13}$ </tex-math></inline-formula>, <inline-formula> <tex-math>$text {ST}_{23}$ </tex-math></inline-formula>, and <inline-formula> <tex-math>$text {ST}_{33}$ </tex-math></inline-formula> together. Extensive experiments demonstrate
{"title":"Polarimeric SAR Ship Detection Based on Sub-Look the Decomposition Technology","authors":"Tao Zhang;Nishang Xie;Sinong Quan;Wei Wang;Feiming Wei;Wenxian Yu","doi":"10.1109/TRS.2025.3631021","DOIUrl":"https://doi.org/10.1109/TRS.2025.3631021","url":null,"abstract":"In the past few years, polarimetric synthetic aperture radar (PolSAR) as an advanced technology has been widely exploited to Earth observation, among which ship detection is an active research topic. Taking the sub-look decomposition technology as the basis, this article proposes a new ship detection method, abbreviated to amplitude-based ship detection metric (ASM). In brief, two single-look complex (SLC) images <inline-formula> <tex-math>$I_{1}$ </tex-math></inline-formula> and <inline-formula> <tex-math>$I_{2}$ </tex-math></inline-formula> are first obtained from the original PolSAR image <inline-formula> <tex-math>$O$ </tex-math></inline-formula> for forming the data group {<inline-formula> <tex-math>$I_{1}$ </tex-math></inline-formula>, <inline-formula> <tex-math>$O$ </tex-math></inline-formula>, and <inline-formula> <tex-math>$I_{2}$ </tex-math></inline-formula>}. Then, the <inline-formula> <tex-math>$H/A/alpha $ </tex-math></inline-formula> decomposition is performed on {<inline-formula> <tex-math>$I_{1}$ </tex-math></inline-formula>, <inline-formula> <tex-math>$O$ </tex-math></inline-formula>, and <inline-formula> <tex-math>$I_{2}$ </tex-math></inline-formula>} so as to yield the <inline-formula> <tex-math>$H/alpha $ </tex-math></inline-formula> plane group {<inline-formula> <tex-math>$P_{1}$ </tex-math></inline-formula>, <inline-formula> <tex-math>$P_{0}$ </tex-math></inline-formula>, and <inline-formula> <tex-math>$P_{2}$ </tex-math></inline-formula>}, which is subsequently used to suppress sea clutter and generate another filtered data group {<inline-formula> <tex-math>$F_{1}$ </tex-math></inline-formula>, <inline-formula> <tex-math>$F_{0}$ </tex-math></inline-formula>, and <inline-formula> <tex-math>$F_{2}$ </tex-math></inline-formula>} that, respectively, corresponds to <inline-formula> <tex-math>$I_{1}$ </tex-math></inline-formula>, <inline-formula> <tex-math>$O$ </tex-math></inline-formula>, and <inline-formula> <tex-math>$I_{2}$ </tex-math></inline-formula>. Thereafter, a new <inline-formula> <tex-math>$3 times 3$ </tex-math></inline-formula> spatial–spectral coherence difference matrix [<inline-formula> <tex-math>$ST$ </tex-math></inline-formula>] is further constructed by {<inline-formula> <tex-math>$F_{1}$ </tex-math></inline-formula>, <inline-formula> <tex-math>$F_{0}$ </tex-math></inline-formula>, and <inline-formula> <tex-math>$F_{2}$ </tex-math></inline-formula>}, wherein the spatial information and spectral information are simultaneously used. Therefore, [<inline-formula> <tex-math>$ST$ </tex-math></inline-formula>] can effectively highlight ships from sea clutter. To verify this point, an ASM is finally built by multiplying the amplitude values of the terms <inline-formula> <tex-math>$text {ST}_{13}$ </tex-math></inline-formula>, <inline-formula> <tex-math>$text {ST}_{23}$ </tex-math></inline-formula>, and <inline-formula> <tex-math>$text {ST}_{33}$ </tex-math></inline-formula> together. Extensive experiments demonstrate ","PeriodicalId":100645,"journal":{"name":"IEEE Transactions on Radar Systems","volume":"4 ","pages":"35-49"},"PeriodicalIF":0.0,"publicationDate":"2025-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145729504","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-03DOI: 10.1109/TRS.2025.3628294
Kuiyu Chen;Chen Liu;Yunchao Song;Lingzhi Zhu
Human activity recognition (HAR) has emerged as a key technology, with applications ranging from security to healthcare. Radar-based HAR, which leverages micro-Doppler signatures, offers strong performance in complex environments. However, most existing systems operate under closed-set assumptions, recognizing only predefined activities. This restricts their effectiveness in real-world scenarios where novel or unseen activities may occur. To address this challenge, this work proposes a virtual prototype learning (VPL) framework for open-set HAR. Inspired by human memory and pattern-matching processes, VPL encodes micro-Doppler spectrograms into abstract representations and compares them with learned prototypes in the metric space. The framework is guided by a combination of Euclidean cross-entropy loss and clustering loss to promote clear separation between different activity classes while preserving consistency within each class. To further improve robustness, VPL incorporates a manifold mixup strategy, generating pseudo-samples that help sharpen the boundary between known and unknown activities. A buffer zone is established in the feature space to reinforce this separation, and hyperspherical decision boundaries are employed to enhance classification accuracy. Experiments with real-world radar data show that VPL outperforms existing methods, achieving higher accuracy for known activities while effectively detecting unknown activities.
{"title":"Open-Set Human Activity Recognition With Micro-Doppler Signatures and Virtual Prototype Learning","authors":"Kuiyu Chen;Chen Liu;Yunchao Song;Lingzhi Zhu","doi":"10.1109/TRS.2025.3628294","DOIUrl":"https://doi.org/10.1109/TRS.2025.3628294","url":null,"abstract":"Human activity recognition (HAR) has emerged as a key technology, with applications ranging from security to healthcare. Radar-based HAR, which leverages micro-Doppler signatures, offers strong performance in complex environments. However, most existing systems operate under closed-set assumptions, recognizing only predefined activities. This restricts their effectiveness in real-world scenarios where novel or unseen activities may occur. To address this challenge, this work proposes a virtual prototype learning (VPL) framework for open-set HAR. Inspired by human memory and pattern-matching processes, VPL encodes micro-Doppler spectrograms into abstract representations and compares them with learned prototypes in the metric space. The framework is guided by a combination of Euclidean cross-entropy loss and clustering loss to promote clear separation between different activity classes while preserving consistency within each class. To further improve robustness, VPL incorporates a manifold mixup strategy, generating pseudo-samples that help sharpen the boundary between known and unknown activities. A buffer zone is established in the feature space to reinforce this separation, and hyperspherical decision boundaries are employed to enhance classification accuracy. Experiments with real-world radar data show that VPL outperforms existing methods, achieving higher accuracy for known activities while effectively detecting unknown activities.","PeriodicalId":100645,"journal":{"name":"IEEE Transactions on Radar Systems","volume":"3 ","pages":"1463-1473"},"PeriodicalIF":0.0,"publicationDate":"2025-11-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145560632","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Polarimetry, which is the ability to measure the scattering response of the environment across orthogonal polarizations, is fundamental to enhancing wireless communication and radar system performance. In this article, we use the Zak-OTFS modulation to enable instantaneous polarimetry within a single transmission frame. We transmit a Zak-OTFS carrier waveform and a spread carrier waveform mutually unbiased to it simultaneously over orthogonal polarizations. The mutual unbiasedness of the two waveforms enables the receiver to estimate the full polarimetric response of the scattering environment from a single received frame. Unlike existing methods for instantaneous polarimetry with computational complexity quadratic in the time–bandwidth product, the proposed method enables instantaneous polarimetry at near-linear complexity in the time–bandwidth product. Via numerical simulations, we show ideal polarimetric target detection and parameter estimation results with the proposed method, with improvements in computational complexity and greater clutter resilience over comparable baselines.
{"title":"Instantaneous Polarimetry With Zak-OTFS","authors":"Nishant Mehrotra;Sandesh Rao Mattu;Robert Calderbank","doi":"10.1109/TRS.2025.3625812","DOIUrl":"https://doi.org/10.1109/TRS.2025.3625812","url":null,"abstract":"Polarimetry, which is the ability to measure the scattering response of the environment across orthogonal polarizations, is fundamental to enhancing wireless communication and radar system performance. In this article, we use the Zak-OTFS modulation to enable instantaneous polarimetry within a single transmission frame. We transmit a Zak-OTFS carrier waveform and a spread carrier waveform mutually unbiased to it simultaneously over orthogonal polarizations. The mutual unbiasedness of the two waveforms enables the receiver to estimate the full polarimetric response of the scattering environment from a single received frame. Unlike existing methods for instantaneous polarimetry with computational complexity quadratic in the time–bandwidth product, the proposed method enables instantaneous polarimetry at near-linear complexity in the time–bandwidth product. Via numerical simulations, we show ideal polarimetric target detection and parameter estimation results with the proposed method, with improvements in computational complexity and greater clutter resilience over comparable baselines.","PeriodicalId":100645,"journal":{"name":"IEEE Transactions on Radar Systems","volume":"3 ","pages":"1413-1420"},"PeriodicalIF":0.0,"publicationDate":"2025-10-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145510157","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-20DOI: 10.1109/TRS.2025.3623966
Zirui Chen;Yifei Ji;Yongsheng Zhang;Zhen Dong;Weijian Liu;Junqiang Song
The nonstationary spatiotemporal distribution of the ionosphere creates multiple irregular propagation paths between the target and transceivers of the skywave over-the-horizon radar (OTHR). The multipath effect fundamentally induces distortions of the target plot signatures in range and Doppler dimensions and thereby significantly degrades the target localization/velocimetry accuracy and detection performance. Building upon the full-link sea clutter model established in Part I, this article develops a comprehensive framework incorporating trans-ionospheric signal modeling, simulation, and impact analysis for multipath targets. First, a variable-step ray-tracing technique generally following the coarse-to-fine search mechanism is developed to identify all propagation paths illuminating targets within a wide radar beam. Second, full-link multipath signal models in the fast-slow-time and range–Doppler (RD) domains are established by integrating ionospheric effects with high-frequency (HF) radar cross section (RCS) of typical targets. Finally, a theoretical analysis of multipath effects on target plot is performed based on the RD model. Three types of typical modes, large-scale multipath, microscale multipath, and multihop multipath, are defined by propagation path characteristics. Their impacts are analyzed for aerial and maritime OTHR detection scenarios. Theoretical and simulation results quantitatively characterize the impact of multipath effects on target signatures, demonstrating that trans-ionospheric multipath effects provide critical information for parameter estimation enhancement. The proposed OTHR full-link model establishes a theoretical framework for understanding trans-ionospheric multipath effects and provides foundational support for enhancing localization/velocimetry accuracy, suppressing false target plots, resolving Doppler ambiguity, and improving detection performance.
{"title":"Skywave OTHR Full-Link Modeling and Simulation—Part II: Trans-Ionospheric Multipath Target Signal","authors":"Zirui Chen;Yifei Ji;Yongsheng Zhang;Zhen Dong;Weijian Liu;Junqiang Song","doi":"10.1109/TRS.2025.3623966","DOIUrl":"https://doi.org/10.1109/TRS.2025.3623966","url":null,"abstract":"The nonstationary spatiotemporal distribution of the ionosphere creates multiple irregular propagation paths between the target and transceivers of the skywave over-the-horizon radar (OTHR). The multipath effect fundamentally induces distortions of the target plot signatures in range and Doppler dimensions and thereby significantly degrades the target localization/velocimetry accuracy and detection performance. Building upon the full-link sea clutter model established in Part I, this article develops a comprehensive framework incorporating trans-ionospheric signal modeling, simulation, and impact analysis for multipath targets. First, a variable-step ray-tracing technique generally following the coarse-to-fine search mechanism is developed to identify all propagation paths illuminating targets within a wide radar beam. Second, full-link multipath signal models in the fast-slow-time and range–Doppler (RD) domains are established by integrating ionospheric effects with high-frequency (HF) radar cross section (RCS) of typical targets. Finally, a theoretical analysis of multipath effects on target plot is performed based on the RD model. Three types of typical modes, large-scale multipath, microscale multipath, and multihop multipath, are defined by propagation path characteristics. Their impacts are analyzed for aerial and maritime OTHR detection scenarios. Theoretical and simulation results quantitatively characterize the impact of multipath effects on target signatures, demonstrating that trans-ionospheric multipath effects provide critical information for parameter estimation enhancement. The proposed OTHR full-link model establishes a theoretical framework for understanding trans-ionospheric multipath effects and provides foundational support for enhancing localization/velocimetry accuracy, suppressing false target plots, resolving Doppler ambiguity, and improving detection performance.","PeriodicalId":100645,"journal":{"name":"IEEE Transactions on Radar Systems","volume":"3 ","pages":"1392-1412"},"PeriodicalIF":0.0,"publicationDate":"2025-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145455818","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Radar-based object detection (OD) is critical for detecting distant objects and ensuring privacy in challenging environments. Existing OD pipelines require extensive preprocessing and complex machine learning (ML) algorithms, which hinders edge deployment. Prior approaches address these challenges by processing raw radar data using an analog-to-digital converter (ADC) or fast Fourier transform (FFT)-based preprocessing. However, as sensing resolution increases, the volume of data generated at sensor nodes grows, leading to increased model complexity and computational overhead on edge systems. In this work, we introduce ChirpNet, a neural network designed for radar-based OD. ChirpNet processes raw ADC data from virtual antennas for each chirp, integrating sequential chirp-based radar sensing directly into the network. This design achieves a $43times $ reduction in model computations and a $5times $ reduction in latency while still maintaining competitive object detection performance. Additionally, the ChirpNet models demonstrate improved robustness in various clutter scenarios compared to prior ML-based detectors.
{"title":"Toward Efficient and Robust Sequential Chirp-Based Data-Driven Radar Processing for Object Detection","authors":"Sudarshan Sharma;Hemant Kumawat;Anuvab Sen;Jinhyeok Park;Saibal Mukhopadhyay","doi":"10.1109/TRS.2025.3622514","DOIUrl":"https://doi.org/10.1109/TRS.2025.3622514","url":null,"abstract":"Radar-based object detection (OD) is critical for detecting distant objects and ensuring privacy in challenging environments. Existing OD pipelines require extensive preprocessing and complex machine learning (ML) algorithms, which hinders edge deployment. Prior approaches address these challenges by processing raw radar data using an analog-to-digital converter (ADC) or fast Fourier transform (FFT)-based preprocessing. However, as sensing resolution increases, the volume of data generated at sensor nodes grows, leading to increased model complexity and computational overhead on edge systems. In this work, we introduce ChirpNet, a neural network designed for radar-based OD. ChirpNet processes raw ADC data from virtual antennas for each chirp, integrating sequential chirp-based radar sensing directly into the network. This design achieves a <inline-formula> <tex-math>$43times $ </tex-math></inline-formula> reduction in model computations and a <inline-formula> <tex-math>$5times $ </tex-math></inline-formula> reduction in latency while still maintaining competitive object detection performance. Additionally, the ChirpNet models demonstrate improved robustness in various clutter scenarios compared to prior ML-based detectors.","PeriodicalId":100645,"journal":{"name":"IEEE Transactions on Radar Systems","volume":"3 ","pages":"1435-1448"},"PeriodicalIF":0.0,"publicationDate":"2025-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145560633","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-16DOI: 10.1109/TRS.2025.3622484
Yihan Su;Lei Wang;Xinan Lu;Cenwei Liu;Yimin Liu
Modern radars face the threat of multiple mainlobe jammings, and the use of distributed radars for jamming suppression has received extensive attention. Most existing studies primarily focus on the narrowband or far-field jamming scenarios, where jamming signals are assumed to be time-aligned across radars. However, for wideband or large-scale distributed radar systems, the time-delay differences of jamming signals across different radar nodes become nonnegligible, leading to the failure of classical algorithms. Considering the jamming delay differences, this article proposes a multijamming suppression method based on reconstruction of the jamming signals, where an alternative iteration is adopted to integrate the jamming signal reconstruction and time-delay difference estimation. Appropriate initialization and waveform design enable the proposed algorithm to be effectively applied across different jamming types, including noise jamming and interrupted sampling repeater jamming (ISRJ). Both the simulation and measured data experiments validate the effectiveness of the proposed algorithm to suppress multiple jammings.
{"title":"Multiple Mainlobe Jamming Reconstruction and Suppression in Wideband Distributed Radars","authors":"Yihan Su;Lei Wang;Xinan Lu;Cenwei Liu;Yimin Liu","doi":"10.1109/TRS.2025.3622484","DOIUrl":"https://doi.org/10.1109/TRS.2025.3622484","url":null,"abstract":"Modern radars face the threat of multiple mainlobe jammings, and the use of distributed radars for jamming suppression has received extensive attention. Most existing studies primarily focus on the narrowband or far-field jamming scenarios, where jamming signals are assumed to be time-aligned across radars. However, for wideband or large-scale distributed radar systems, the time-delay differences of jamming signals across different radar nodes become nonnegligible, leading to the failure of classical algorithms. Considering the jamming delay differences, this article proposes a multijamming suppression method based on reconstruction of the jamming signals, where an alternative iteration is adopted to integrate the jamming signal reconstruction and time-delay difference estimation. Appropriate initialization and waveform design enable the proposed algorithm to be effectively applied across different jamming types, including noise jamming and interrupted sampling repeater jamming (ISRJ). Both the simulation and measured data experiments validate the effectiveness of the proposed algorithm to suppress multiple jammings.","PeriodicalId":100645,"journal":{"name":"IEEE Transactions on Radar Systems","volume":"3 ","pages":"1362-1374"},"PeriodicalIF":0.0,"publicationDate":"2025-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145405286","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：