The evolution beyond 5G and the advent of new 6G applications pose increasing challenges for communication in dense urban and hotspot areas. Photonics-assisted millimeter-wave (mm-Wave) communication systems have emerged as attractive solutions to meet these demands. Various enhancement schemes based on traditional digital signal processing (DSP) and neural networks have been proposed. However, the limitations of traditional DSP algorithms and the high complexity of supervised neural networks hinder real-time transmission. To address this issue, this article proposes an unsupervised DNN-based equalizer built upon the traditional constant modulus algorithm (CMA) for blind equalization of PAM signals, with structured pruning applied to facilitate practical deployment. Compared to the conventional CMA, the blind DNN equalizer effectively handles both linear and nonlinear impairments and achieves better bit error rate (BER) performance. Structured pruning reduces model floating-point operations (FLOPs) by 34%, successfully transforming the originally nondeployable architecture into a compact model executable in real-time on a field programmable gate array (FPGA), at the cost of a slight BER increase from $3.53times 10^{-3}$ to $6.84times 10^{-3}$ . We demonstrate a real-time photonics-assisted W-band wireless system based on an FPGA, employing the blind DNN equalizer to enhance performance. The system achieves real-time transmission of 14.7456-Gb/s PAM4 signals over a 200-m free-space link, attaining a BER of $6.84times 10^{-3}$ . This work explores high-capacity real-time communication in dense AI environments and serves as a representative case of integrating traditional DSP algorithms with data-driven neural networks.
{"title":"Real-Time Deployment of Pruned Unsupervised DNN for Blind Equalization in a Photonics-Aided W-Band Wireless System","authors":"Jie Zhang;Wen Zhou;Qihang Wang;Sheng Hu;Sicong Xu;Chengzhen Bian;Jingtao Ge;Jingwen Lin;Siqi Wang;Zhihang Ou;Tengsheng Zhang;Tong Wang;Jianjun Yu","doi":"10.1109/TMTT.2025.3636545","DOIUrl":"https://doi.org/10.1109/TMTT.2025.3636545","url":null,"abstract":"The evolution beyond 5G and the advent of new 6G applications pose increasing challenges for communication in dense urban and hotspot areas. Photonics-assisted millimeter-wave (mm-Wave) communication systems have emerged as attractive solutions to meet these demands. Various enhancement schemes based on traditional digital signal processing (DSP) and neural networks have been proposed. However, the limitations of traditional DSP algorithms and the high complexity of supervised neural networks hinder real-time transmission. To address this issue, this article proposes an unsupervised DNN-based equalizer built upon the traditional constant modulus algorithm (CMA) for blind equalization of PAM signals, with structured pruning applied to facilitate practical deployment. Compared to the conventional CMA, the blind DNN equalizer effectively handles both linear and nonlinear impairments and achieves better bit error rate (BER) performance. Structured pruning reduces model floating-point operations (FLOPs) by 34%, successfully transforming the originally nondeployable architecture into a compact model executable in real-time on a field programmable gate array (FPGA), at the cost of a slight BER increase from <inline-formula> <tex-math>$3.53times 10^{-3}$ </tex-math></inline-formula> to <inline-formula> <tex-math>$6.84times 10^{-3}$ </tex-math></inline-formula>. We demonstrate a real-time photonics-assisted W-band wireless system based on an FPGA, employing the blind DNN equalizer to enhance performance. The system achieves real-time transmission of 14.7456-Gb/s PAM4 signals over a 200-m free-space link, attaining a BER of <inline-formula> <tex-math>$6.84times 10^{-3}$ </tex-math></inline-formula>. This work explores high-capacity real-time communication in dense AI environments and serves as a representative case of integrating traditional DSP algorithms with data-driven neural networks.","PeriodicalId":13272,"journal":{"name":"IEEE Transactions on Microwave Theory and Techniques","volume":"74 1","pages":"1086-1097"},"PeriodicalIF":4.5,"publicationDate":"2025-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146049272","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","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/TMTT.2025.3622320
Dantong Liu;Zehao Zhang;Yifeng Wang;Yunxiao Zhao;Qizhi Wang;Yuanming Shi;Pingqiang Zhou;Xiong Wang
Cerebral disease has always been a major threat to human health, of which hemorrhagic stroke poses one of the greatest dangers. As a novel imaging modality, microwave-induced thermoacoustic tomography (MITAT) serves as a potential noninvasive, time and cost-effective technique to detect cerebral diseases. However, the traditional MITAT technique can only provide qualitative rather than quantitative information of tissues, which limits biomedical applications of MITAT. In this article, a deep-learning-enabled MITAT (DL-MITAT) brain imaging approach is presented to perform transcranial quantitative dual reconstruction of dielectric and acoustic properties of the brain tissues. We design a novel network architecture to extract the tissue properties and mitigate the acoustic inhomogeneity issue caused by the skull. With sufficient simulation and ex vivo experimental testing, we demonstrate that this method can effectively recover the quantitative dielectric constant, conductivity, and speed of sound (SOS) distributions of the applied brain models in a transcranial manner. Different cases are studied to test the generalization ability of the proposed approach. This is the first reported work that can simultaneously and quantitatively reconstruct both the dielectric and acoustic properties. This work provides a viable pathway for transcranial quantitative reconstruction of brain tissues’ dielectric properties and SOS, which is very meaningful for cerebral disease diagnosis. The proposed DL-MITAT technique holds the potential to alleviate the acoustic distortion issue due to the skull-induced acoustic inhomogeneity.
{"title":"Deep-Learning-Based Transcranial Quantitative Microwave-Induced Thermoacoustic Tomography for Dual Reconstruction of Dielectric and Acoustic Properties","authors":"Dantong Liu;Zehao Zhang;Yifeng Wang;Yunxiao Zhao;Qizhi Wang;Yuanming Shi;Pingqiang Zhou;Xiong Wang","doi":"10.1109/TMTT.2025.3622320","DOIUrl":"https://doi.org/10.1109/TMTT.2025.3622320","url":null,"abstract":"Cerebral disease has always been a major threat to human health, of which hemorrhagic stroke poses one of the greatest dangers. As a novel imaging modality, microwave-induced thermoacoustic tomography (MITAT) serves as a potential noninvasive, time and cost-effective technique to detect cerebral diseases. However, the traditional MITAT technique can only provide qualitative rather than quantitative information of tissues, which limits biomedical applications of MITAT. In this article, a deep-learning-enabled MITAT (DL-MITAT) brain imaging approach is presented to perform transcranial quantitative dual reconstruction of dielectric and acoustic properties of the brain tissues. We design a novel network architecture to extract the tissue properties and mitigate the acoustic inhomogeneity issue caused by the skull. With sufficient simulation and ex vivo experimental testing, we demonstrate that this method can effectively recover the quantitative dielectric constant, conductivity, and speed of sound (SOS) distributions of the applied brain models in a transcranial manner. Different cases are studied to test the generalization ability of the proposed approach. This is the first reported work that can simultaneously and quantitatively reconstruct both the dielectric and acoustic properties. This work provides a viable pathway for transcranial quantitative reconstruction of brain tissues’ dielectric properties and SOS, which is very meaningful for cerebral disease diagnosis. The proposed DL-MITAT technique holds the potential to alleviate the acoustic distortion issue due to the skull-induced acoustic inhomogeneity.","PeriodicalId":13272,"journal":{"name":"IEEE Transactions on Microwave Theory and Techniques","volume":"73 12","pages":"10632-10643"},"PeriodicalIF":4.5,"publicationDate":"2025-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145778254","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A W-band frequency-modulated continuous-wave (FMCW) radar, implemented in 65-nm CMOS, is proposed for intelligent transportation system (ITS) applications in this article. The system integrates four transmitters (TXs) and four receivers (RXs), along with a frequency synthesizer and a local oscillator (LO) distribution network. Both the low-noise amplifier (LNA) and the power amplifier (PA) adopt multistage cascaded topologies with magnetically coupled resonators (MCRs) to enable broadband operation. An ultrawideband class-B mixer, implemented with only an active switching core, supports continuous operation from 20 to 110 GHz. Furthermore, an LO distribution network featuring three cascaded frequency doublers achieves frequency octupling from an 11–13-GHz synthesizer, enabling wide-bandwidth (BW) modulation. Under the default configuration, the four TX and RX channels achieve a maximum TX output power of 13.4 dBm with 12.8% drain efficiency, an RX conversion gain (CG) of 64.4 dB, a minimum RX $text {NF}_{text {ssb}}$ of 8.4 dB, an RX in-band (IB) IP1dB from −49.4 to −43.6 dBm @3 MHz offset, and an RX out-of-band (OOB) IP1dB from −17 to −11.1 dBm at 10-kHz offset across 90–98 GHz. The measured phase noise is −94.06 dBc/Hz at 1-MHz offset with a 90.4-GHz carrier. The root-mean-square (rms) error is 3.52 MHz (0.044%) for a sawtooth chirp with an 8-GHz range and a 20-MHz/$mu $ s chirp rate. Each TX/RX element consumes 208.5/76.5 mW, respectively, and the entire chip occupies a $4.5times 3.8$ mm2 area. To validate the radar operation, a slot substrate-integrated waveguide (SIW) antenna array, with a flip-chip chip-scale package (FCCSP) transceiver, is designed and fabricated on a Rogers 3003G2 PCB. The multiple-input–multiple-output (MIMO) radar prototype achieves a distance resolution of 2.85 cm and an angular resolution of 13° with a field of view (FOV) of 144°.
{"title":"A W-Band FMCW Radar Transceiver Supporting Broadband Modulation in 65-nm CMOS for Intelligent Transportation System Applications","authors":"Shengjie Wang;Jiangbo Chen;Quanyong Li;Jingwen Xu;Wenyan Zhao;Nayu Li;Huaicheng Zhao;Xiaokang Qi;Yen-Cheng Kuan;Gaopeng Chen;Chunyi Song;Qun Jane Gu;Zhiwei Xu","doi":"10.1109/TMTT.2025.3621072","DOIUrl":"https://doi.org/10.1109/TMTT.2025.3621072","url":null,"abstract":"A W-band frequency-modulated continuous-wave (FMCW) radar, implemented in 65-nm CMOS, is proposed for intelligent transportation system (ITS) applications in this article. The system integrates four transmitters (TXs) and four receivers (RXs), along with a frequency synthesizer and a local oscillator (LO) distribution network. Both the low-noise amplifier (LNA) and the power amplifier (PA) adopt multistage cascaded topologies with magnetically coupled resonators (MCRs) to enable broadband operation. An ultrawideband class-B mixer, implemented with only an active switching core, supports continuous operation from 20 to 110 GHz. Furthermore, an LO distribution network featuring three cascaded frequency doublers achieves frequency octupling from an 11–13-GHz synthesizer, enabling wide-bandwidth (BW) modulation. Under the default configuration, the four TX and RX channels achieve a maximum TX output power of 13.4 dBm with 12.8% drain efficiency, an RX conversion gain (CG) of 64.4 dB, a minimum RX <inline-formula> <tex-math>$text {NF}_{text {ssb}}$ </tex-math></inline-formula> of 8.4 dB, an RX in-band (IB) IP1dB from −49.4 to −43.6 dBm @3 MHz offset, and an RX out-of-band (OOB) IP1dB from −17 to −11.1 dBm at 10-kHz offset across 90–98 GHz. The measured phase noise is −94.06 dBc/Hz at 1-MHz offset with a 90.4-GHz carrier. The root-mean-square (rms) error is 3.52 MHz (0.044%) for a sawtooth chirp with an 8-GHz range and a 20-MHz/<inline-formula> <tex-math>$mu $ </tex-math></inline-formula>s chirp rate. Each TX/RX element consumes 208.5/76.5 mW, respectively, and the entire chip occupies a <inline-formula> <tex-math>$4.5times 3.8$ </tex-math></inline-formula> mm2 area. To validate the radar operation, a slot substrate-integrated waveguide (SIW) antenna array, with a flip-chip chip-scale package (FCCSP) transceiver, is designed and fabricated on a Rogers 3003G2 PCB. The multiple-input–multiple-output (MIMO) radar prototype achieves a distance resolution of 2.85 cm and an angular resolution of 13° with a field of view (FOV) of 144°.","PeriodicalId":13272,"journal":{"name":"IEEE Transactions on Microwave Theory and Techniques","volume":"74 1","pages":"1037-1053"},"PeriodicalIF":4.5,"publicationDate":"2025-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146049268","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Microwave dielectric spectroscopy (MDS) is a powerful technique for analyzing the electromagnetic properties of biological substances, offering advantages over lower-frequency methods such as impedometry, which are prone to concentration polarization effects. While flexible MDS sensors and broadband MDS sensors have been independently demonstrated, no prior work has successfully combined both capabilities into a single platform. In this work, we present the first flexible, broadband MDS sensor capable of measuring the complex permittivity of biological liquids directly within their original vials, across a frequency range of 2–$19{,}$ GHz. The novelty of this sensor is enabled by a theoretical framework that describes the excitation of waveguide modes in cylindrical coplanar waveguides (CCPWs). By accurately modeling the influence of these modes on the device’s scattering parameters, we establish a reliable method for extracting the complex permittivity of liquid samples contained within standard laboratory vials. The sensor’s mechanical flexibility allows it to conform to vials of varying shapes and sizes, facilitating noninvasive, contactless measurements. This feature is particularly advantageous for hazardous materials, where minimizing human exposure is essential, and for sensitive biological samples, which are susceptible to contamination if transferred from the containers in which they were originally collected. The proposed sensor addresses key limitations of existing MDS platforms, providing a safe, accurate, and practical solution for broadband dielectric characterization of biological and chemical substances.
{"title":"Broadband Flexible Sensor for Microwave Dielectric Spectroscopy of Liquids in Vials","authors":"Benyamin Harkinezhad;Tomislav Markovic;Robin Evans;Kamran Ghorbani;Efstratios Skafidas;Dominique Schreurs","doi":"10.1109/TMTT.2025.3624149","DOIUrl":"https://doi.org/10.1109/TMTT.2025.3624149","url":null,"abstract":"Microwave dielectric spectroscopy (MDS) is a powerful technique for analyzing the electromagnetic properties of biological substances, offering advantages over lower-frequency methods such as impedometry, which are prone to concentration polarization effects. While flexible MDS sensors and broadband MDS sensors have been independently demonstrated, no prior work has successfully combined both capabilities into a single platform. In this work, we present the first flexible, broadband MDS sensor capable of measuring the complex permittivity of biological liquids directly within their original vials, across a frequency range of 2–<inline-formula> <tex-math>$19{,}$ </tex-math></inline-formula> GHz. The novelty of this sensor is enabled by a theoretical framework that describes the excitation of waveguide modes in cylindrical coplanar waveguides (CCPWs). By accurately modeling the influence of these modes on the device’s scattering parameters, we establish a reliable method for extracting the complex permittivity of liquid samples contained within standard laboratory vials. The sensor’s mechanical flexibility allows it to conform to vials of varying shapes and sizes, facilitating noninvasive, contactless measurements. This feature is particularly advantageous for hazardous materials, where minimizing human exposure is essential, and for sensitive biological samples, which are susceptible to contamination if transferred from the containers in which they were originally collected. The proposed sensor addresses key limitations of existing MDS platforms, providing a safe, accurate, and practical solution for broadband dielectric characterization of biological and chemical substances.","PeriodicalId":13272,"journal":{"name":"IEEE Transactions on Microwave Theory and Techniques","volume":"74 1","pages":"893-904"},"PeriodicalIF":4.5,"publicationDate":"2025-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146049286","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","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/TMTT.2025.3619548
Zehao Zhang;Dantong Liu;Yunxiao Zhao;Hongjia Liu;Guoqiang Liu;Xufeng Kou;Xiong Wang
Transcranial imaging is an indispensable method for the diagnosis of cerebral diseases that are major threats to human health. Microwave-induced thermoacoustic tomography (MITAT) is a promising hybrid technique for nonionizing, noninvasive, and time and cost-effective modality for transcranial imaging with a compact hardware system. Nevertheless, the conventional MITAT technique cannot efficiently deal with the acoustic inhomogeneity issue caused by the skull, which leads to low image quality. Although MITAT combined with deep learning (DL) has shown compelling ability in reconstructing high-quality images in a transcranial manner, the requirement for too many training datasets may hinder potential applications. In this work, we propose a new DL-based MITAT modality that leverages a physics-informed neural network (PINN) to improve the image quality of transcranial imaging using much less training data. The PINN is based on the acoustic reciprocity theorem (ART), and the proposed method is named as DL-MITAT-ART. We perform ex vivo 2-D experimental testing employing intact cynomolgus monkey skulls and blood vessel phantoms. The imaging results demonstrate that the proposed DL-MITAT-ART method can faithfully recover the blood vessel phantoms in a transcranial manner applying only 175 training datasets, more than ten times fewer than those for the traditional DL-MITAT methods. This work provides a novel paradigm for PINN-based MITAT technique for transcranial imaging. It is highly meaningful for cerebral disease diagnosis based on MITAT or ultrasonography and microwave imaging applications involving an inhomogeneous environment.
{"title":"Transcranial Blood Vessel Imaging Through Intact Cynomolgus Monkey Skulls Applying Microwave-Induced Thermoacoustic Tomography Based on a Physics-Informed Neural Network","authors":"Zehao Zhang;Dantong Liu;Yunxiao Zhao;Hongjia Liu;Guoqiang Liu;Xufeng Kou;Xiong Wang","doi":"10.1109/TMTT.2025.3619548","DOIUrl":"https://doi.org/10.1109/TMTT.2025.3619548","url":null,"abstract":"Transcranial imaging is an indispensable method for the diagnosis of cerebral diseases that are major threats to human health. Microwave-induced thermoacoustic tomography (MITAT) is a promising hybrid technique for nonionizing, noninvasive, and time and cost-effective modality for transcranial imaging with a compact hardware system. Nevertheless, the conventional MITAT technique cannot efficiently deal with the acoustic inhomogeneity issue caused by the skull, which leads to low image quality. Although MITAT combined with deep learning (DL) has shown compelling ability in reconstructing high-quality images in a transcranial manner, the requirement for too many training datasets may hinder potential applications. In this work, we propose a new DL-based MITAT modality that leverages a physics-informed neural network (PINN) to improve the image quality of transcranial imaging using much less training data. The PINN is based on the acoustic reciprocity theorem (ART), and the proposed method is named as DL-MITAT-ART. We perform ex vivo 2-D experimental testing employing intact cynomolgus monkey skulls and blood vessel phantoms. The imaging results demonstrate that the proposed DL-MITAT-ART method can faithfully recover the blood vessel phantoms in a transcranial manner applying only 175 training datasets, more than ten times fewer than those for the traditional DL-MITAT methods. This work provides a novel paradigm for PINN-based MITAT technique for transcranial imaging. It is highly meaningful for cerebral disease diagnosis based on MITAT or ultrasonography and microwave imaging applications involving an inhomogeneous environment.","PeriodicalId":13272,"journal":{"name":"IEEE Transactions on Microwave Theory and Techniques","volume":"73 12","pages":"10644-10656"},"PeriodicalIF":4.5,"publicationDate":"2025-11-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11224392","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145778178","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A self-calibrated testing method for Mach–Zehnder modulator (MZM) chips operating up to 110 GHz is proposed based on fiber-free-coupled and self-pilot-based detection. Through introducing and tracking the self-pilot signal in the low-frequency region, the wideband combined frequency response of the cascaded microwave adapter network and the MZM chip under single-tone driving can be extracted without extra optical-to-electrical (O/E) calibration. In the system calibration, the one-port calibration consisting of the power-leveling technique and short–open–load (SOL) calibration is used to de-embed the uneven degradation response of the adapter network, which is attributed to impedance mismatch and transmission attenuation. Finally, the intrinsic half-wave voltage and relative frequency response are both extracted up to 110 GHz with sub-MHz photodetection. Using low-frequency photodetection with large-area photodetectors (PDs), fiber-free coupling replaces traditional waveguide-to-fiber coupling with single-mode-fiber pigtailed PDs, achieving better alignment tolerance without sacrificing performance. The proposed method features O/E self-calibration, single-tone modulation, and fiber-free coupling, which is favorable for on-chip microwave characterization of high-speed MZMs.
{"title":"Fiber-Free-Coupled and Self-Pilot-Based Detection for On-Chip Measurement of Mach–Zehnder Modulators Up to 110 GHz","authors":"Junfeng Zhu;Chao Jing;Simou Wang;Xinhai Zou;Yali Zhang;Sha Zhu;Yong Liu;Ning Hua Zhu;Shang Jian Zhang","doi":"10.1109/TMTT.2025.3622737","DOIUrl":"https://doi.org/10.1109/TMTT.2025.3622737","url":null,"abstract":"A self-calibrated testing method for Mach–Zehnder modulator (MZM) chips operating up to 110 GHz is proposed based on fiber-free-coupled and self-pilot-based detection. Through introducing and tracking the self-pilot signal in the low-frequency region, the wideband combined frequency response of the cascaded microwave adapter network and the MZM chip under single-tone driving can be extracted without extra optical-to-electrical (O/E) calibration. In the system calibration, the one-port calibration consisting of the power-leveling technique and short–open–load (SOL) calibration is used to de-embed the uneven degradation response of the adapter network, which is attributed to impedance mismatch and transmission attenuation. Finally, the intrinsic half-wave voltage and relative frequency response are both extracted up to 110 GHz with sub-MHz photodetection. Using low-frequency photodetection with large-area photodetectors (PDs), fiber-free coupling replaces traditional waveguide-to-fiber coupling with single-mode-fiber pigtailed PDs, achieving better alignment tolerance without sacrificing performance. The proposed method features O/E self-calibration, single-tone modulation, and fiber-free coupling, which is favorable for on-chip microwave characterization of high-speed MZMs.","PeriodicalId":13272,"journal":{"name":"IEEE Transactions on Microwave Theory and Techniques","volume":"74 1","pages":"975-983"},"PeriodicalIF":4.5,"publicationDate":"2025-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146049287","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Frequency-modulated continuous-wave (FMCW) radar has been widely used in motion sensing applications. However, inherent nonlinear errors can significantly degrade its performance. To the best of our knowledge, the impact of nonlinear frequency modulation (NLFM) on displacement motion sensing has not been studied. This article presents the first investigation into the influence of NLFM on displacement motion sensing, supported by theoretical derivations, simulations, and experimental validations. The linearity requirements for accurate motion sensing are analyzed and verified, providing theoretical guidance for displacement motion sensing applications. Building on the demonstrated robustness against NLFM, a voltage-controlled oscillator (VCO) in open-loop FMCW radar architecture is validated for displacement motion sensing. Comparative experiments show that, despite reduced hardware complexity, this architecture achieves performance comparable to that of VCO in closed-loop FMCW radar, with a normalized root-mean-squared error (NRMSE) difference below 1.6%. In robustness experiments, with ±0.2-V supply voltage variation and $9.5~^{circ }$ C temperature drop of VCO, this architecture still achieves precise motion sensing with an NRMSE below 3.2%. The miniaturized prototype based on this architecture successfully reconstructs differential microwave cardiograms (D-MCGs) and senses four types of gestures. These results demonstrate the architecture’s suitability for low-cost, low-complexity, high-accuracy, and compact short-range displacement motion sensing applications.
{"title":"Analysis and Experiments on the Impact of Frequency Nonlinearity on Displacement Motion Sensing With FMCW Radar","authors":"Zhiwei Zhang;Jingtao Liu;Jiayu Zhang;Yijing Guo;Changzhan Gu","doi":"10.1109/TMTT.2025.3622958","DOIUrl":"https://doi.org/10.1109/TMTT.2025.3622958","url":null,"abstract":"Frequency-modulated continuous-wave (FMCW) radar has been widely used in motion sensing applications. However, inherent nonlinear errors can significantly degrade its performance. To the best of our knowledge, the impact of nonlinear frequency modulation (NLFM) on displacement motion sensing has not been studied. This article presents the first investigation into the influence of NLFM on displacement motion sensing, supported by theoretical derivations, simulations, and experimental validations. The linearity requirements for accurate motion sensing are analyzed and verified, providing theoretical guidance for displacement motion sensing applications. Building on the demonstrated robustness against NLFM, a voltage-controlled oscillator (VCO) in open-loop FMCW radar architecture is validated for displacement motion sensing. Comparative experiments show that, despite reduced hardware complexity, this architecture achieves performance comparable to that of VCO in closed-loop FMCW radar, with a normalized root-mean-squared error (NRMSE) difference below 1.6%. In robustness experiments, with ±0.2-V supply voltage variation and <inline-formula> <tex-math>$9.5~^{circ }$ </tex-math></inline-formula>C temperature drop of VCO, this architecture still achieves precise motion sensing with an NRMSE below 3.2%. The miniaturized prototype based on this architecture successfully reconstructs differential microwave cardiograms (D-MCGs) and senses four types of gestures. These results demonstrate the architecture’s suitability for low-cost, low-complexity, high-accuracy, and compact short-range displacement motion sensing applications.","PeriodicalId":13272,"journal":{"name":"IEEE Transactions on Microwave Theory and Techniques","volume":"74 1","pages":"1154-1166"},"PeriodicalIF":4.5,"publicationDate":"2025-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146049303","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-31DOI: 10.1109/TMTT.2025.3624870
Bohui Guo;Junjie Wang;Ran Sui;Yan Ma;Dejun Feng
Temporal-coded metasurface (TCM) provides a new paradigm for radar target feature modulation through dynamic electromagnetic parameter control. However, the false peaks generated by TCM modulation are limited by the sinc envelope constraint of the Fourier transform, and the regular false peaks are easily filtered out by constant false alarm rate (CFAR) detectors. This article proposes a nonuniform bipulse modulation waveform based on the TCM, and the core innovation lies in introducing a time-delay factor $gamma $ as the core degree of freedom to construct a time-domain encoded sequence with nonuniform pulse intervals. This design breaks the periodic constraints of traditional uniform modulation from a physical perspective, enabling nonuniform distribution of echo energy in the feature space and significantly improving the randomness and flexibility of amplitude distribution of the false peaks. Microwave darkroom experiments have shown that by adjusting the time-delay factor and duty cycle, the amplitude of specific harmonic false peaks can be precisely controlled and even eliminated. Finally, CFAR detection further confirms its ability to significantly increase the number of effective false peaks, providing a new approach to breaking through the bottleneck of radar adaptive anti-interference.
{"title":"Nonuniform Bipulse Modulation on Temporal-Coded Metasurface Against Radar CFAR Detection","authors":"Bohui Guo;Junjie Wang;Ran Sui;Yan Ma;Dejun Feng","doi":"10.1109/TMTT.2025.3624870","DOIUrl":"https://doi.org/10.1109/TMTT.2025.3624870","url":null,"abstract":"Temporal-coded metasurface (TCM) provides a new paradigm for radar target feature modulation through dynamic electromagnetic parameter control. However, the false peaks generated by TCM modulation are limited by the sinc envelope constraint of the Fourier transform, and the regular false peaks are easily filtered out by constant false alarm rate (CFAR) detectors. This article proposes a nonuniform bipulse modulation waveform based on the TCM, and the core innovation lies in introducing a time-delay factor <inline-formula> <tex-math>$gamma $ </tex-math></inline-formula> as the core degree of freedom to construct a time-domain encoded sequence with nonuniform pulse intervals. This design breaks the periodic constraints of traditional uniform modulation from a physical perspective, enabling nonuniform distribution of echo energy in the feature space and significantly improving the randomness and flexibility of amplitude distribution of the false peaks. Microwave darkroom experiments have shown that by adjusting the time-delay factor and duty cycle, the amplitude of specific harmonic false peaks can be precisely controlled and even eliminated. Finally, CFAR detection further confirms its ability to significantly increase the number of effective false peaks, providing a new approach to breaking through the bottleneck of radar adaptive anti-interference.","PeriodicalId":13272,"journal":{"name":"IEEE Transactions on Microwave Theory and Techniques","volume":"74 1","pages":"1054-1068"},"PeriodicalIF":4.5,"publicationDate":"2025-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146049294","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Wi-Fi sensor node (SN) identification is necessary for efficient beam control in analog beam forming for wireless Internet of Things (IoT) communication. A method for SN identification using Wi-Fi backscatter, incorporating self-mixing with the backscattered signal from SN and the transmitted (Tx) Wi-Fi signal, has been proposed for efficient multiple SN identification. The method can be easily applied by just adding an RF switch and clock oscillator to the backscatter module without changing the Wi-Fi communication standard. To theoretically reveal the principle, this article proposes a simplified analysis model, assuming the mixer operates ideally as a multiplier. Based on the proposed model, the self-mixing signal spectrum can be simulated, and its signal-to-noise ratio (SNR) is estimated. In a multicarrier modulation (e.g., IEEE802.11g) situation, intermodulation of its subcarriers affects SNR by self-mixing. The noise distribution can also be analyzed by the proposed model. We evaluate the model by measurement with IEEE802.11g (2.4-GHz band) Wi-Fi signal using the fabricated backscatter module and confirm that the measured SNR shows a good agreement with the proposed model.
{"title":"Simplified Model and Its Evaluation of Wi-Fi Sensor Node Identification Method Using Wi-Fi Backscatter and Self-Mixing Receiver","authors":"Yuki Fujiya;Koki Edamatsu;Tomoyuki Furuichi;Takashi Shiba;Noriharu Suematsu","doi":"10.1109/TMTT.2025.3619412","DOIUrl":"https://doi.org/10.1109/TMTT.2025.3619412","url":null,"abstract":"Wi-Fi sensor node (SN) identification is necessary for efficient beam control in analog beam forming for wireless Internet of Things (IoT) communication. A method for SN identification using Wi-Fi backscatter, incorporating self-mixing with the backscattered signal from SN and the transmitted (Tx) Wi-Fi signal, has been proposed for efficient multiple SN identification. The method can be easily applied by just adding an RF switch and clock oscillator to the backscatter module without changing the Wi-Fi communication standard. To theoretically reveal the principle, this article proposes a simplified analysis model, assuming the mixer operates ideally as a multiplier. Based on the proposed model, the self-mixing signal spectrum can be simulated, and its signal-to-noise ratio (SNR) is estimated. In a multicarrier modulation (e.g., IEEE802.11g) situation, intermodulation of its subcarriers affects SNR by self-mixing. The noise distribution can also be analyzed by the proposed model. We evaluate the model by measurement with IEEE802.11g (2.4-GHz band) Wi-Fi signal using the fabricated backscatter module and confirm that the measured SNR shows a good agreement with the proposed model.","PeriodicalId":13272,"journal":{"name":"IEEE Transactions on Microwave Theory and Techniques","volume":"74 1","pages":"1110-1120"},"PeriodicalIF":4.5,"publicationDate":"2025-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146049296","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-30DOI: 10.1109/TMTT.2025.3618460
Tom Shvartzman;Erez Zolkov;Emanuel Cohen
This work presents a technique for reciprocal mixing (RM) cancellation caused by the interaction of close-in phase noise (PN) with a strong out-of-band (OOB) blocker, leveraging local oscillator (LO) delay and the baseband (BB) impedance transparency of N-path filters. By connecting two N-path filters with 90° phase shift (PS) elements and introducing LO delay, a separation is created between the signal and RM content in the BB portion. This enables digital-domain RM cancellation with minimal noise figure (NF) penalty. Implemented in 65-nm TSMC, the proposed technique achieves 16–17-dB RM suppression, with NF increase of approximately 2 dB over the 3-dB NF of a standalone N-path mixer-based receiver, yielding a total NF of 5 dB. The RM suppression is achieved for a 1-dBm blocker located 270 MHz from the 1-GHz in-band frequency while using an LO delay of 0.65 ns.
{"title":"Wideband N-Path Receiver With Reciprocal Mixing Phase Noise Cancellation and Reduced Delay Path","authors":"Tom Shvartzman;Erez Zolkov;Emanuel Cohen","doi":"10.1109/TMTT.2025.3618460","DOIUrl":"https://doi.org/10.1109/TMTT.2025.3618460","url":null,"abstract":"This work presents a technique for reciprocal mixing (RM) cancellation caused by the interaction of close-in phase noise (PN) with a strong out-of-band (OOB) blocker, leveraging local oscillator (LO) delay and the baseband (BB) impedance transparency of N-path filters. By connecting two N-path filters with 90° phase shift (PS) elements and introducing LO delay, a separation is created between the signal and RM content in the BB portion. This enables digital-domain RM cancellation with minimal noise figure (NF) penalty. Implemented in 65-nm TSMC, the proposed technique achieves 16–17-dB RM suppression, with NF increase of approximately 2 dB over the 3-dB NF of a standalone N-path mixer-based receiver, yielding a total NF of 5 dB. The RM suppression is achieved for a 1-dBm blocker located 270 MHz from the 1-GHz in-band frequency while using an LO delay of 0.65 ns.","PeriodicalId":13272,"journal":{"name":"IEEE Transactions on Microwave Theory and Techniques","volume":"74 1","pages":"1098-1109"},"PeriodicalIF":4.5,"publicationDate":"2025-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146049299","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","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：