Pub Date : 2025-03-08DOI: 10.1109/JMW.2025.3561533
{"title":"IEEE Journal of Microwaves Table of Contents","authors":"","doi":"10.1109/JMW.2025.3561533","DOIUrl":"https://doi.org/10.1109/JMW.2025.3561533","url":null,"abstract":"","PeriodicalId":93296,"journal":{"name":"IEEE journal of microwaves","volume":"5 3","pages":"C4-C4"},"PeriodicalIF":6.9,"publicationDate":"2025-03-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10994217","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143925003","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-08DOI: 10.1109/JMW.2025.3560420
Wantao Li;Raúl Criado;William Thompson;Gabriel Montoro;Kevin Chuang;Pere L. Gilabert
This paper presents a feature selection technique based on $ell _{1}$ regularization to select the most relevant weights of artificial neural networks (ANNs) for digital predistortion (DPD) linearization of wideband radio-frequency (RF) power amplifiers (PAs). The proposed pruning method is applied to the first hidden layer of a feed-forward real-valued time-delay neural network, commonly used for DPD purposes. In addition, this paper presents the ANN-based DPD implementation using a graphic processing unit (GPU) with compute unified device architecture (CUDA) units. Thanks to the proposed pruning strategy, it is possible to reduce the ANN complexity significantly, thereby achieving a higher data throughput with the GPU-based implementation. The trade-off among RF performance metrics, number of model parameters and throughput of the GPU implementation is evaluated considering the linearization of a high-efficiency pseudo-Doherty load modulated balanced amplifier (LMBA). The linearized PA operating at an RF frequency of 2 GHz delivers a mean output power of 40 dBm with approximately 50% power efficiency when excited with 5G new radio (NR) signals with up to 200 MHz bandwidth and an 8 dB peak-to-average power ratio (PAPR). The real-time GPU implementation of the ANN-based DPD can meet the linearity specifications with a throughput circa 1 GSa/s.
提出了一种基于$ well _{1}$正则化的特征选择技术，为宽带射频功率放大器（pa）的数字预失真（DPD）线性化选择最相关的人工神经网络权值。将所提出的剪枝方法应用于用于DPD目的的前馈实值时滞神经网络的第一隐层。此外，本文还介绍了使用图形处理单元（GPU）和计算统一设备架构（CUDA）单元实现基于人工神经网络的DPD。由于所提出的修剪策略，可以显著降低人工神经网络的复杂性，从而通过基于gpu的实现实现更高的数据吞吐量。考虑到高效伪doherty负载调制平衡放大器（LMBA）的线性化，评估了射频性能指标、模型参数数量和GPU实现吞吐量之间的权衡。工作在2 GHz射频频率下的线性化PA，当被带宽高达200 MHz、峰值平均功率比（PAPR）为8 dB的5G新无线电（NR）信号激发时，平均输出功率为40 dBm，功率效率约为50%。基于人工神经网络的DPD的实时GPU实现可以满足线性度要求，吞吐量约为1gsa /s。
{"title":"GPU-Based Implementation of Pruned Artificial Neural Networks for Digital Predistortion Linearization of Wideband Power Amplifiers","authors":"Wantao Li;Raúl Criado;William Thompson;Gabriel Montoro;Kevin Chuang;Pere L. Gilabert","doi":"10.1109/JMW.2025.3560420","DOIUrl":"https://doi.org/10.1109/JMW.2025.3560420","url":null,"abstract":"This paper presents a feature selection technique based on <inline-formula><tex-math>$ell _{1}$</tex-math></inline-formula> regularization to select the most relevant weights of artificial neural networks (ANNs) for digital predistortion (DPD) linearization of wideband radio-frequency (RF) power amplifiers (PAs). The proposed pruning method is applied to the first hidden layer of a feed-forward real-valued time-delay neural network, commonly used for DPD purposes. In addition, this paper presents the ANN-based DPD implementation using a graphic processing unit (GPU) with compute unified device architecture (CUDA) units. Thanks to the proposed pruning strategy, it is possible to reduce the ANN complexity significantly, thereby achieving a higher data throughput with the GPU-based implementation. The trade-off among RF performance metrics, number of model parameters and throughput of the GPU implementation is evaluated considering the linearization of a high-efficiency pseudo-Doherty load modulated balanced amplifier (LMBA). The linearized PA operating at an RF frequency of 2 GHz delivers a mean output power of 40 dBm with approximately 50% power efficiency when excited with 5G new radio (NR) signals with up to 200 MHz bandwidth and an 8 dB peak-to-average power ratio (PAPR). The real-time GPU implementation of the ANN-based DPD can meet the linearity specifications with a throughput circa 1 GSa/s.","PeriodicalId":93296,"journal":{"name":"IEEE journal of microwaves","volume":"5 3","pages":"726-738"},"PeriodicalIF":6.9,"publicationDate":"2025-03-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10994208","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143925333","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-08DOI: 10.1109/JMW.2025.3564466
Artem Boriskin;Massinissa Ziane;Mariem Mafamane;Shoaib Muhammad Anwar;Lars Jacob Foged;Maxim Zhadobov
The increasing use of the upper part of the microwave spectrum for wireless communications requires appropriate methods and instrumentation for user exposure assessment. In this context, the IEC TC106 is developing a new international standard for user exposure compliance testing of the next generation 5G/6G wireless devices operating above 6 GHz. As a part of this initiative, the development of a universal reference skin model (RSM) is fundamental for definition of reference data to be included in specifications for body phantom design. In this study, we systematically analyze the impact of the human body near-surface tissue structure on the electromagnetic field (EMF) reflection from the skin surface in the 6–100 GHz range. A conventional multi-layer model is used to calculate skin reflectance as a function of the tissue thickness for the range of thicknesses corresponding to that of typical human skin and near-surface body tissues at four body sites concerned by the 5G/6G wireless use-case scenarios, namely: head, torso, forearm, and palm. The dominant contribution from the epidermis/dermis (ED) layer to the skin reflectance is demonstrated for all body sites in the considered frequency range. A high variation in the reflectance of the palm skin at frequencies above 20 GHz is demonstrated and explained by the matching layer effect associated with a thick stratum corneum (SC). The dry skin model, represented by a semi-infinite homogeneous medium with complex permittivity equivalent to that of the ED tissue, is shown to be an appropriate RSM both for the experimental and numerical evaluation of the absorbed power density (APD) in the 6–100 GHz range. The reference data for the antenna loading and APD at the skin surface are provided for standard reference feeds at 10 GHz, 30 GHz, 60 GHz, and 90 GHz.
{"title":"Universal Electromagnetic Reference Skin Model for APD Evaluation at 6–100 GHz","authors":"Artem Boriskin;Massinissa Ziane;Mariem Mafamane;Shoaib Muhammad Anwar;Lars Jacob Foged;Maxim Zhadobov","doi":"10.1109/JMW.2025.3564466","DOIUrl":"https://doi.org/10.1109/JMW.2025.3564466","url":null,"abstract":"The increasing use of the upper part of the microwave spectrum for wireless communications requires appropriate methods and instrumentation for user exposure assessment. In this context, the IEC TC106 is developing a new international standard for user exposure compliance testing of the next generation 5G/6G wireless devices operating above 6 GHz. As a part of this initiative, the development of a universal reference skin model (RSM) is fundamental for definition of reference data to be included in specifications for body phantom design. In this study, we systematically analyze the impact of the human body near-surface tissue structure on the electromagnetic field (EMF) reflection from the skin surface in the 6–100 GHz range. A conventional multi-layer model is used to calculate skin reflectance as a function of the tissue thickness for the range of thicknesses corresponding to that of typical human skin and near-surface body tissues at four body sites concerned by the 5G/6G wireless use-case scenarios, namely: head, torso, forearm, and palm. The dominant contribution from the epidermis/dermis (ED) layer to the skin reflectance is demonstrated for all body sites in the considered frequency range. A high variation in the reflectance of the palm skin at frequencies above 20 GHz is demonstrated and explained by the matching layer effect associated with a thick stratum corneum (SC). The dry skin model, represented by a semi-infinite homogeneous medium with complex permittivity equivalent to that of the ED tissue, is shown to be an appropriate RSM both for the experimental and numerical evaluation of the absorbed power density (APD) in the 6–100 GHz range. The reference data for the antenna loading and APD at the skin surface are provided for standard reference feeds at 10 GHz, 30 GHz, 60 GHz, and 90 GHz.","PeriodicalId":93296,"journal":{"name":"IEEE journal of microwaves","volume":"5 3","pages":"543-554"},"PeriodicalIF":6.9,"publicationDate":"2025-03-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10994204","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143925223","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-07DOI: 10.1109/JMW.2025.3539957
Jiajia Shi;Yihan Zhu;Jiaqing He;Zhihuo Xu;Liu Chu;Robin Braun;Quan Shi
Human activity recognition (HAR) technology is increasingly utilized in domains such as security surveillance, nursing home monitoring, and health assessment. The integration of multi-sensor data improves recognition efficiency and the precision of behavioral analysis by offering a more comprehensive view of human activities. However, challenges arise due to the diversity of data types, dimensions, sampling rates, and environmental disturbances, which complicate feature extraction and data fusion. To address these challenges, we propose a HAR approach that fuses millimeter-wave radar and inertial navigation data using bimodal neural networks. We first design a comprehensive data acquisition framework that integrates both radar and inertial navigation systems, with a focus on ensuring time synchronization. The radar data undergoes range compression, moving target indication (MTI), short-time Fourier transforms (STFT), and wavelet transforms to reduce noise and improve quality and stability. The inertial navigation data is refined through moving average filtering and hysteresis compensation to enhance accuracy and reduce latency. Next, we introduce the Radar-Inertial Navigation Multi-modal Fusion Attention (T-C-RIMFA) model. In this model, a Convolutional Neural Network (CNN) processes the 1D inertial navigation data for feature extraction, while a channel attention mechanism prioritizes features from different convolutional kernels. Simultaneously, a Vision Transformer (ViT) interprets features from radar-derived micro-Doppler images. Experimental results demonstrate significant improvements in HAR tasks, achieving an accuracy of 0.988. This approach effectively leverages the strengths of both sensors, enhancing the accuracy and robustness of HAR systems.
{"title":"Human Activity Recognition Based on Feature Fusion of Millimeter Wave Radar and Inertial Navigation","authors":"Jiajia Shi;Yihan Zhu;Jiaqing He;Zhihuo Xu;Liu Chu;Robin Braun;Quan Shi","doi":"10.1109/JMW.2025.3539957","DOIUrl":"https://doi.org/10.1109/JMW.2025.3539957","url":null,"abstract":"Human activity recognition (HAR) technology is increasingly utilized in domains such as security surveillance, nursing home monitoring, and health assessment. The integration of multi-sensor data improves recognition efficiency and the precision of behavioral analysis by offering a more comprehensive view of human activities. However, challenges arise due to the diversity of data types, dimensions, sampling rates, and environmental disturbances, which complicate feature extraction and data fusion. To address these challenges, we propose a HAR approach that fuses millimeter-wave radar and inertial navigation data using bimodal neural networks. We first design a comprehensive data acquisition framework that integrates both radar and inertial navigation systems, with a focus on ensuring time synchronization. The radar data undergoes range compression, moving target indication (MTI), short-time Fourier transforms (STFT), and wavelet transforms to reduce noise and improve quality and stability. The inertial navigation data is refined through moving average filtering and hysteresis compensation to enhance accuracy and reduce latency. Next, we introduce the Radar-Inertial Navigation Multi-modal Fusion Attention (T-C-RIMFA) model. In this model, a Convolutional Neural Network (CNN) processes the 1D inertial navigation data for feature extraction, while a channel attention mechanism prioritizes features from different convolutional kernels. Simultaneously, a Vision Transformer (ViT) interprets features from radar-derived micro-Doppler images. Experimental results demonstrate significant improvements in HAR tasks, achieving an accuracy of 0.988. This approach effectively leverages the strengths of both sensors, enhancing the accuracy and robustness of HAR systems.","PeriodicalId":93296,"journal":{"name":"IEEE journal of microwaves","volume":"5 2","pages":"409-424"},"PeriodicalIF":6.9,"publicationDate":"2025-03-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10916995","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143654924","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-05DOI: 10.1109/JMW.2025.3540560
Shih-Shen Fan;Tsu-Te Huang;Jia-Shiang Fu
All-passnetworks (APNs) have long been used in the design of phase shifters. A transmission-line-based quasi-all-pass network (TL-based QAPN) is a series-$C$-configured second-order common-mode APN with its two inductors respectively replaced by two transmission lines. In this work, the TL-based QAPN is, for the first time, analyzed in the context of phase-shifter design. Through the analysis, the conditions for the TL-based QAPN to exhibit zero reflection and the design equations are derived. Two generic switched-network phase-shifter topologies incorporating the TL-based QAPNs are investigated. For experimental verification, a Ka/Q-band 5-bit phase shifter is designed and realized in a 90-nm CMOS technology. The design procedure for using the TL-based QAPN as a phase-shifting bit is described. The measured RMS phase and amplitude errors of the 5-bit phase shifter are less than 4$^{circ }$ and 0.3 dB, respectively, from 29.0 to 45.1 GHz, which translates into a bandwidth of 43.5%. The performance is among the best results in the literature, demonstrating the usefulness of the proposed TL-based QAPN in phase-shifter design.
{"title":"Analysis of Transmission-Line-Based Quasi-All-Pass Network for Phase-Shifter Design","authors":"Shih-Shen Fan;Tsu-Te Huang;Jia-Shiang Fu","doi":"10.1109/JMW.2025.3540560","DOIUrl":"https://doi.org/10.1109/JMW.2025.3540560","url":null,"abstract":"All-passnetworks (APNs) have long been used in the design of phase shifters. A transmission-line-based quasi-all-pass network (TL-based QAPN) is a series-<inline-formula><tex-math>$C$</tex-math></inline-formula>-configured second-order common-mode APN with its two inductors respectively replaced by two transmission lines. In this work, the TL-based QAPN is, for the first time, analyzed in the context of phase-shifter design. Through the analysis, the conditions for the TL-based QAPN to exhibit zero reflection and the design equations are derived. Two generic switched-network phase-shifter topologies incorporating the TL-based QAPNs are investigated. For experimental verification, a Ka/Q-band 5-bit phase shifter is designed and realized in a 90-nm CMOS technology. The design procedure for using the TL-based QAPN as a phase-shifting bit is described. The measured RMS phase and amplitude errors of the 5-bit phase shifter are less than 4<inline-formula><tex-math>$^{circ }$</tex-math></inline-formula> and 0.3 dB, respectively, from 29.0 to 45.1 GHz, which translates into a bandwidth of 43.5%. The performance is among the best results in the literature, demonstrating the usefulness of the proposed TL-based QAPN in phase-shifter design.","PeriodicalId":93296,"journal":{"name":"IEEE journal of microwaves","volume":"5 2","pages":"462-475"},"PeriodicalIF":6.9,"publicationDate":"2025-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10910119","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143655003","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Simultaneous Localization and Mapping (SLAM) of indoor scenarios is usually based on camera or lidar sensors as data sources. However, radar based room scanning offers several complementary advantages to these systems. Among other features, radar sensors are more robust to optical opacities, for example, those caused by smoke and dust in emergency scenarios or in harsh environments. Furthermore, the coherent measurement principle of radar sensors provides highly precise distance information that can be utilized to track the exact position and dimensions of the visible objects. In contrast to camera and lidar, radar applications in room exploration are up to now limited by reduced spatial, i.e. mostly angular/lateral, resolution. This is due to the comparably large wavelength of the utilized signals. This work demonstrates the capabilities of ultrawideband millimeterwave Frequency Modulated Continuous Wave (FMCW) radar sensors operating around 80 GHz in conjunction with the Synthetic Aperture Radar (SAR) imaging method. To allow imaging whilst moving, self-localization techniques based on sub-aperture processing are evaluated. Therefore, we demonstrate a detailed mapping procedure for room exploration applications by exploiting large absolute bandwidths of more than 20 GHz with high resolution imaging techniques on a mobile robot platform.
{"title":"Simultaneous Localization and Mapping (SLAM) for Room Exploration Using Ultrawideband Millimeterwave FMCW Radar","authors":"Tobias Körner;Aman Batra;Thomas Kaiser;Nils Pohl;Christian Schulz;Ilona Rolfes;Jan Barowski","doi":"10.1109/JMW.2025.3541789","DOIUrl":"https://doi.org/10.1109/JMW.2025.3541789","url":null,"abstract":"<underline>S</u>imultaneous <underline>L</u>ocalization <underline>a</u>nd <underline>M</u>apping (SLAM) of indoor scenarios is usually based on camera or lidar sensors as data sources. However, radar based room scanning offers several complementary advantages to these systems. Among other features, radar sensors are more robust to optical opacities, for example, those caused by smoke and dust in emergency scenarios or in harsh environments. Furthermore, the coherent measurement principle of radar sensors provides highly precise distance information that can be utilized to track the exact position and dimensions of the visible objects. In contrast to camera and lidar, radar applications in room exploration are up to now limited by reduced spatial, i.e. mostly angular/lateral, resolution. This is due to the comparably large wavelength of the utilized signals. This work demonstrates the capabilities of ultrawideband millimeterwave <underline>F</u>requency <underline>M</u>odulated <underline>C</u>ontinuous <underline>W</u>ave (FMCW) radar sensors operating around 80 GHz in conjunction with the <underline>S</u>ynthetic <underline>A</u>perture <underline>R</u>adar (SAR) imaging method. To allow imaging whilst moving, self-localization techniques based on sub-aperture processing are evaluated. Therefore, we demonstrate a detailed mapping procedure for room exploration applications by exploiting large absolute bandwidths of more than 20 GHz with high resolution imaging techniques on a mobile robot platform.","PeriodicalId":93296,"journal":{"name":"IEEE journal of microwaves","volume":"5 2","pages":"344-355"},"PeriodicalIF":6.9,"publicationDate":"2025-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10909660","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143654913","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-03DOI: 10.1109/JMW.2025.3539953
Martina Falchi;Angelica Masi;Pierpaolo Usai;Agostino Monorchio;Danilo Brizi
Resonant inductive Wireless Power Transfer (WPT) offers a practical solution for supplying energy to consumer, industrial and medical devices. However, conventional WPT systems face severe limitations if one is interested to the dynamic and arbitrary control of the magnetic field distribution. Consequently, our paper explores the design and implementation of an electronically reconfigurable 5×5 magnetic metasurface for low-frequency WPT applications, operating at 3 MHz. The reconfigurable array is excited by a resonant transmitting coil operating in its near-field region. Through an analytical approach, the metasurface operation can be arbitrarily driven, obtaining the unit-cells current distribution which optimally reshapes the magnetic field for a desired application. In addition, the method also enables the precise determination of capacitive loads of each unit-cell for effectively synthetizing the metasurface response. Then, the reconfigurability process is accomplished by integrating varactor diodes within each unit-cell, providing real-time control of the currents pattern across the metasurface according to the analytical approach outputs. Finally, numerical simulations and experimental measurements on a fabricated prototype are presented, fully demonstrating the system's capability to efficiently switch between arbitrary configurations, either concentrating the magnetic field in specific areas or creating a uniform distribution. This dynamic adaptability addresses important challenges in WPT, such as reduction in alignment sensitivity and efficiency loss over distance, with enhanced flexibility, reliability, and safety.
{"title":"An Electronically Reconfigurable Magnetic Metasurface for Enhanced Low-Frequency Wireless Power Transfer Applications","authors":"Martina Falchi;Angelica Masi;Pierpaolo Usai;Agostino Monorchio;Danilo Brizi","doi":"10.1109/JMW.2025.3539953","DOIUrl":"https://doi.org/10.1109/JMW.2025.3539953","url":null,"abstract":"Resonant inductive Wireless Power Transfer (WPT) offers a practical solution for supplying energy to consumer, industrial and medical devices. However, conventional WPT systems face severe limitations if one is interested to the dynamic and arbitrary control of the magnetic field distribution. Consequently, our paper explores the design and implementation of an electronically reconfigurable 5×5 magnetic metasurface for low-frequency WPT applications, operating at 3 MHz. The reconfigurable array is excited by a resonant transmitting coil operating in its near-field region. Through an analytical approach, the metasurface operation can be arbitrarily driven, obtaining the unit-cells current distribution which optimally reshapes the magnetic field for a desired application. In addition, the method also enables the precise determination of capacitive loads of each unit-cell for effectively synthetizing the metasurface response. Then, the reconfigurability process is accomplished by integrating varactor diodes within each unit-cell, providing real-time control of the currents pattern across the metasurface according to the analytical approach outputs. Finally, numerical simulations and experimental measurements on a fabricated prototype are presented, fully demonstrating the system's capability to efficiently switch between arbitrary configurations, either concentrating the magnetic field in specific areas or creating a uniform distribution. This dynamic adaptability addresses important challenges in WPT, such as reduction in alignment sensitivity and efficiency loss over distance, with enhanced flexibility, reliability, and safety.","PeriodicalId":93296,"journal":{"name":"IEEE journal of microwaves","volume":"5 2","pages":"312-321"},"PeriodicalIF":6.9,"publicationDate":"2025-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10908628","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143654895","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-26DOI: 10.1109/JMW.2025.3535525
Lukas Engel;Jonas Mueller;Eduardo Javier Feria Rendon;Eva Dorschky;Daniel Krauss;Ingrid Ullmann;Bjoern M. Eskofier;Martin Vossiek
This paper presents a deep learning-enabled method for human pose estimation using radar target lists, obtained through a low-cost radar system with three transmitters and four receivers in a multiple-input multiple-output setup. We address challenges in previous research that often relied on extracting ground truth poses from RGB data, which are constrained by the need for 3D mapping and vulnerability to occlusions. To overcome these limitations, we utilized optical motion capture, which is widely recognized as the gold standard for precise human motion analysis. We conducted an extensive optical motion capture study involving various recorded movement activities, which resulted in mmRadPose, a new dataset that enhances existing benchmarks for radar-based pose estimation. This dataset has been made publicly accessible. Building on this approach, we designed an application-tailored radar signal processing chain to generate suitable input for the machine learning algorithm. We further developed an attentional recurrent-based deep learning model, PntPoseAT, which predicts 24 keypoints of human poses using radar target lists. We employed cross validation to thoroughly evaluate the model. This model surpasses previous approaches and achieves an average mean per-joint position error of $6.49 ,mathrm{c}mathrm{m}$ with a standard deviation of $3.74 ,mathrm{c}mathrm{m}$ on totally unseen test data. This excellent accuracy of the reconstructed keypoint positions is particularly remarkable when you consider that a very simple radar was used for the measurements. Additionally, we conducted a comprehensive analysis of the model's performance by exploring aspects such as network architecture, the use of long short-term memory versus gated recurrent units, input data selection, and the integration of multi-head self-attention mechanisms.
{"title":"Advanced Millimeter Wave Radar-Based Human Pose Estimation Enabled by a Deep Learning Neural Network Trained With Optical Motion Capture Ground Truth Data","authors":"Lukas Engel;Jonas Mueller;Eduardo Javier Feria Rendon;Eva Dorschky;Daniel Krauss;Ingrid Ullmann;Bjoern M. Eskofier;Martin Vossiek","doi":"10.1109/JMW.2025.3535525","DOIUrl":"https://doi.org/10.1109/JMW.2025.3535525","url":null,"abstract":"This paper presents a deep learning-enabled method for human pose estimation using radar target lists, obtained through a low-cost radar system with three transmitters and four receivers in a multiple-input multiple-output setup. We address challenges in previous research that often relied on extracting ground truth poses from RGB data, which are constrained by the need for 3D mapping and vulnerability to occlusions. To overcome these limitations, we utilized optical motion capture, which is widely recognized as the gold standard for precise human motion analysis. We conducted an extensive optical motion capture study involving various recorded movement activities, which resulted in <italic>mmRadPose</i>, a new dataset that enhances existing benchmarks for radar-based pose estimation. This dataset has been made publicly accessible. Building on this approach, we designed an application-tailored radar signal processing chain to generate suitable input for the machine learning algorithm. We further developed an attentional recurrent-based deep learning model, <italic>PntPoseAT</i>, which predicts 24 keypoints of human poses using radar target lists. We employed cross validation to thoroughly evaluate the model. This model surpasses previous approaches and achieves an average mean per-joint position error of <inline-formula><tex-math>$6.49 ,mathrm{c}mathrm{m}$</tex-math></inline-formula> with a standard deviation of <inline-formula><tex-math>$3.74 ,mathrm{c}mathrm{m}$</tex-math></inline-formula> on totally unseen test data. This excellent accuracy of the reconstructed keypoint positions is particularly remarkable when you consider that a very simple radar was used for the measurements. Additionally, we conducted a comprehensive analysis of the model's performance by exploring aspects such as network architecture, the use of long short-term memory versus gated recurrent units, input data selection, and the integration of multi-head self-attention mechanisms.","PeriodicalId":93296,"journal":{"name":"IEEE journal of microwaves","volume":"5 2","pages":"373-387"},"PeriodicalIF":6.9,"publicationDate":"2025-02-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10904474","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143655019","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-26DOI: 10.1109/JMW.2025.3533026
Miguel Rodríguez;Raúl Cervera;Carlos Alcaide;Pablo González;Pablo Soto;José V. Morro;Vicente E. Boria;César Miquel España;David Raboso
Multipactor is a harmful effect that may challenge the correct operation of satellite communication systems by limiting the system power-handling capability, and hence its overall performance. Therefore, the development of techniques for predicting the multipactor threshold is of great practical interest. For narrowband components, rough estimations can be obtained from experimental charts, normally resulting in conservative thresholds. More accurate predictions can be obtained with particle simulators, at the expense of a much higher computational effort. This study proposes an approach based on circuital models for swift and accurate multipactor threshold predictions, specifically addressing short-term discharges induced by multicarrier signals in narrowband samples. The use of reduced, but representative, electromagnetic (EM) models of the critical gap region is discussed in detail. Through the use of these models, it is possible to avoid simulating the complete structure, increasing the computational efficiency and enabling the fulfillment of the power requirements at early design stages. The proposed technique is validated through commercial particle simulators, showcasing its efficacy, efficiency, and key benefits.
{"title":"Multipactor Breakdown Threshold Estimation Based on Circuital Models and Particle Simulations for Multicarrier Signals in RF Filters","authors":"Miguel Rodríguez;Raúl Cervera;Carlos Alcaide;Pablo González;Pablo Soto;José V. Morro;Vicente E. Boria;César Miquel España;David Raboso","doi":"10.1109/JMW.2025.3533026","DOIUrl":"https://doi.org/10.1109/JMW.2025.3533026","url":null,"abstract":"Multipactor is a harmful effect that may challenge the correct operation of satellite communication systems by limiting the system power-handling capability, and hence its overall performance. Therefore, the development of techniques for predicting the multipactor threshold is of great practical interest. For narrowband components, rough estimations can be obtained from experimental charts, normally resulting in conservative thresholds. More accurate predictions can be obtained with particle simulators, at the expense of a much higher computational effort. This study proposes an approach based on circuital models for swift and accurate multipactor threshold predictions, specifically addressing short-term discharges induced by multicarrier signals in narrowband samples. The use of reduced, but representative, electromagnetic (EM) models of the critical gap region is discussed in detail. Through the use of these models, it is possible to avoid simulating the complete structure, increasing the computational efficiency and enabling the fulfillment of the power requirements at early design stages. The proposed technique is validated through commercial particle simulators, showcasing its efficacy, efficiency, and key benefits.","PeriodicalId":93296,"journal":{"name":"IEEE journal of microwaves","volume":"5 2","pages":"438-450"},"PeriodicalIF":6.9,"publicationDate":"2025-02-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10904461","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143655081","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-26DOI: 10.1109/JMW.2025.3538856
Ali Kheirdoost;Maysam Haghparast;Behzad Ahmadi;Matteo Oldoni
In this paper, the unitary condition of scattering parameters for lossless and lossy filters is investigated. The primary objective is to establish a mathematical framework that explains the limitations observed in filter amplitude response sharpness at passband edges during practical measurements. It is achieved by analytically evaluating the unitary condition using a closed form coupling matrix model. Various numerical examples are presented to demonstrate the impact of loss on the unitary condition and filter sharpness in practical scenarios. Moreover, we propose a mathematical form of the T term based on the extension to the unitary condition, serving as a gauge to quantify the rounded passband edges in filter response. This T term is related to filter design parameters and the resonators' quality factor, enabling a deeper understanding of filter performance in real-world conditions. The developed theory implication is discussed in analyzing the measurement results of three different filter in different frequency bands.
{"title":"Unitary Condition, Group Delay and Quality Factor in Microwave Filters","authors":"Ali Kheirdoost;Maysam Haghparast;Behzad Ahmadi;Matteo Oldoni","doi":"10.1109/JMW.2025.3538856","DOIUrl":"https://doi.org/10.1109/JMW.2025.3538856","url":null,"abstract":"In this paper, the unitary condition of scattering parameters for lossless and lossy filters is investigated. The primary objective is to establish a mathematical framework that explains the limitations observed in filter amplitude response sharpness at passband edges during practical measurements. It is achieved by analytically evaluating the unitary condition using a closed form coupling matrix model. Various numerical examples are presented to demonstrate the impact of loss on the unitary condition and filter sharpness in practical scenarios. Moreover, we propose a mathematical form of the T term based on the extension to the unitary condition, serving as a gauge to quantify the rounded passband edges in filter response. This T term is related to filter design parameters and the resonators' quality factor, enabling a deeper understanding of filter performance in real-world conditions. The developed theory implication is discussed in analyzing the measurement results of three different filter in different frequency bands.","PeriodicalId":93296,"journal":{"name":"IEEE journal of microwaves","volume":"5 2","pages":"451-461"},"PeriodicalIF":6.9,"publicationDate":"2025-02-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10904455","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143654855","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","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：