This paper presents the design, fabrication, and characterization of broadband liquid crystal (LC) reconfigurable intelligent surfaces (RISs) operating around 60 GHz and scaling up to 750 radiating elements. The RISs employ a delay line architecture (DLA) that decouples the phase shifting and radiating layer, enabling wide bandwidth, continuous phase control exceeding $360^{\circ}$, and fast response times with a micrometer-thin LC layer of 4.6 µm. Two prototypes with 120 and 750 elements are realized using identical unit cells and column-wise biasing. Measurements demonstrate beam steering over $\pm60^{\circ}$ and -3 dB bandwidths exceeding 9% for both apertures, confirming the scalability of the proposed architecture. On top of a measured nanowatt power consumption per unit cell, aperture efficiencies above 20% are predicted by simulations. While the measured efficiencies are reduced to 9.2% and 2.6%, a detailed analysis verifies that this reduction can be attributed to technological challenges in a laboratory environment. Finally, a comprehensive comparison between the applied DLA-based LC-RIS and a conventional approach highlights the superior potential of the applied architecture.

In this work, we propose interconnection structures for substrateless all-silicon dielectric waveguides, a platform for terahertz (THz) hybrid integration that currently relies on traditional tapered interfaces. We demonstrate two dielectric waveguide-to-waveguide interconnect approaches: a truncated termination and a quarter-wave-matched (QWM) slot termination. Both designs maintain a confined coupling point, enabling consistent coupling. Experimental results show maximum interconnection coupling losses of $5\,\mathrm{dB}$ and $2.5\,\mathrm{dB}$ for the truncated and QWM-slot terminations, respectively, across the 220–330 GHz band. The QWM-slot transmission coefficient remains above $-3\,\mathrm{dB}$ with a misalignment tolerance up to $100\,\mu\mathrm{m}$, representing a $2.5\times$ improvement over the truncated design. Offset characterization along three axes reveals smooth decoupling behavior over a 1 mm range, validating the potential for broadband, contactless interconnects between dielectric waveguide interfaces. These results establish a compact THz interconnect concept with substantially relaxed translational alignment tolerances, making it well-suited for deployment in scalable 6G multi-chip arrays and THz sensing modules.

Terahertz near-field imaging enables high-resolution biomedical and material characterization by combining the non-ionizing nature of terahertz radiation with sub-wavelength spatial resolution. However, the scalability of existing solid-state near-field imagers is fundamentally limited by shared illumination schemes and laterally arranged imaging components, restricting pixel miniaturization and array density. This work overcomes these limitations by introducing a novel pixel architecture that vertically integrates the illumination source, near-field sensing element, and power detector within each pixel, replacing conventional lateral implementations. The proposed architecture reduces the pixel footprint while enabling every pixel to operate as an independent terahertz near-field sensor. Based on this concept, a fully monolithic 32 × 32 (1024-pixel) near-field imager was implemented in a 130-nm SiGe BiCMOS HBT technology. The sensor operates in rolling-shutter mode and achieves real-time imaging at up to 40 fps. The chip consumes 280 mW, corresponding to 273 μW per pixel, and achieves a pixel density of 200 pixels/mm2, the highest reported for monolithically integrated terahertz near-field imaging systems above 100 GHz. The paper presents the scalable architecture together with the design, characterization, and performance analysis of the individual imaging components and discusses the trade-offs introduced by vertical integration.

A miniaturized pentagonal slotted microstrip patch antenna (MPSMA) is designed using FR4 substrate at 4.2 GHz for non-invasive glucose sensing applications. The MPSMA design was optimized for maximal impedance matching, S11, Voltage Standing Wave Ratio (VSWR), efficiency, gain, and directivity. Simulated results were experimentally validated using a vector network analyzer inside an anechoic chamber. The proposed MPSMA achieved a notable return loss of −36.36 dB, a VSWR of 1.0308, efficiency of 75.4 %, gain of 1.5 dBi, and directivity of 2.879 dBi, confirming its performance. A detailed analysis was carried out by placing the antenna near various concentrations of glucose solution (10%–50%), which indicates a significant shift in the resonant frequency and reflection characteristics of the antenna. Principal component analysis was applied for identifying distinct characteristics between various glucose concentrations. The glucose sensing capability of the proposed MPSMA was validated by using a multilayer finger model designed in CST Studio software that mimics real tissue characteristics (Cole–Cole model). The detailed analysis on the multilayer finger model as well as the beaker solution setup highlights the potential of MPSMA in non-invasive glucose monitoring. Further deep learning is implemented using the experimental data for the prediction of glucose from the S11 data. Convolutional neural network classifier results are balanced across most classes. Around most glucose concentrations, true and predicted values are close, with minor overestimation around 4000 mg/dL. The proposed solution is inexpensive and has the potential for diabetes treatment, which ensures the feasibility integration of microstrip antenna technology into next-generation biomedical sensing platforms.

This work presents a broadband two-stage pseudo-differential low-noise amplifier (LNA) fabricated in an advanced 130-nm SiGe BiCMOS technology with $f{_t}$/$f_{max}$ of 470/650 GHz. The design combines Electromagnetic (EM)-based device co-simulations with novel broadband transformer matching networks. The input transformer employs asymmetric broadside-coupled lines to achieve wideband impedance transformation with insertion losses below 2 dB. The novel interstage and output transformers are designed as tapered asymmetric broadside coupled lines, providing a distributed transformation with a high impedance transformation ratio of 5:1 and 4:1, respectively, over a bandwidth (BW) of 100 GHz. The LNA provides a peak small-signal gain of 11.8 dB at 295 GHz with a 111 GHz 3-dB bandwidth spanning 211–322 GHz. The simulated noise figure is between 9.7 dB at 226 GHz and 14 dB at 330 GHz. The LNA has a saturated output power of 4.65 dBm at 280 GHz and an $OP_{1dB}$ of 3.6 dBm, including balun losses and consumes 118 mW DC power. It has a measured group delay (GD) variation of $\pm 3.8$ ps, making it suitable for broadband wireless communication.

Microwave hyperthermia (HT) is a promising adjunct therapy for breast tumors. This study investigates gold nanoparticle (GNP)-enhanced HT through three-dimensional electromagnetic (EM) simulation of a realistic breast phantom containing an embedded 5 mm tumor, excited by an eight-element fractal octagonal ring antenna (FORA) array operating at 2.45 GHz. A 5.85 nM concentration of 21 nm peptide-capped GNPs confined to the tumor region produces a 65–75$\%$ increase in peak specific absorption rate (SAR) and improves the target-to-breast quotient by 39–47$\%$ across both optimization methods, without introducing additional hotspots in healthy tissue. Two complementary optimization strategies are compared: Python-based genetic algorithm (GA) post-processing and direct optimization within CST. The Python/GA approach achieves a six- to seven-fold reduction in computational time (22–24 vs. 145–158 minutes), enabling rapid treatment planning for varying tumor geometries and nanoparticle concentrations. Baseline thermal modeling confirms that tumor temperatures reach 43–45$^\circ$C, within the therapeutic window for HT. These results demonstrate that GNPs effectively enhance EM energy focusing within the tumor region. The FORA antenna delivers the microwave energy, while the GNPs modify the target’s dielectric profile to selectively boost energy absorption-confining the heat to the tumor while safeguarding adjacent healthy tissue.’ Overall, this work provides computational evidence supporting the feasibility of GNP-enhanced microwave HT, while highlighting key optimization trade-offs between computational efficiency and protection of healthy tissue.

Unmanned aerial vehicles (UAVs) are increasingly used to accelerate radio frequency electromagnetic field (RF-EMF) studies and performance assessments of mobile networks, particularly in locations that are difficult or hazardous to access. However, a conductive carbon-fiber airframe located in the reactive near-field of a receive antenna may detune the antenna and distort local field conditions, potentially compromising far-field assumptions. This study quantitatively investigates UAV-induced measurement errors in cellular frequency bands and their dependence on antenna–frame separation. A controlled laboratory procedure compares reference measurements with UAV-mounted configurations while varying the receive antenna distance from the airframe over the 700–3500 MHz range. Results indicate that, for the examined UAV platform and antenna placements, the measured differences relative to the reference are mostly below 3 dB across the investigated frequency range. Increasing the antenna–frame separation reduces these differences by approximately 1–1.5 dB for positions beyond $4\cdot(\lambda/8)$ compared with closer positions. These findings suggest that, under controlled conditions, UAV-mounted configurations can provide sufficiently accurate results for RF-EMF and performance assessment applications. However, the results are specific to the investigated UAV platform, antenna configuration, and laboratory conditions, and cannot be directly generalized without further validation.

A reconfigurable multiphase (MP) modulator is presented that avoids continuous generation and distribution of MP clocks by reconstructing switching waveforms from a memory-stored wavetable using an event-driven pulse generator. Reconfiguration is achieved by updating the wavetable, enabling changes in phase count (e.g., 8–16 phases) and duty cycle to tune spectral performance without modifying the pulse-generator core. A waveform-level proof-of-concept transmitter chain with a commercially available RF amplifier is demonstrated at carrier frequencies of 0.9, 2.4, 3.7, and 6 GHz. Orthogonal frequency-division multiplexing signals with bandwidths up to $500\,\mathrm{MHz}$ are used. In 8-phase, ${50}\%$ duty cycle operation, the measured out-of-band and in-band linearity achieves ${-48.3}\,\mathrm{dBc}$ adjacent-channel leakage ratio (ACLR) with ${1.56}\%$ error vector magnitude (EVM). Scaling from 8 to 16 phases via wavetable updates increases delivered output power by up to ${1.61}\,\mathrm{dB}$ while maintaining comparable ACLR and EVM. In 6-phase, ${33}\%$ duty cycle operation, the measured second- and third-harmonic rejection is approximately ${42}\,\mathrm{dBc}$ and ${45}\,\mathrm{dBc}$, respectively. These results validate the proposed architecture at the waveform level and motivate future integrated implementations with reconfigurable phase count and duty cycle.

A sparse high-gain multifaceted circularly polarized antenna array based on partially reflecting surfaces (PRSs) is designed. Circular polarization is obtained by implementing the sequential phase rotation method at the source antenna level. Beamforming is implemented using phase compensation combined with amplitude weighting proportional to the radiated intensity of the given facet in the targeted direction, following the maximum ratio transmission concept. The radiation pattern is scanned across the angular region specified by the angular gap between adjacent facets. The obtained beamformed result shows a maximum gain fluctuation smaller than 0.5 dB across the scanned sector, together with a sidelobe suppression of 10.27 dB, and the obtained embedded element pattern exhibits a flat top.

A microstrip low-pass filter with sharp roll-off and excellent suppression in both the near- and far-stopbands is proposed. The design incorporates a stepped-impedance hairpin-type resonator (SIHTR) along with a pair of asymmetric T-shaped resonators (TSRs). While the SIHTR offers a tunable 3 dB cutoff frequency and roll-off rate (ROR), its near-stopband suppression degrades at higher ROR values. To address this limitation and further extend the stopband bandwidth, two TSRs with distinct resonant frequencies are positioned at the input and output ports. By fixing their resonant frequencies and fine-tuning their stopband characteristics, the overall suppression bandwidth and stopband attenuation are significantly enhanced. The proposed filter features a compact layout, high selectivity, a wide stopband, and excellent out-of-band suppression performance. Measured results demonstrate an insertion loss of less than 1 dB and a return loss of greater than 15.1 dB in the passband from DC to 7.5 GHz. The 3 dB cutoff frequency is 8.23 GHz, and the transition band is 34.58 dB/GHz from -3dB to -40dB. In the stopband, attenuation exceeds 31.5 dB across the frequency range from 9.17 GHz to 40 GHz.

The rapid advancement of the miniaturization techniques for high-frequency transceivers, radars, and communication systems operating in the D-band (110–170 GHz) has positioned circular dielectric waveguides (CDWGs) as a promising transmission-line alternative. Yet, their potential has not been fully exploited. This work fills this gap by introducing novel, compact CDWG-fed lens antenna solutions that deliver excellent gain and impedance matching performance. A low-loss, low-cost, and flexible CDWG made of low-density polyethylene is used to excite dielectric lens antennas of various sizes, contours, and materials, enabling a systematic evaluation of the impact of lens geometry on antenna performance. Through detailed electromagnetic simulations and experimental validations, lenses of different sizes and contours are designed. The proposed CDWG-lens configuration achieves measured broadband matching better than 17 dB and stable high gain across 115–130 GHz. Measured gains of 21–23 dBi for lenses with diameter of 10.6 mm and gain factor of 27–28 dBi for lenses with diameter of 20 mm confirm the high aperture efficiency and novelty of this approach. A strong agreement between the simulation and the measurement results is also observed. These compact and efficient lens antennas combined with the advantages of CDWG feed, can be integrated as a part of advanced system applications in the D-band frequency range like radar sensing and communications.

This paper presents a novel, independently controllable dual-polarized reconfigurable intelligent surface (RIS) unit cell for 6G upper mid-band ($7.125$–$24.25\,\mathrm{GHz}$) applications. The proposed unit cell is based on the diagonal placement of mutually orthogonal tunable radiators (loaded with varactor diodes) alongside static dual-polarized radiators. Full-wave electromagnetic simulations of the unit cell demonstrate a phase shift range of $270^{\circ}$, a maximum reflection loss of $4.5\,\mathrm{dB}$, and $61\,\mathrm{dB}$ cross-polarization isolation (XPI) within a $400\,\mathrm{MHz}$ frequency range centered at $15\,\mathrm{GHz}$. Furthermore, full-wave simulations of a $16 \times 16$ RIS panel were conducted for various reflection angles under both TE- and TM-polarized illuminations. The results demonstrate a wide-angle beam-steering capability of $\pm60^\circ$, with a maximum beam-pointing error of only $1.7^{\circ}$. The proposed RIS also exhibits accurate and stable reflection control under wide-angle excitation. A $3 \times 3$ prototype was fabricated to validate the unit cell’s performance. Measurement results agree with the simulations, yielding a $270^{\circ}$ phase shift, approximately $7\,\mathrm{dB}$ reflection loss, and $24\,\mathrm{dB}$ XPI. The proposed RIS is a promising candidate for future 6G communication systems.

In this work, a novel hybrid algorithm is introduced that combines the range migration algorithm (RMA) and the back projection algorithm (BPA). The traditional image reconstruction algorithms used for synthetic aperture radar each have trade-offs between computational efficiency and imaging accuracy. The proposed hybrid RMA–BPA approach leverages the computational efficiency of RMA for the initial image reconstruction and object detection, followed by BPA for the refined, high resolution of the cropped region of data. The method of focusing computational resources on the smaller cropped datasets that contain the objects significantly reduces the processing time compared to that of traditional standalone BPA. The hybrid approach’s performance was evaluated over three different scenarios, providing a reduction in computation time for each scenario. Due to the algorithm’s approach to crop the dataset for the specific object, the increased efficiency varied. The different scenarios each produced different times to compute; however, the most impressive result delivered a 75.2% reduction in computation time compared to traditional BPA, without sacrificing the accuracy of the image. The hybrid approach is especially suited for applications that require precise object detection in healthcare, oil and gas, security, and industrial inspections.

Medical microwave imaging (MMWI) is emerging as a promising noninvasive diagnostic modality, offering a safe, portable, and cost-effective alternative to conventional imaging methods. Despite its potential, MMWI poses significant challenges due to the ill-posed and nonlinear nature of the underlying inverse scattering problem. Two main strategies are typically adopted: qualitative imaging, which yields rapid but approximate reconstructions, and quantitative imaging, which aims to recover the full permittivity distribution of tissues. Both approaches substantially benefit from accurate patient-specific background models that enhance reliability and reduce computational complexity. Building upon preliminary results presented at EuCAP 2025, this work advances the development of patient-specific head models for quantitative brain imaging. The study employs anatomically conformal basis functions and extends the model to include cerebrospinal fluid, which represents a strong scatterer with markedly distinct dielectric properties. The proposed framework improves reconstruction accuracy and convergence efficiency, thereby contributing to the advancement of quantitative MMWI toward practical clinical applications.

This paper introduces a methodology for detecting and classifying railway infrastructure elements using ground-penetrating radar (GPR). Seven trips, covering $57.4\,$km across a German track network were recorded. The network includes ballasted tracks, switches, bridges with dominant metal elements, and overpasses. GPR data from a multi-channel sensor array is transformed through a preprocessing pipeline to generate spatial B-scans. Statistical and structural features, including energy, depth, and skewness, are extracted via an overlapping sliding window. The detection algorithm operates as a streaming process, combining correlation-based pattern matching with feature-based thresholds in a state-machine architecture. The evaluation results demonstrate overall good performance for infrastructure detection and classification ($98$–$100\%$), with only minor misclassifications occurring for switch orientation estimation in scenarios involving closely spaced or physically interconnected switches ($82$–$100\%$).

This paper investigates the use of large Reconfigurable Intelligent Surfaces (RISs), also known as “Smart Skins,” to improve outdoor-to-indoor millimeter-wave (mmWave) propagation, taking advantage of ventilation holes commonly found – or easily implemented – in European buildings. By using a transmissive, focusing RIS, the signal can be concentrated and routed through the opening, enhancing signal penetration into the indoor space even in modern, highly insulated buildings that particularly hinder signal penetration. Using Ray Tracing simulations, we compare scenarios with and without RIS at 29 GHz, demonstrating significant indoor signal coverage improvements in the first case. The study highlights the use of RIS – or Smart-Skins – as a cost-effective solution to address the challenges of mmWave propagation in modern building designs.

A novel family of $Ku$-band highly selective multi-band bandpass filters (BPFs) realized using groove-gap-waveguide (GGW) technology is presented. The conceived multi-band BPF structures are derived from a quasi-elliptic-type single-band prototype by applying a split-type single-to-multi-band frequency transformation. The baseline single-band prototype incorporates extracted GGW-rectangular-cavity sections at both the input and output accesses, which are preserved in the multi-band extension as multi-cavity tails. These extracted cavities enable the generation of multiple stopband transmission zeros (TZs) to obtain steep-attenuation characteristics in fully inductive-coupling filter structures without the need for cross-coupling mechanisms. Theoretical design examples of fourth-order dual-band and sixth-order triple-band filtering responses are developed using the coupling-routing-diagram formalism. These examples demonstrate the versatility of the proposed multi-band BPF approach in terms of bandwidths and center frequencies for passbands, spectral symmetry or asymmetry, and out-of-band TZ placement. For experimental-validation purposes, a proof-of-concept GGW dual-band BPF prototype exhibiting fourth-order passbands centered at 14 and 14.4 GHz and five stopband TZs is designed, manufactured, and measured. A compact two-layer vertically stacked GGW configuration is adopted for its physical implementation, confirming the practical feasibility of the proposed architecture. In addition, electromagnetic-simulation-based sensitivity analyses are performed to assess fabrication robustness, and a quantitative comparison with prior-art designs is provided to evaluate the RF-performance merits of the developed Ku band GGW dual-band BPF.

This work proposes a band-switchable electrically small antenna (ESA) system for a wearable health monitoring device. The system operates at multiple 5G-NR bands (875–934 MHz and 1.72–2.2 GHz) using a switchable configuration. The antenna consists of multi-resonant shorted strips that have been miniaturized to fit in meandered form into an area of 63.2$\times$25 mm$^2$ (0.19$\times$0.08$\lambda^2$) available in the intended health monitoring device. To enable operation at two distinct modes, the antenna feedline has been integrated with a switch, which is then matched using lumped matching components. Results indicate that the ESA system operates at 875–900 MHz in Configuration 1, and at 898–934 MHz and at 1.72–2.2 GHz in Configuration 2. The device has been fabricated and integrated into a compact 3D printed casing prior to experimental evaluations. In addition to studying the switchable reflection and radiation performance, the ESA is also evaluated in the presence of a human body phantom. Simulated specific absorption rate levels are less than 0.47 W/kg averaged over 1 g of tissue, which complies with international standards.

This paper presents a novel dual-band cylindrical dielectric resonator antenna featuring a distinctive cross-ring structure that enables circular polarization (CP) across two critical frequency bands. The proposed antenna operates efficiently in both the 8–16 GHz microwave band and the 24–30 GHz millimeter-wave band, making it particularly suitable for ultra-wideband (UWB) and emerging 6G applications. Examination of the $S_{11}$ parameters reveals excellent impedance matching below $-10$ dB, yielding an impressive bandwidth of approximately 8 GHz (64.26%) in the lower band and approximately 6 GHz (23.72%) in the upper band. The antenna achieves stable CP through degenerate orthogonal modes, i.e., $\mathrm{TM}^{x}_{11\delta}$ and $\mathrm{TM}^{y}_{11\delta}$, in the cylindrical dielectric resonator, with 3-dB axial-ratio bandwidth extending across both operating bands (approximately 8 GHz in the UWB band, representing 54.91% bandwidth, and approximately 6 GHz in the mmWave band, representing 20.12% bandwidth). The antenna delivers consistent gain performance, averaging 7.0 dBi in the UWB band and 6.5 dBic in the mmWave band. The novelty of this design lies in its elegant cross-ring feeding structure that simultaneously enables dual-band operation, excellent CP characteristics, and high radiation efficiency in a compact form factor. Simulation results closely align with measured data, confirming the antenna’s suitability for next-generation wireless communication systems.