A multi-functional 7-axis antenna measurement system has been implemented using robots for field measurements. We achieved a high level of positional precision, resulting in repeatable and accurate measurement results, even in the sub -THz band, which extends up to 500 GHz. We are developing a robot-based field measurement for sub-THz antennas and AiPs (Antenna in Package) for future comm applications. The system and measurement results will be presented for open discussions.

The increasing direct RF bandwidth needs for optoelectronic links for data centers and other applications has led to increased attention on mm-wave optical converter (OE-EO) measurements. This talk will cover some of the changing uncertainty contributions above 140 GHz RF bandwidth and possible method modifications.

We present a technique for measuring millimeter-wave electromagnetic fields using a miniature fiber-coupled field probe, constructed from electro-optic (EO) crystal-based dielectric material. This method ensures minimal distortion of the electromagnetic waves during the measurement process. A WR-28 waveguide, transmitting Ka-band electromagnetic waves, is utilized to generate an absolute electric field that can be analytically computed. The miniature EO probe is positioned along the propagation direction of the electromagnetic waves at the waveguide's aperture, enabling endoscopic measurements of the electric field within the waveguide. Additionally, the method includes the measurement of standing waves within the waveguide, induced by reflected waves at the aperture. This approach offers an innovative technique for evaluating the waveguide’s transmission characteristics in terms of the V/m absolute electric field.

This session explores advanced measurement techniques at sub-terahertz (sub-THz) frequencies to support the development of 6G components and systems. Key topics include wideband performance characterization and the implementation of next-generation single-sweep solutions utilizing the 0.5 mm connector.
Additionally, the session highlights the role of spectrum analyzers in evaluating over-the-air (OTA) system-level characteristics, offering insights into practical applications and measurement strategies for emerging 6G technologies.

In this paper, we examine the fundamental aspects of traceability that underpin measurements in the sub-THz range. We provide an overview of the research and implementation of new standards at National Metrology Institutes (NMIs), which serve as the primary sources of traceability. Additionally, we discuss the measurement products that facilitate the implementation of sub-THz test systems. Calibration and performance verification are crucial for the reliable operation of test systems in the sub-THz range. We explore various methodologies for implementing these processes, ensuring that the systems deliver accurate and consistent results. Our findings include measurement results and simplified uncertainty calculations, offering insights into the achievable performance levels.

Noise sources are a key technology for the characterization and calibration of THz instrumentation ranging from amplifiers to radiometers. This talk will describe the development of noise sources, both diode and transistor based, with a focus on increasing ENR (excess noise ratio) to enable a wide range of applications. In addition, the characterization methods and error analysis of the noise sources will be presented.

This presentation explains in detail how traceability for the 0.8 mm coaxial connector is established at PTB. It covers SI-traceable calibration for coaxial measurements up to 167 GHz, using seven calculable offset short standards within the 0.8 mm coaxial system.
The talk will present accurate dimensional measurements, from which the S-parameters of the standards are analytically calculated. In the primary experiment, an over-determined calibration approach is used, treating material and surface properties as optimization parameters. The standards, derived from dimensional data, serve as input for an over-determined least-squares calibration.
Measurement results of both the primary and the transfer experiment, as well as associated uncertainty budgets are presented and discussed for selected representative calibration standards. Additionally, the main contributors to the overall uncertainty will be analyzed. The repeatability of the connector interface and the influence of a minimal pin gap are investigated, and challenges are discussed.
This provides the audience with valuable insight and a deeper understanding of the sources of measurement uncertainty in high-frequency coaxial calibrations.

Accurate S-parameter measurements for 6G components can be challenging due to using broader and higher frequencies. To overcome this challenge, 3D electromagnetics simulations as well as lumped circuit simulations may be used to enhance the measurement’s accuracy. Examples will be presented for open discussions on their validities and possible improvements.

Advances in semiconductor technologies are driving optimized and well-performing integrated system designs for 6G. Meeting the broadband, high data rate, and complex modulation demands of 6G require accurate, reliable, wideband modeling and characterization of the active components at 6G frequencies, such as sub-THz frequencies. This presentation will highlight the latest large signal characterization and measurement techniques, including automated electromechanical impedance tuner based passive setups and active/hybrid load-pull methods up to 1.1 THz. The performance expectations of these setups and the unique challenges associated with large-signal measurements will also be discussed.

Ensuring confidence in wafer-level mm-wave measurements requires rigorous calibration methodologies, robust validation strategies, and careful control of influencing factors. This presentation reviews practical methods for increasing trust in both calibration and characterization processes at mm-wave frequencies. Key topics include quantifying and reducing uncertainty, improving data repeatability, validating calibration procedures, and mitigating environmental and setup-induced variations. Real-world examples will illustrate how these recommended methods contribute to higher data integrity and reproducibility at the wafer level in the mm-wave range

This overview presentation will be organized into three parts. Part I will briefly repeat the fundamentals of RF power calibrations traceable to SI, and examples for RF microcalorimeter realizations up to 220 GHz will be shown. Part II will explain how the quantity RF power is disseminated using suitable transfer standards as well as different commercial sensors and receiver systems. In part III different methods of accurate RF power measurements of signal sources will be presented and discussed.
In an open discussion the audience will have the chance to ask questions from the complete field of RF power.

Silicon based phased arrays have been widely adopted in sub-100GHz band for wireless communication and sensing. However, scaling towards > 100GHz band and more array elements is hindered by the shrinking chip area and degrading RF efficiency. This talk will introduce our most recent works on mmW/THz active phased array, which are featured by half-wavelength array spacing, multi-chip signal distribution and low cost packaging.

Sub-terahertz (sub-THz) frequency bands are expected to be a promising platform for ultra-high-speed wireless communication in beyond 5G (B5G) systems. Si CMOS technology, with its advantages in scalability and cost efficiency, is an ideal solution for realizing the widespread adoption of THz wireless communication. However, in the 300 GHz band where wider bandwidths are available, its limited maximum oscillation frequency (fmax) presents significant challenges for RF front-end circuit design.
In this workshop, we will introduce our efforts to address these challenges through the development of Si CMOS-based RF front-end technologies. We present sub-THz transmitters and receivers with the potential for high-speed wireless communication, beam-steering-enabled phased-array circuits and modules capable of forming highly directional beams with high gain, and a hybrid wireless communication system that combines millimeter-wave and terahertz-wave bands.

The rapid advancement of wireless technology has accelerated the informatization and digitalization of human society, the terahertz spectrum holds vast application potential in areas such as mobile communications, intelligent sensing, aerospace satellites, security inspection, and bioimaging. Driven by developments in silicon-based process technology, miniaturized low-cost terahertz radiation sources and terahertz phased array transceiver systems have become focal research topics worldwide. This report will start by examining the technical bottlenecks faced by silicon-based processes in the terahertz frequency band. It will then review the latest global research progress in this field, introduce the research team's recent achievements in terahertz radiator arrays and phased array transceiver systems, and conclude with prospects for future research work.

The frequency range of 110-170GHz is known as D-band, and recently it has been receiving increased attention due to its wide available spectrum and relatively low atmospheric loss compared to neighboring frequency band. Practical D-band radio systems for communication, radar, and imaging require high-power sources ranging from 100mW to 1W, or even beyond. In this talk, recent advances in D-band power amplifiers (PAs) will be discussed, including Sungkyunkwan University's recent PAs producing 100mW to 600mW.

The utilization of the abundant bandwidth in the 300-GHz band is anticipated to enable multi-tens-of-Gb/s wireless communication in the future. To achieve high gain and output power in this band, III-V compound semiconductor-based transistors with a high maximum oscillation frequency (fmax), such as HBTs and HEMTs, are usually employed. However, these transistors are less suitable for digital circuit design compared to the reliable CMOS technology. This presentation introduces 300-GHz-band front-ends designed for short-range (0.1–10 m) and medium-range (10–100 m) high-speed wireless communication. The RF circuits in these front-ends are implemented using InP-HEMT, InP-HBT, and CMOS technologies. Furthermore, various packaging techniques aimed at implementing low-loss modules and on-PCB arrays will be discussed.

THz technology is a promising candidate for future 6G wireless communication systems. While CMOS offers low cost and high integration, it faces challenges like low supply voltage and limited transistor speed. This talk presents our comprehensive efforts to overcome these limitations, spanning from device-level innovations to system-level demonstrations, including an EM-optimized transistor layout enabling a 302.5-GHz amplifier with 30.9-dB gain, a 340-GHz LO source delivering –4.4-dBm output and supporting full 360° phase shifting, a CMOS PLL operating from 324 to 360-GHz with –6-dBm output, a THz heterogeneous integration platform, and a 240-GHz transmitter featuring I/Q phase calibration.

Large signal operation of RF power devices requires accurate and reliable compact models for advanced design applications. However, large-signal measurement setups are the most complicated configurations in the field of RF measurements. Source-pull and load-pull have long been essential device characterization methods that provide comprehensive information about optimal device performance under large-signal operating conditions. The extensive data obtained from load-pull analysis can be used for various purposes, including compact model verification, behavioral modeling, matching network design, and reliability assessment. In this session, we will briefly revisit the large signal operation and load-pull concept. We will then review the fundamentals of modern passive, active, and hybrid-active load-pull techniques used to characterize devices at mm-wave and sub-THz frequencies, and present the benefits and drawbacks of each implementation. We will also present several applications and their impacts, such as double RF-pulse for trapping analysis, load-pull using modulated signals, and stability analysis. Finally, we will share some user experiences.

This presentation will focus on a model of trapping effects in GaN HEMTs and its applications. Trapping effects are well-known characteristic challenges of GaN devices, which degrade model accuracy and, in GaN LNAs, cause recovery phenomena after overdriven input power. This talk will present a highly accurate trapping effects model, as well as a trap-compensation circuit designed based on this model to improve recovery performance.

Machine-learning (ML) and artificial-intelligence (AI) techniques are rapidly becoming indispensable for modeling active and passive devices, especially those operating in the millimeter-wave and terahertz regimes, where traditional compact models fall short. However, truly representative measurement datasets are inevitably sparse and expensive, so deep neural network (DNN) regressors often deliver sub-par accuracy and may even violate basic device physics. Addressing this “data-starved” regime is therefore the pivotal challenge in next-generation device and circuit modelling.
For DC modeling of devices, we demonstrate that a DNN trained on approximately 79,000 measured NMOS-HV samples achieves an 11% mean absolute percentage error (MAPE). By augmenting each bias region with GAN-generated synthetic data, the dataset expands to approximately 109k points; region-aware loss weighting then drives the MAPE down to 9.8%, with sub-4% error in both linear and saturation regions. This part of the workshop details the augmentation strategy and weighted-loss design.
For high-frequency modeling of devices, we present a physics-informed, frequency-aware neural architecture that separately learns frequency features and device/bias features before fusion. The model achieves wide-band interpolation errors of less than 0.043 (MAE) and an R² of approximately 0.97 for key S-parameters on held-out test points, while remaining within the intrinsic device-to-device variability during extrapolation to 65 GHz from a trained model up to 45 GHz.

Large-signal characterisation of GaN-on-Si (MIS)HEMTs is essential for both compact modelling and design and technology co-optimisation. GaN-on-Si brings reliability-linked nonlinearities to the foreground: surface and buffer trapping that drive current collapse, knee walk-out and dispersion, threshold-voltage drift that shifts the intended bias class, and self-heating at elevated drain voltage.
This workshop talk outlines a measurement flow tailored to GaN-on-Si and shows how the dominant risks differ with supply targets, from low VDQ rails for user equipment, where bias-class control and knee behaviour dominate, to higher VDQ infrastructure operation, where thermal margins and probe stability are critical. Using representative results from previous studies, it compares front- and back-barrier choices and channel scaling, and translates their impact on Imax, Vknee and thermal resistance into first-order expectations for Psat and PAE. The goal is to provide a compact set of methods and DTCO guidelines that make GaN-on-Si large-signal data model-ready and reduce surprises between characterisation, modelling and PA design.

The noise behaviour of microwave amplifiers plays a significant role in determining the power budget of communication and radar sensing systems. Minimizing the noise figure of the active circuits and reducing the noise power of high-gain modules and benchtop test amplifiers, improve the performance of RF systems and measurement setups.
Historically, Gallium Arsenide (GaAs) technology has set the standard for superior noise performance and has maintained a dominant position in the market for several years. It is well-regarded for its low noise figures, making it the go-to choice for various applications.
However, Gallium Nitride (GaN) technology is gaining popularity due to its ability to handle large breakdown voltages, along with offering competitive gain and noise performance.
In this session, we will begin by exploring noise parameter measurement methods that extend into the sub-terahertz (THz) frequency bands. Following that, we will discuss the noise modeling process, including various noise models and their applications. In the final section, we will look at some GaN noise modeling examples and recent applications, providing practical insights.

This session delves into how mmWave-SDR enables seamless frequency conversion (6–30 GHz) and its applications in FR2/FR3, SATCOM, radar, and Integrated Sensing and Communication (ISAC), extending its capabilities to integrate with FR2/FR3 OAI and the O-RAN architecture. Attendees will explore advanced beamforming techniques, efficient resource allocation strategies, practical deployment solutions, and a demonstration of SDR integration with mmWave technology to enhance network capacity through Dynamic RIS and MIMO.

In this talk, an intuitive approach to assessing advantages of beamforming in 5G/6G wireless communication is proposed as a novel try and practical demonstration of importance of alignment between the transmitter's and receiver's beams working in millimeter-wave frequency bands. The effects of the misalignment and alignment between beams need to be checked for, which was conducted with a horn antenna and the beamforming/beam-steering metamaterial antennas for RF-to-RF wireless connectivity between the horn and the antennas of interest including RIS reflectarrays, concerning the changing angle of the beam from the aforementioned antenna, was tested, showing over a substantial enhancement in received power. This direct electromagnetic link test was accompanied by examining 64-QAM constellations for beam-angle change. This verification is suppoted by the use of TMYTEK BBox.

Integrated Sensing and Communication (ISAC) is gaining prominence as we transition towards 6G communication. Global research initiatives encompass algorithm development and hardware prototyping. This presentation provides an up-to-date overview of ongoing ISAC research, encompassing diverse sensing, radar algorithms, and beam management techniques. Additionally, hardware prototypes for the 6G ISAC will be showcased. The presentation delves into wireless localization and sensing in both far-field and near-field regions. Concluding remarks will highlight recent advancements in wireless sensing utilizing Rydberg atomic receivers.

The classic filter design methods are tapped into first to remind the attendees of the low-pass filter prototype, frequency transformation, Richardson’s formula and Kuroda identity to make basic bandpass filters. This is followed by the filter synthesis methods covering the canonical configuration, derivation of the coupling coefficient matrix and conducting the similarity transform. Additionally, stochastic approaches to building up the filter’s matrix to meet the user’s requirement are introduced. They are realized by the RH transmission-lines and waveguides and furthered by zeroth-order resonator (ZOR) based BPFs as CRLH metamaterials to be compact and performing better demanded by modern mobile and satellite communication.

This tutorial talk provides a comprehensive overview of the fundamentals in modern microwave filters, covering all aspects from the synthesis theory to the design techniques. It covers a series of design steps including lowpass prototype filters with generalized Chebyshev functions as transfer functions, synthesis of coupling matrices, and bandpass filter design using coupling coefficients and external Q factors, through illustrative examples.

Using filters and other passive components, wireless transceivers are developed for high data transmission rate connectivity. Active components are designed considering their neighbors, say, filters, power-dividers, couplers and antenna elements interconnecting in a friendly manner. Especially, in order to make sure the entire system should meet the bandwidth and noise figure, channel selectivity and insertion loss of the filters as important factors are handled from theory to practice. Also, their role in the sensor working as a UWB device is mentioned with design examples.

High fidelity measurement of qubit gates require low noise amplifiers. Superconducting parametric (passive) amplifers promise high signal to noise ratio as the added noise level approaches to the quantum limit and gain enhancing mechanisms are progressing. In this session the current state of the art of the field in high fidelity qubit readout will be reviewed and a broad picture of the field is skteched for the audience.

Strong nonlinearity of Josephson junctions (JJ) in parametric amplifiers mandates using large-signal and nonlinear simulation techniques because small-signal analysis and analytic models based on coupled-mode theory cannot fully capture the mechanisms behind amplification e.g., pump depletion. In this presentation, we show how nonlinear elements based on JJ e.g., SQUID, and SNAILs are modeled mathematically. Then the mechanisms of gain in three-wave mixing and four-wave mixing amplifiers and exchange of power between different pump harmonics are introduced. The concept of dispersion and bandgap engineering are presented. We show how the harmonic balance (HB) method faithfully captures the physics of amplification in TWPAs. A few real design examples in which the simulation results are in close match with the experimental measurement, are demonstrated.

In superconducting qubit-based quantum computing circuits, multiplexed readout is used to measure several qubits via a single transmission channel. Achieving optimal measurement performance in this framework requires a broadband quantum amplifier that covers the frequency range of the qubits. The Josephson Traveling Wave Parametric Amplifier (JTWPA) is widely utilized, providing the necessary broadband amplification. To minimize the complexity of optimizing thousands of Josephson junctions, the band amplification of a JPA has been explored by employing an auxiliary resonator of which resonant frequency matches the JPA's effective resonance. In practical implementations, achieving precise frequency matching between these components faces significant challenges so that it is crucial to understand the tolerance for detuning. In this study, we investigate the effects of resonance detuning on bandwidth enhancement and signal gain. In addition, we optimize JPA performance based on this context.

Josephson parametric amplifiers are crucial components for quantum state measurement of superconducting qubits. Compared to HEMT amplifiers that operate at 4 K with noise temperatures of 2-5 K, Josephson parametric amplifiers can operate at 10 mK and add noise close to the quantum limit. Early Josephson parametric amplifiers had very narrow bandwidths of tens of MHz, enabling only single qubit measurements. However, as the number of qubits increases, broadband Josephson parametric amplifiers capable of simultaneous readout of multiple qubits have become necessary. Recently, traveling wave parametric amplifiers using thousands of Josephson junctions or nonlinearity of superconducting materials have been widely studied. However, due to fabrication challenges, this study introduces a Josephson parametric amplifier utilizing an impedance transformer. An impedance transformer combining quarter and half wavelength CPW was employed. The fabricated device demonstrated a 10-fold bandwidth enhancement based on 15 dB gain criteria. This device is expected to be utilized for multi-qubit readout, squeezing, and other applications.

The Traveling-Wave Parametric Amplifier (TWPA) plays a crucial role in qubit measurement systems, especially in cryogenic environments where conventional amplifiers generate additional noise that degrades signal quality. TWPA amplifies signals while minimizing added noise to quantum-limited levels, significantly enhancing the accuracy of qubit state readout. In large-scale qubit systems, multiplexed readout is implemented for efficient measurement. In this configuration, TWPA amplifies multi- frequency readout signals from multiple qubits, ensuring stable and accurate state discrimination while improving the Signal-to-Noise Ratio (SNR) to reduce measurement uncertainty and enhance readout fidelity. In this presentation, we report experimental results obtained by optimizing parameters to achieve high measurement fidelity in a multiplexed qubit readout. We evaluated TWPA performance by applying a cost function reflecting key performance metrics, along with an optimization algorithm. This approach successfully stabilized signal gain and minimized information loss, resulting in a significant improvement in qubit measurement fidelity. This study provides a systematic methodology for TWPA optimization in large-scale quantum computing architectures.

Design and simulation of superconducting quantum computing chips and amplifiers relies a lot on circuit-level, 3D or 2.5D EM simulation, layout generation, and post-layout analysis. Switching between different tools and packages to perform these steps renders the design and production cycle of chips very tedious and time/consuming. This also makes the design prone to errors and mistakes. In this part of the workshop, we show how a unified and seamless design process can be performed from the idea to the pre-production layout and EM simulation using Keysight ADS and Quantum Pro package. We demonstrate design steps of a superconducting qubit chip and a Josephson junction-based traveling wave parametric amplifier. The conversion of schematic to layout and post-layout simulation with FEM and MoM methods will be demonstrated. The RF parameters of pre-fabrication layout of a parametric amplifier are extracted, and co-simulation is performed with harmonic balance (HB) method which shows a striking match with the experimental results.