This paper presents the development and characterization of a wideband noise source, involving Commercial Off-The-Shelf components. The noise source relies on avalanche noise generation by driving the base-emitter junction of a packaged Si–Ge Heterojunction Bipolar Transistor into reverse breakdown. The paper discusses the noise source operation principle and its extensive characterization in both mm-Wave K band, as well as in C and X bands. Two prototypes were implemented without including output impedance matching, such as to preserve the wideband capabilities of the noise source. Performances were validated in terms of output Excess Noise Ratio (ENR), values reaching 10.8 dB were obtained for the K band at 6.71 mA breakdown current, in a 24–32 GHz bandwidth and $21-102^{\circ}\mathrm{C}$ device temperature excursion. A calibration model is also provided, which fits ENR fluctuations with an average error under 0.05 dB, when considering the maximum current and temperature excursions, as compared with 0.8 dB ENR drift reported for the non-calibrated source. The C and X band validation in 4–6 and 10–12 GHz frequency ranges highlights ENR reaching 25.6 and 22.6 dB, respectively, at 6.9 mA bias current.

In this paper, we introduce and validate signal processing techniques for the estimation of the individual rotation rates of multicopter’s Unmanned Aerial Vehicle (UAV), by exploiting a multistatic radar echoes. To validate the techniques, which have been introduced in our previous works, in this paper, we present a simulator for the multistatic radar echoes scattered by a UAV that integrates quadcopter’s aerodynamics with electromagnetic modeling to generate realistic radar return, characterized by blades rotating in different directions and with different rates depending on the flight trajectory to be traveled. This simulator enables the validation of signal processing.

We leverage the simulator to assess the effectiveness of autocorrelation and cross-correlation (XCF) techniques in separating multiple propellers, both in hovering and along a realistic flight trajectory. Simulated results confirm that XCF allows distinguishing counter-rotating propellers, while co-rotating ones remain unresolved due to their similar speeds. The simulator also demonstrates how variations in rotation rates can be used to infer the presence and weight of a payload. Experimental validation with a C-band continuous wave radar confirms the findings and highlights the impact of material properties on resolution. Finally, we exploit the simulator to investigate the effect of higher carrier frequencies, showing that increasing the operating frequency improves the ability to discriminate co-rotating propellers, supporting improved UAV classification, payload estimation, and trajectory prediction for anti-drone applications.

A multi-finger radio frequency (RF) transistor has been divided into multiple gate sections which can be biased independently. This provides a system designer the ability to dynamically reconfigure the output power and power gain of the device while maintaining good power efficiency and without changing the input drive power. By selectively switching the gate biases below pinch-off to effectively reduce the device’s active periphery, the maximum current of the device can be tuned to “follow” a reduced drain bias voltage, so that the optimum impedance at lower power remains similar to the one at full power, and a fixed matching network can be used to accommodate all power modes. The concept has been tested in a large signal load–pull characterization campaign on a test cell and implemented in a K-band power amplifier (PA) prototype. Measurements on the PA confirm the effectiveness of the method, achieving 30% efficiency at around 4.8–4.9 dB of output power tunability when maintaining a constant input power.

This paper explores the impact of handle silicon substrate resistivity on substrate noise coupling and its influence on the spectral purity of voltage-controlled oscillators (VCOs). Three VCOs were designed using the 28 nm fully depleted silicon-on-insulator (FD-SOI) technology and fabricated on process-of-reference wafer featuring a handle Si substrate resistivity value of 10 Ω.cm and also on high-resistivity (HR) Si handle wafer of 1 kΩ.cm. The output spectrum of the VCOs was measured under two conditions: with and without a 0 dBm noise signal injected into the substrate. The results demonstrate that passivated HR substrates achieve more than 26 dB reduction in parasitic spurs induced by substrate noise. To the best of the authors’ knowledge, this work presents the first fabrication and measurement of VCOs on HR substrates in FD-SOI technology, highlighting their effectiveness in mitigating substrate noise coupling.

This paper presents a highly isolated diplexer-antenna and a dual-band filtenna with high-frequency selectivity. The precise design procedure for the diplexer based on substrate-integrated waveguide technology is presented and effectively integrated with magnetoelectric dipole antenna/microstrip patch antennas to achieve a diplexer-antenna and dual-band filtenna. The proposed configuration enables the design and control of the filtering response of the channels individually. To verify the proposed method, a diplexer-antenna and a dual-band filtenna with operating frequencies of 26.5–27.5 and 28.5–29.5 GHz, a fourth-degree Chebyshev response, and two symmetric radiation nulls for each band are designed and simulated. Finally, the proposed dual-band filtenna is fabricated and measured. The measured peak realized gains for the lower and higher bands are 4 and 4.5 dBi, respectively. Besides, a high out-of-band suppression with a deep roll-off of better than 20 dBi is obtained.

This study presents an innovative method for in situ measurements of electrical conductivity at microwave frequencies using a dielectric resonator (DR)-based probe. The DR, excited in the transverse electric mode with azimuthal index 0, radial index 1, and an open (non-integer) axial field variation denoted by δ (TE01δ mode) by a ridged waveguide, interacts with metallic layers as small as 5 × 5 mm2 on a dielectric substrate, enabling precise conductivity measurements through reflection coefficient analysis and electromagnetic simulations. Validation is performed by comparing the probe’s conductivity extraction results at 13.28 GHz with a conventional resonant cavity method at 10 GHz, demonstrating strong agreement. The used samples are therefore much smaller than what other comparable methods would require. The method is further applied to additively manufactured metallic deposits produced via a micro-dispensing technique, providing insights into their high-frequency (HF) conductivity and the effects of surface roughness. Additionally, a second-generation contact-based probe is developed to extend the characterization capabilities to larger samples and perform HF surface conductivity mapping. This advancement enables localized evaluations of surface properties and correlations with roughness, offering a valuable tool for optimizing additively manufactured components for HF applications.

D-band waveguide transitions that combine a resonant patch on the printed circuit board with multiple waveguide irises are presented and shown to achieve a relative bandwidth of up to 35%. The irises implement an impedance-matching network that is described by a lumped-element circuit model. Parametric sweeps and Monte Carlo simulations quantify the influence of iris dimensions and manufacturing tolerances. The transitions’ group delay is characterized, and its influence on time-domain signals is demonstrated. As an application example, a transition design with six irises is integrated into a multichannel radar frontend with $40\,\mathrm{GHz}$ bandwidth, and measured range profiles are presented. The derived design guidelines enable straightforward customization of iris-based transitions for various applications and frequency bands.