This research presents a novel circularly polarized filtering antenna with high selectivity and harmonic suppression. The feeding network integrates out-of-phase power dividers and fourth-order filters, which are co-designed with the radiation patch to introduce multiple resonant poles, thus effectively broadening the bandwidth. Circular polarization is attained through equal-amplitude unequal-phase feeding of the power divider. Additionally, cross-coupling introduces radiation zeros on either side of the passband, thereby enhancing passband selectivity. The selectivity at the lower and upper frequencies is quantified as 287 dB/GHz and 269 dB/GHz, respectively. To enhance performance, a periodic square slot structure is incorporated into the coupling section of the filter to achieve a transmission zero in the stopband, which effectively suppresses high-order harmonics. The fabricated filtering antenna demonstrates an S11 ≤ -18 dB, axial ratio below 3 dB, and peak gain of 6.45 dBi over the passband from 3.4 to 3.65 GHz (approximately 7.1%). In the extended stopband, it exhibits an attenuation greater than 16.04 dBi at 2.37f0, confirming its superiority in suppressing clutter signals. These results make the designed antenna as promising candidate for 5G application.

This study presents an implicit unconditionally stable and divergence-preserving alternating-direction-implicit finite-difference time-domain (DP-ADI-FDTD) method for modeling the multi-pole lossy Debye dispersive medium, developed using the Z-transform technique. To address ground-penetrating radar (GPR) problems with enhanced computational efficiency and accuracy, the DP-ADI-FDTD method is integrated with the traditional FDTD method, culminating in an advanced hybrid subgrid framework. In this approach, electromagnetic field components in the coarse grid region are computed using the explicit FDTD scheme, while those in the fine grid region are solved using the implicit DP-ADI-FDTD scheme. The unconditional stability of the DP-ADI-FDTD method permits the use of a uniform time step on both coarse and fine grids, thereby improving computational efficiency and simplifying numerical implementation. The proposed Z-transform-based DP-ADI-FDTD method is validated through theoretical analysis and numerical experiments, demonstrating its divergence-preserving property, unconditional stability, and the effectiveness of the extended hybrid subgrid technique.

A divergence-preserving hybrid implicit–explicit finite-difference time-domain (DP-HIE-FDTD) method is proposed for accurate and efficient electromagnetic (EM) simulations. In contrast to traditional semi-implicit or fully implicit finite-difference time-domain (FDTD) schemes, which often fail to preserve the divergence condition required by Gauss’s law—leading to potential inaccuracies—the proposed method provides a more physically faithful solution. It strictly enforces the divergence-free condition in source-free regions, effectively eliminating nonphysical artifacts. To simulate open-region problems, a complex frequency-shifted perfectly matched layer (CFS-PML) is incorporated into the DP-HIE-FDTD framework, enabling stable and accurate truncation of computational domains. Subsequently, a comprehensive analysis is carried out to evaluate the method’s DP property, numerical stability, dispersion characteristics, and computational complexity. The accuracy and effectiveness of the DP-HIE-FDTD method are finally validated through several representative numerical examples, including a frequency-selective surface (FSS), a slot-loaded shielding structure, a microstrip branch-line coupler, and a dual-band microstrip patch antenna.

The electromagnetic parameters of practical materials are inherently susceptible to stochastic fluctuations arising from individual variability and environmental perturbations, thereby introducing uncertainty into the electromagnetic response of complex systems. To comprehensively capture such randomness, a stochastic complying-divergence implicit finite-difference time-domain (S-CDI-FDTD) scheme is developed, in which the statistical Delta method is systematically incorporated to quantify and propagate parameter-induced uncertainties. This approach enables the simultaneous acquisition of both the mean and variance of EM fields through a single simulation, thus providing an efficient framework for modeling EM systems involving random media. Numerical analyses confirm that the proposed method strictly adheres to Gauss’s divergence law while maintaining unconditional stability. Several numerical examples further verify the accuracy and effectiveness of the S-CDI-FDTD scheme for EM simulations in random media.

All‐perovskite four‐terminal tandem solar cells offer a promising platform for high‐efficiency photovoltaics due to their electrical independence and flexible subcell optimization. However, optical losses such as interfacial reflection and parasitic absorption limit device performance. In this study, a systematic light‐management optimization framework was established, and multiphysics simulations were employed to reveal how perovskite layer thickness, intermediate light‐coupling layer (ILCL) materials and thickness, and top cell structural inversion collaboratively regulate light distribution, electromagnetic field phase, and transmission and reflection characteristics. Optimizing the perovskite layer thickness balances light absorption between subcells, increasing the power conversion efficiency (PCE) from 25.0% to 26.1%. Further introduction of the ILCL with phase‐control design enhances optical coupling, raising the PCE to 28.10%. Numerical simulations indicate that top cell structural inversion effectively suppresses long‐wavelength reflection and enhances bottom cell absorption, resulting in a simulated PCE of 33.73%, approaching the theoretical limit predicted by a semiempirical model guided by experimental data. Quantitative analysis based on admittance and phase matching elucidates the optical mechanisms, providing generalizable guidance for the design of multijunction photovoltaic devices. These results demonstrate that a unified light‐management strategy not only systematically enhances device performance but also provides deep insights into the optical physics of all‐perovskite tandem solar cells.

In this brief, we propose an RF module that integrates a Doherty power amplifier (PA), a low noise amplifier (LNA), and a transceiver (TRX) switch. This configuration enables transmitter (TX) and receiver (RX) mode conversion using a single MOSFET as a switch and a hybrid stacked transformer (HSTF) that covers just 0.0182 mm2. The HSTF co-designs the output matching network (OMN) of the Doherty PA and the input matching network (IMN) of the LNA, thereby reducing insertion loss and significantly miniaturizing the RF front-end. The introduction of an adjustable capacitor bank helps in regulating the parasitic capacitance generated by different operating modes in the HSTF. The 35–41 GHz TRX front-end is implemented in a 65 nm CMOS (1P9M) process. In the TX mode, the high saturated output power (Psat), 1-dB output compression point (OP1dB) and peak power added efficiency (PAE) are measured as 18.6 dBm, 17.7 dBm, and 25.8% at 38 GHz, respectively. The measured PAE at 6-dB and 9-dB power back-off (PBO) efficiency are 18.3% and 13.7%, resulting in efficiency enhancement ratios of 1.46 and 1.51 when compared with an ideal class-B PA. In the RX mode, the minimum noise figure (NF) is 4.8 dB at 38 GHz. The overall chip size of the TRX, including the three-stage PA/LNA and I/O pads, is approximately $1.34\times 0.98$ mm2.