Interference-margin (IM)-based graded screening (GS) is widely used in engineering practice for equipment-to-equipment electromagnetic interference prediction (EMIP) and has been well established for analog modulation scenarios. However, the power spectra of digitally modulated signals exhibit pronounced morphological features and quasi-periodic fluctuations. When conventional power spectrum threshold criteria are directly applied, noticeable spectral ripples can be introduced, leading to unstable screening outcomes and degraded prediction accuracy and repeatability. To address this issue, this article investigates representative digital modulation schemes, including ASK, phase-shift keying (PSK), and frequency-shift keying (FSK), and proposes a power spectral envelope (PSE)-based modeling and application method. The proposed approach parameterizes the PSE by constraining envelope morphology and out-of-band energy roll-off behavior, and employs a genetic algorithm to optimize the model parameters. In addition, a unified database admission criterion is established to evaluate the validity and stability of the fitted envelopes, thereby constructing a PSE model database for digital signals. The resulting database is then leveraged in the GS frequency-screening stage to refine and strengthen IM evaluation. Simulations and experimental validations are conducted using communication equipment employing the above modulation schemes as the test objects. Compared with susceptibility test results, the proposed method achieves a prediction accuracy of no less than 84.81% within a $\pm \text{3}\;\text{dB}$ error bound and no less than 91.84% within a $\pm \text{6}\;\text{dB}$ bound, demonstrating its engineering applicability for GS-based EMIP in fundamental digital modulation settings.
The massive deployment of frequency modulated continuous wave (FMCW) radars in autonomous vehicles will lead to severe electromagnetic interference issues. Mutual interference mitigation is vital for ensuring the detection performance of automotive FMCW radars. Existing methods struggle to maintain a balance between interference mitigation performance and computational efficiency in complex scenarios with multiple interferences, which restricts the applicability in practical scenarios requiring rapid response. To address the predicament, this article proposes an efficient method for mutual interference mitigation among automotive FMCW radars using the short-time Fourier transform (STFT) spectrogram sliding cancellation. The interferences and target signals exhibit different time–frequency characteristics. After sliding the STFT spectrogram along the time dimension, the intensity of interference bins fluctuates dramatically while that of target bins remains virtually unchanged. Therefore, this intensity variation is employed to detect and cancel interference bins in the STFT domain. The effectiveness of the proposed method is validated through both simulations and experimental results in different interference scenarios. The results show that the proposed method can effectively eliminate the interferences and recover the target signals even if 100% of the samples are contaminated by the interference, and it is computationally efficient.
Millimeter-wave (mmWave) links in greenhouses are challenged by dense vegetation, which blocks line-of-sight and produces scattering-dominated channels. We present a hybrid ray-tracing (RT) to full-wave finite-difference time-domain (FDTD) framework to jointly assess propagation and exposure at 26 GHz in a realistic tomato-greenhouse geometry. RT simulations (Sionna) were performed for 440 receiver (RX) locations and the dominant multipath components were mapped to a coherent multiplane-wave excitation in FDTD (Sim4Life) to compute absorbed metrics for a Bombus terrestris model (total absorbed power) and a human head phantom with absorbed power density averaged over 4 cm² (APD$_{4\text{cm}^{2}}$). For isotropic transmission, the mean coherent path gain of approximately −122 dB corresponds to a noise-limited regime, with insect absorbed power typically in the 10−14–10−12 W range. With directional illumination, a reconfigurable intelligent surface (RIS) increases the mean coherent path gain by approximately 56.9 dB relative to the isotropic baseline and narrows the central 90% gain interval to approximately [−76, −50] dB and produces substantially stronger location-dependent optimized links; the resulting insect absorbed power is typically 10−10–10−8 W and reaches 1.5 × 10−6 W under the strongest optimized RX condition. Human exposure remains below 10−4 W/m2 in all scenarios, is confined to superficial tissues, and remains more than 53 dB below the international commission on non-ionizing radiation protection (ICNIRP) 2020 general-public local basic restriction. The results quantify the connectivity gains of RIS-assisted mmWave links in vegetation-dense layouts and the spatial variability of the corresponding optimized propagation and exposure conditions.
An automated, sensitive-feature identification–visualization method is proposed in this article to address the challenges—namely, the limited identification accuracy, reliance on single-dimensional static criteria, and bottlenecks in Black box modeling efficiency—in evaluating the conducted susceptibility of analog and mixed-signal (AMS) integrated circuits (ICs) in complex electromagnetic environments. The proposed method is based on the data selection initialization of an improved k-means clustering approach, which does not require predefined output failure criteria for evaluation. The developed method addresses issues inherent to traditional clustering techniques, including sensitivity to initial conditions and random clustering outcomes, by leveraging the chip’s baseline response characteristics under undisturbed conditions as a physical constraint and integrating a multi-indicator fusion algorithm. The proposed method generates heatmaps of electromagnetic interference susceptibility distributions, providing intuitive quantification and visualization of the chip’s susceptibility evolution patterns in the 2-D frequency–power interference space. Experimental results demonstrate that this approach had good consistency with the susceptibility boundaries determined according to traditional standards and, more critically, revealed the continuous degradation trajectory of AMS ICs from the linear operating region through the nonlinear distortion transition zone to the functional failure zone. Furthermore, the partition modeling strategy guided by clustering results reduced data acquisition and model training costs by eliminating redundant test points and focusing on critically sensitive regions while maintaining the black-box model’s prediction accuracy (the model iteration convergence speed improved by approximately 33.7%, and the total time was reduced by 25.9%). This work provides an objective evaluation tool for the electromagnetic susceptibility of AMS ICs without predefined output failure criteria, characterizes the susceptibility response evolution under broadband interference, and establishes a robust theoretical and methodological foundation for chip screening, in situ replacement, and efficient electromagnetic behavior modeling in high-reliability scenarios.
This article investigates the conducted electromagnetic susceptibility (EMS) mechanism of commercial Gigabit Ethernet switches and its relationship with port-to-port conducted electromagnetic interference (EMI) transfer. The EMS results reveal the susceptible frequency band from 70 to 100 MHz, with the critical threshold occurring around 90 MHz. This frequency-dependent behavior is explained by the combined effects of frequency-selective cable response associated with reflection and standing-wave enhancement, high-frequency parasitic coupling through the Ethernet isolation transformer, and the effective baseband response and filtering characteristics of the PHY receiver. The agreement between the explained resonance-related sensitive band and the measured minimum-threshold frequency supports the proposed EMS mechanism. Furthermore, a high-frequency equivalent model is developed to quantify port-to-port EMI transfer by incorporating transformer parasitic capacitance, inter-port capacitive crosstalk, and common-ground transfer impedance. The simulated transfer characteristics are validated by measurements, confirming that the model captures the dominant frequency-dependent inter-port coupling behavior. A comparative analysis of the EMS thresholds and port-to-port transfer characteristics provides frequency-domain evidence for identifying the dominant coupling paths in multiport Ethernet switches. The proposed mechanism-based analysis and validated model offer a reference for EMC assessment and robustness optimization of multiport Ethernet equipment.
A hybrid model is proposed for predicting the response of printed circuit board (PCB) traces excited by electromagnetic (EM) field radiated from the other PCB. The proposed hybrid model combines the equivalent dipole model with the modified Kron–Branin (KB) method. The equivalent dipole model is based on electric and magnetic dipoles, which can produce the same EM fields as the actual electromagnetic interference source. The modified KB method can calculate the terminal responses of PCB traces by incorporating the illumination of the external EM field and discontinuous structures in actual PCB traces. The modified KB method is verified using two actual trace models in the frequency range of 0.1to 3 GHz, and the maximum absolute difference of terminal responses compared with the full-wave simulation is less than 5 dB. Based on the validation of the modified KB method, EM coupling between two parallel PCBs is investigated. The radiation fields of the emitter PCB are modeled by the equivalent dipole model via near-field scanning experiments and are used to calculate the coupling voltages with two victim PCBs in different slots from 0.1 to 3 GHz. The results show generally consistent trends with the measurements, particularly for slot B, where good agreement is obtained across the investigated band. For slot C, the proposed method still captures the main variation tendency and response level, although relatively larger deviations are observed at some frequencies. For the two victim PCBs in slots B and C, the mean root-mean-square error between the proposed hybrid model and that of the experimental measurements is 3.91and 7.06 dB, respectively. These results indicate that the proposed hybrid model is potentially useful for estimating EM coupling effects of multiple PCBs placed in one module with different layouts.
The increasing power density of electric-vehicle (EV) inverters has intensified the demand for compact electromagnetic-interference (EMI) filters. However, large low-frequency common-mode (CM) currents generated during inverter operation can bias the magnetic cores of CM chokes, reducing their CM impedance and causing EMI-filter attenuation to be overestimated under conventional small-signal characterization. This article investigates AC CM current-induced core saturation in EMI filters for EV inverter applications. Time-domain measurements are first performed to quantify the inverter-generated CM current and identify its dominant low-frequency component. A dedicated AC-bias characterization setup is then developed to superimpose controlled kHz-range CM excitation while measuring the MHz-range impedance and equivalent permeability of magnetic cores. Measurements on MnZn, NiZn, and nanocrystalline cores show material-dependent permeability and impedance degradation under AC bias, with core impedance reductions exceeding 30% under selected AC-bias conditions. In addition, a single-stage EMI filter prototype is designed, and its performance is evaluated through equivalent CM impedance analysis. The measured filter-level impedance degradation and resonance-frequency shift show good agreement with simulations up to 10 MHz. These results demonstrate that AC CM current-induced core saturation should be considered in compact EMI-filter design for high-power inverters.
In this work, a common-mode (CM) equivalent-circuit model ensuring accurate prediction of the CM conducted emissions (CE) of real electric drive systems (eAxles) is proposed and validated. The proposed model allows for computationally effective frequency-domain (FD) simulation, without the need for additional time-to-frequency domain postprocessing as usually required by three-phase models available in the literature. In addition, an enhanced equivalent-circuit model of the electric motor is derived starting from CM and differential-mode impedance measurements, experimentally validated over the frequency range of interest from 1 kHz up to 20 MHz, and used for time-domain and FD simulation. The CM CE spectrum obtained by integrating the enhanced motor model into the developed CM equivalent-circuit model of a powertrain is eventually validated against measurements carried out on a real eAxle system with a high degree of integration and complexity.
In this article, we present an experimental analysis of electromagnetic interference in mobile communication systems caused by unmodulated continuous wave (CW) noise. The CW noise model represents a harmonic component of a clock signal that is unintentionally radiated from digital circuits based on electromagnetic noise measurement. The effects of CW noise on mobile communication systems are evaluated through wired and wireless measurements using multiple commercial user equipment. The results indicate that interference characteristics caused by CW noise strongly depend on the relationship between the CW noise frequency and channel mapping of the communication system. Severe degradation occurs when CW noise interferes with reference signals, while a slight impact is observed when it aligns with the center of channels that allocate only data bits. These trends were consistently observed across all tested equipment. This finding clarifies conditions under which CW noise used to represent a harmonic component of clock signal radiated from digital circuits, leads to strong interference and provides practical insights to set an emission limit and design high-reliability mobile communication systems.
This study presents a comprehensive dosimetry and thermal modeling framework for highly focused millimeter-wave (mmWave) exposure on rat dorsal skin. A 60-GHz dual-lens exposure system was developed to generate a localized beam with a spot size of approximately $\mathbf {\text{1}}\; {\text{cm}}^{2}$, facilitating precise heating. To address the multiscale computational challenge posed by the exposure distance and fine anatomical structures, a hybrid numerical method combining plane-wave spectrum expansion with the finite-difference time-domain method was employed. This approach yielded accurate dosimetry computation of the absorbed power density and resultant skin temperature rise within high-resolution anatomical models. While thermal simulation initially matched the transient experimental heating, discrepancies emerged in the steady-state phase due to dynamic thermoregulatory responses. To resolve this, a Weibull distribution-based approximation was introduced to characterize the time-dependent temperature rise behavior. Through an inverse fitting procedure, the blood perfusion rates of the skin and muscle layers, identified as the dominant cooling factors, were estimated and integrated into the Pennes’ bioheat transfer equation. The optimized thermal model demonstrated good agreement with experimental measurement. These findings establish a calibrated, experimentally validated framework for accurate mmWave dosimetry assessment, providing methodological support for future refinement of thermally based exposure assessment approaches in human protection guidelines.
This article presents a dedicated verification and harmonic reference kit for contact harmonic measurement, consisting of a printed circuit board (PCB) for contact harmonics testing, an S-parameter verification PCB, and a harmonic reference source PCB with a controlled nonlinear element. Experimental results demonstrate that the proposed reference source provides predictable harmonic-versus-power behavior, good repeatability, a well-defined operating range, and adjustable harmonic output. The proposed kit provides a practical method for verifying the radio frequency (RF) signal path, checking the system self-harmonic baseline, and comparing the measured harmonic response with a repeatable nonlinear reference. It improves measurement reliability and provides a basis for future interlaboratory comparison of contact harmonic measurements.
Prior studies have revealed that a double shield achieves higher shielding effectiveness (SE) than a single shield of the same total conductor thickness. This letter generalizes this framework to an arbitrary number of layers, examining SE improvement by partitioning a single conducting slab into equally thick subsheets interleaved with uniform air gaps. We show that when the product of the total thickness (sheets plus gaps) and the slab thickness exceeds the skin depth squared, increasing either the layer count or the total thickness significantly enhances SE. As layers are added, SE initially rises, then declines, and finally stabilizes at the homogenized limit. In addition, half-wavelength resonant effects can cause SE dips at certain higher frequencies, but these can be eliminated by further increasing the layer count.
The disruption, damage, or destruction of critical infrastructures can have a significant impact on the stability and safety of modern societies. Therefore, the electromagnetic compatibility (EMC) of critical infrastructures, including the protection for intentional electromagnetic interferences, is of growing concern. Critical infrastructures are designed as networked and distributed systems with an in-service period of up to several decades. This article provides some reflections on how the necessary level of EMC in critical infrastructure can be ensured by the application of a holistic EMC risk management. The EMC risk management is the most comprehensive and formal application of risk management principles and processes to EMC. The EMC risk management process is a process that involves the systematic application of procedures and practices to identify, analyze, evaluate, treat, monitor, and review EMC risks throughout the entire life cycle of critical infrastructures. The objective of this article is to introduce the concept and principles of risk management.
Previous work demonstrated that a simple pseudoinverse estimator can be used to reconstruct cavity fields from sparse probe measurements in the fundamental and undermoded regimes (Wallace et al., 2026). That work is extended herein by developing an optimization framework for probe placement and frequency-dependent mode selection. We derive a noise-aware mean-squared error (MSE) objective that enables greedy mode selection and guides probe placement. Since MSE depends on the covariance of mode coefficients, a low complexity model is presented that captures the relative power of mode coefficients, based on knowledge of actual or possible slots on the cavity boundary and a numerical description of cavity modes. An iterative greedy algorithm is then proposed to jointly optimize probes and modes; when probe locations are fixed, the same framework reduces to a single greedy mode optimization step. Numerical and experimental results show that optimizing the mode sets substantially reduces reconstruction error, even for suboptimal probe placements.
Standardized test procedures for radiated emission measurements in anechoic chambers face challenges when applied to electrically large devices under test. The established maximum value method suffers from poor reproducibility and comparability due to undersampling of the highly inhomogeneous radiation patterns. This work proposes a robust statistical procedure based on the spatial root-mean-square value of the electric field strength to assess radiated emissions from electrically large unintentional radiators in anechoic chambers with fewer angular samples.
This article investigates the radiated emissions characteristics of large-scale +/- 500 kV modular multilevel converter-based high-voltage direct current (MMC-HVDC) systems in comparison with conventional line commutated converter-based high voltage direct current (LCC-HVDC) systems. The analysis reveals that the high switching frequency of IGBT devices, the massive number of submodules, and the commutations within each valve control period introduce complex disturbance sources that significantly affect high-frequency (HF) radiated electric field. To address these challenges, a wideband equivalent circuit model is developed from a multilevel perspective spanning "submodulevalve group-system," capturing MHz-level disturbance sources and their conductive and radiative propagation paths. A corresponding radiated emissions simulation method is further proposed and validated, enabling accurate prediction of HF current components. Simulation results show that the switching superposition effect arising from submodule commutations is an important factor influencing MHz-level radiated emissions. The findings provide theoretical foundations and methodological support for electromagnetic compatibility design and radiated emissions mitigation in large-capacity MMC-HVDC projects.
To improve the accuracy of equivalent radiation source modeling, this article proposes a phaseless single-plane iterative radiation source reconstruction method based on adaptive singular value decomposition (ASVD) with Tikhonov regularization. Unlike traditional singular value decomposition, which suffers from the limitations of one-time matrix decomposition and high sensitivity to noise, the proposed ASVD approach performs iterative decomposition of the transfer coefficient matrix and dynamically updates the singular values using a threshold determined by the maximum singular value. This strategy effectively suppresses numerical errors and enhances reconstruction stability. Furthermore, five accelerated iterative schemes-Nesterov accelerated gradient, fast iterative shrinkage threshold algorithm, preconditioned conjugate gradient, alternating direction multiplier method, and Barzilai-Borwein-are incorporated to further improve computational efficiency. Both simulation and experimental results verify that the reconstructed sources achieve a relative error below 3% compared with measurements, demonstrating the proposed method's high accuracy and efficiency.
