This work presents a low-cost, easy-to-fabricate, kirigami-inspired deployable MIMO antenna operating at 2 GHz, featuring dual polarization and angular pattern diversity. Inspired by the Yagi antenna design, the proposed MIMO antenna comprises two orthogonally oriented antennas, each consisting of a rectangular radiating element (monopole) and one parasitic rectangular strip (director). These are integrated onto a foldable, staircase-shaped Kirigami structure made of polyethylene terephthalate (PET) sheet. Each antenna includes a reflector attached to the Kirigami structure’s back. Operating at 2 GHz, the antenna achieves a measured peak gain of 9 dBi with a -10 dB impedance bandwidth of 630 MHz (1.82 GHz to 2.45 GHz). Mutual coupling between the radiating elements is maintained below -15 dB, and the total efficiency exceeds 80% across the operating band. The antenna system meets MIMO specifications, demonstrating an Envelope Correlation Coefficient (ECC) below 0.0025 and a Channel Capacity Loss (CCL) below 0.35 bps/Hz. This combination of high performance, low cost, and ease of fabrication makes the proposed design a promising candidate for diverse microwave applications.

Reliable assessment of the global corrosion state of metallic plates, including thickness loss and surface inhomogeneity, is critical in many applications, such as the shipbuilding industry. Nondestructive testing has attracted considerable attention due to its ability to detect material degradation at an early stage without damaging structures, thereby ensuring integrity and extending service life. Among recent advances, guided ultrasonic waves have shown strong potential for evaluating statistical corrosion characteristics, including the mean and standard deviation of thickness reduction. However, their practical use remains challenging because of the complex interaction between wave propagation and spatially varying thickness profiles. This work presents a novel approach for accurate estimation of global corrosion parameters based on behavioral surrogate modeling. The corrosion profile is represented as a stochastic field, while guided wave responses are obtained through numerical simulations within a sensor network. An inverse model is then constructed to map wave response features onto the corresponding corrosion parameters. A key difficulty lies in the non-uniqueness of the problem, as a single set of corrosion parameters may correspond to infinitely many stochastic field realizations. To address this issue, we extract a statistical representation of the guided wave responses in the form of a multivariate normal distribution, characterized by its mean vector and covariance matrix. This formulation allows individual sensor signals to be interpreted as realizations of the same distribution. Prediction accuracy is further enhanced by aggregating responses from multiple independent stochastic fields, which improves the robustness of feature extraction. Extensive validation studies confirm the high predictive accuracy of the proposed method and its advantage over alternative feature definitions that treat response characteristics, such as wave packet position, width, and amplitude, as statistically independent. Moreover, the approach significantly outperforms conventional time-of-flight-based identification techniques derived from dispersion curve analysis.

An SIW-based dual-circularly polarized (CP) metasurface (MS) antenna with high gain and high isolation is proposed in this article for millimeter-wave (MMW) applications. The orthogonal higher order modes (HOMs) TE230/TE(320 )are utilized to generate CP. A dual-CP cavity-backed antenna with curved slots serves as the primary radiator, whereas a 4 x 4 square patch MS is employed as a superstrate. The characteristic mode analysis (CMA) of the antenna is utilized to identify and align the HOMs with a 90 degrees phase difference to enable dual-CP radiation. CMA is also applied to the MS to synthesize modal phase differences of the primary antenna's modes, enabling broadband phase compensation. The MS height is optimized to induce a 180 degrees phase-shifted reflection, canceling undesired coupled fields and enhancing port isolation. Additionally, the symmetric electric field distribution across the MS improves broadside gain and suppresses cross-polarization. The fabricated antenna achieves a peak gain of 13.0 dBic, isolation >38 dB at 30.2 GHz, and broadside cross-polarization not exceeding 30 dB with a maximum cross-polarization of less than -18 dB. These results confirm the antenna's potential for advanced MMW applications requiring efficient, dual-CP performance.

This paper presents a high-performance dual-polarized super-wideband (SWB) antenna that employs optimized slots and an Artificial Material (AM) array to achieve remarkable enhancements in both bandwidth and gain. The proposed design features a compact Vivaldi configuration with integrated arrow-shaped slots, enabling SWB operation while maintaining a low-profile size of 0.66 x 0.49 x 0.01 lambda(3) . A constrained trust-region gradient-based optimization is used to refine the slot geometry and antenna parameters, achieving an operating range from 8.4 to 42.8 GHz with a stable gain of approximately 10 dBi at frequencies above 17 GHz. The incorporation of AMs further enhances the gain and extends the bandwidth. The AM antenna, with dimensions of 0.94 x 0.49 x0.01 lambda(3), achieves peak gains of 13.3 dBi at 24 and 38 GHz and covers a wide operating range of 8.3-45 GHz. Dual polarization is realized by employing two orthogonally oriented Vivaldi elements, resulting in a high polarization isolation exceeding 25 dB across the entire operating bandwidth. The experimental results demonstrate excellent agreement with the simulation-based findings. Featuring SWB performance, high gain, strong polarization isolation, polarization diversity, and a compact low-profile structure, the proposed antenna is well suited for 5G millimeter-wave (MMW) and satellite remote sensing applications.

This paper presents a geometrically simple substrate-integrated waveguide (SIW)-based self-quintuplexing antenna (SQA). The antenna structure consists of a modified X-shaped slot embedded in the SIW cavity. The antenna operates at five distinct frequencies: 3.28 GHz, 3.63 GHz, 3.89 GHz, 4.68 GHz and 5.62 GHz, while ensuring isolation better than 22.6 dB across the entire range. An equivalent circuit model is provided to validate the proposed model. The fabricated prototype, realized on a Rogers RT/5880 substrate, shows close alignment between simulated and measured outcomes. Furthermore, the measured antenna gains are 4.1 dB, 3.95 dB, 4.23 dB, 4.35 dB, and 5.7 dB at the respective resonant frequencies. The proposed device features a simple design, good isolation, and enhanced gain, making it suitable for a wide range of L, S, and C-band applications.

This paper presents a novel and compact decoupling methodology for ultra-compact planar antenna arrays with center-to-center spacing smaller than 0.3λ. In such densely packed configurations, strong multi-directional mutual coupling significantly degrades the array performance. To address this challenge, a comprehensive decoupling strategy is proposed for a 2 × 2 array operating in the 5.8–6 GHz band. The approach integrates three coordinated mechanisms: I-shaped metallic strips for coupling-path cancellation, a dumbbell-shaped defected ground structure (DGS) for ground-current suppression, and dielectric perturbation for reducing substrate-mediated coupling. The proposed structure achieves effective two-dimensional decoupling while maintaining a low-profile and easy-to-fabricate configuration. By performing a systematic parametric study and Genetic Algorithm (GA) the structure is optimized in the multidimensional design space. Both simulated and measured results demonstrate that the proposed method successfully reduces the transmission coefficients between array elements to below −20 dB across the target frequency band, satisfying the critical design requirement. The achieved isolation enhancement does not degrade the impedance matching, and the measured radiation patterns further demonstrate that the proposed decoupling structure improves the realized gain by 2.85 dB and 4.34 dB at the lower and upper edges of the operating frequency band, respectively.

A novel dual-functional frequency-selective metasurface (MS) with low backscattering is proposed by integrating a reflective cross-polarization (LRCP) converter with a bandpass frequency-selective surface (BP-FSS). The proposed MS comprises three layers, with a cross-dipole FSS in the middle layer that provides high-efficiency bandpass transmission. The thickness, inter-gap, and interlayer spacing of the diagonally oriented metallic strip gratings (MSGs), printed on the top and bottom layers, are meticulously optimized to achieve efficient LRCP conversion while maintaining high transmittance transparency for the out-of-band frequencies. The proposed MS simultaneously exhibits two independent functionalities: co-polarized bandpass transmission from 3.74 to 5.12 GHz with a simulated insertion loss of above-1 dB, and LRCP conversion from 5.72 to 7.33 GHz with a polarization conversion ratio (PCR) exceeding 0.99. The design concept is validated using two equivalent circuit models (ECMs) corresponding to the BP-FSS and the cascaded MS configuration, as well as through measurements of the fabricated prototype comprising 16 & times; 16 unit-cell arrays arranged in a chessboard configuration. Both the ECM predictions and measurements show good alignment with the full-wave simulation findings. In addition, the measurement results showcase significant radar cross-section (RCS) reduction across the frequency range from 5.94-7.21 GHz. Unlike conventional LRCP-MSs, the proposed work eliminates the need for a complete ground plane, and hence, its dual functionality makes it a promising candidate for Radome and communication platforms.

Achieving multifunctional features within a compact, single-layer metasurface (MS) without employing biasing networks or multilayer configurations remains a significant challenge in recent research. To address this, a novel multifunctional bi-anisotropic omega metasurface (BiOMS) is proposed that simultaneously achieves polarization transformation in both reflection and transmission, along with bandstop, bandpass, absorption, and beam-splitting functionalities, using two distinct geometries on opposite sides of the same substrate. When a y-polarized electric-field (E-t) impinges from +Z-side (forward-direction, (E-i(y))(+Z)), the MS exhibits reflective cross-and circular-polarization conversion, including bandpass, bandstop, and beam-splitting features. The MS exhibits similar characteristics, while additionally providing absorption and partial transmissive cross-polarization (TCP) conversion, when excited by x-polarized wave ((E-t(x))(+Z). Under-Z-side (backward-direction) illumination ((E-t(y,x))(-Z)), the structure functions as a beam-splitter and retains bandpass and bandstop functionalities at different frequency bands without altering the E-t polarization. For both ((E-t(y))(+/- Z))-sided excitations, the MS maintains symmetric transmission, featuring a low-loss bandpass and a wide bandstop window. The proposed concept is validated through an equivalent circuit model (ECM), current distribution, electric-field analysis, and prototype measurements, all of which show good agreement with simulations. Compared with existing BiOMSs, the proposed work uniquely combines multiple functionalities, making it attractive for radar-cross-section, absorption, filtering, and beam-splitting applications.

The small physical size of high-frequency components is an essential prerequisite in an increasing number of applications. While size reduction of microwave circuits can be achieved with a suitable choice of architecture, ensuring the best available compromise between electrical parameters and size necessitates meticulous adjustment of all system dimensions. Due to the complexity of electrically small structures, parameter adjustment should be executed globally. Nevertheless, it is an expensive endeavor because it entails a large number of system evaluations executed through electromagnetic (EM) simulation. This article introduces an innovative approach to expedited miniaturization of microwave passives, in which the primary goal is miniaturization, while the conditions imposed on performance metrics are treated as constraints. Our framework operates in three stages, including parameter space pre-screening, machine-learning (ML)-driven globalized search, and gradient-based fine-tuning. The first stage involves random sampling aimed at identifying the most promising parameter space region. The ML process launched therein employs simplex-based regression models, the use of which contributes to a competitive computational efficiency. On the other hand, the final tuning improves design quality at a low cost, as it is realized as a local optimization process. The performance of the suggested algorithm is illustrated using two planar devices and demonstrated to be superior to several benchmark methods. The major advantages of our technique include low running costs, equivalent to only a few dozen full-wave analyses, and straightforward implementation.

This work presents a geometrically simple topology for developing an ultra-wideband directional coupler with improved coupling and directivity. A short-ended coupled-line structure is used to achieve an ultra-wideband, tightly coupled symmetric three-section coupler using the microstrip line technology. The proposed design demonstrates an explicit improvement of approximately 1.2 dB in coupling compared to conventional multi-section directional couplers. Calculated, simulated, and measured responses validate the effectiveness of the proposed configuration in terms of low-ripple coupling bandwidth, low insertion loss, and improved directivity performance compared to respective responses of the conventional structure. Couplers featuring a higher number of sections to implement different bandwidths and couplings can be fabricated using the presented structure due to its transmission line-based approach. A prototype of the three-section directional coupler with coupling of 7.6 dB, 8.1 dB, and 8.3 dB and corresponding bandwidths of 104%, 123% and 133% is designed, fabricated, and measured. The experimental results confirm that the coupler can reliably achieve higher coupling with ultra-wideband response from 0.75 GHz to 3.75 GHz (5:1) with 8.3 ± 1.4 dB (ripple). Additionally, the design yields promising performance with return loss > 16 dB, isolation > 20 dB, a phase difference of 90 ± 4°, and directivity > 30 dB, and the maximum circuit size is 0.067λ02. This work aligns with SDG 9: Industry, Innovation and Infrastructure by advancing high-performance microwave components that support efficient, reliable, and scalable communication infrastructure.

This article presents the development of a novel cavity-backed dual-feed miniaturized antenna with an intrinsic self-diplexing property for millimeter-wave dual-frequency 5G applications. The proposed device is designed using a rectangular substrate integrated cavity loaded with a K-shaped slot, which is powered by two 50 Ω feed lines to enable radiation at two millimeter-wave bands. The designable parameters of the K-shaped slot are used for independent frequency tuning. The impedance matching between the rectangular cavity and the feed lines is achieved by utilizing the inline feeding method. A basic design process, working approach, radiation methodology, and equivalent circuit analysis are all elaborated in depth. Finally, a prototype of the miniaturized antenna diplexer operating at 22 GHz and 27 GHz is fabricated and experimentally validated. The proposed antenna diplexer has a compact footprint of 0.448λg2. The antenna prototype exhibits a return loss of − 25.7 dB (− 17.8 dB) and isolation greater than 31.2 dB (25.1 dB) at 22 GHz (27 GHz). The antenna prototype achieves EM and measured realized gains exceeding 4.38 dBi (4.23 dBi) and 4.1 dBi (4.27 dBi), with efficiencies better than 88% and 83% at 22 and 27 GHz, respectively. The proposed antenna diplexer offers a low cross-polarization level well below − 30 dB at both millimeter-wave bands. Additionally, the antenna enables wide frequency tunability ranging from 21.85 to 22.15 GHz and from 21.85 to 22.15 GHz around 22 and 27 GHz, respectively, which makes it a suitable device for dual-band millimeter-wave systems. Furthermore, the proposed design supports Sustainable Development Goal (SDG) 9 by advancing innovative and energy-efficient communication infrastructure for future millimeter-wave 5G applications. Furthermore, the proposed design supports Sustainable Development Goal (SDG) 9 by advancing innovative and energy-efficient communication infrastructure for future millimeter-wave 5G applications.

This research proposes a novel methodology for rapid multi-objective (MO) electromagnetic (EM)-driven design of antenna systems. Our approach integrates machine learning (ML) with neural network metamodels. At each iteration, the surrogate, optimized using an MO-oriented evolutionary algorithm, generates multiple candidate Pareto-optimal designs. These solutions are further refined locally using approximate response Jacobians estimated from the surrogate model. The metamodel itself is continuously updated with EM data collected during the optimization process. By combining surrogate-driven exploration with gradient-based refinement, the method accelerates convergence and enables more accurate identification of Pareto fronts compared to conventional techniques. The framework has been validated on two planar antennas, showcasing excellent cost efficiency: the mean MO expenses are under 300 EM simulations, corresponding to roughly 80

Surrogate modeling is a key enabler in antenna engineering, facilitating efficient simulation-based design, including parametric optimization. However, developing reliable behavioral models remains challenging, with most existing methods focusing solely on electrical characteristics (e.g., reflection coefficients) or integral performance metrics (e.g., gain). Modeling full radiation patterns-critical for practical antenna design-has received limited attention in the literature. This article presents an innovative and comprehensive modeling framework for antennas based on customized deep neural networks (DNNs). The proposed models capture both electrical responses and radiation patterns across a wide frequency spectrum, various antenna dimensions, and diverse substrate parameters, making them suitable for solving design tasks. Frequency-domain characteristics (e.g., reflection, gain) are accurately modeled using recurrent neural networks (RNNs) incorporating long short-term memory (LSTM) layers, treating frequency as a sequential input. Radiation patterns are learned through decoder-style convolutional neural networks (CNNs), with deconvolutional (or upsampling) layers inspired by the design principles of Transformer decoders, to provide reliable predictions over wide frequency ranges and azimuth angles as functions of input geometry parameters. Extensive validation involving two planar antennas demonstrates the exquisite accuracy of the presented surrogates compared to several deep learning approaches. Practical design applications, including center frequency tuning, gain maximization, bandwidth widening, and front-to-back ratio improvement, confirm the models' effectiveness in real-world optimization scenarios at a moderate computational cost. The methodology paves the way for robust, multiresponse antenna modeling and expands the utility of data-driven surrogates in modern antenna design.

Rigorous optimization methods have become standard practice in antenna engineering, gradually replacing traditional interactive design approaches that relied on parametric studies and engineering intuition. Nevertheless, antenna optimization remains computationally expensive due to its reliance on electromagnetic (EM) analysis. To mitigate this, accelerated strategies have been introduced by limiting the occurrences of EM simulations at the algorithmic level or through surrogate modeling. Multi-fidelity approaches offer another avenue, though existing frameworks typically restrict themselves to just two levels (low and high fidelity). In this work, we propose an innovative model management scheme that adaptively adjusts EM model resolution across a continuous fidelity spectrum. Model selection is guided by the optimization’s convergence status and design quality indicators. The process begins with the lowest usable resolution, which is progressively refined as the optimization approaches convergence and the objective value improves. This strategy lowers computational costs by exploiting faster, lower-fidelity simulations when far from the optimum, while ensuring reliability by incorporating high-fidelity models near convergence. Extensive numerical experiments involving two microstrip antennas showcase the efficacy of the presented framework, showing speedups of exceeding 70

A planar metamaterial lens-based single-element circularly polarized (CP) antenna for millimeter wave (mm-wave) band applications is presented. The proposed antenna consists of a modified patch excited by a single-point-fed coaxial probe and two displaced layers of a novel meta-lens design. The modified structure allows for the simultaneous excitation of orthogonal components with equal magnitudes. To realize the gain enhancement of the proposed design, a novel meta-lens is designed based on meta-atoms of subwavelength size arranged in a disconnected cross-shape repeated pattern. To effectively focus the outgoing CP wave radiated by the antenna, the focal distance is meticulously optimized. Two layers of the same lens are used to enhance the antenna gain. Following a rigorous numerical analysis and optimization, the proposed design is fabricated and experimentally validated. The comparison of the results with the lens and without the lens illustrates that a 4 dB gain improvement is attained with the compact lens configuration. Furthermore, the antenna features a wide impedance bandwidth (S11) from 24 GHz to 31 GHz and the axial ratio (AR) below 3 dB within the same operating band. The proposed design offers multiple advantages, including a simple geometrical configuration, light in weight, and ease of integration due to the planar lens structure. The proposed antenna is suitable for multiple modern communication systems, including short-range radar systems and other line-of-sight mm-wave applications requiring fixed-beam and high data rates.

Assessing corrosion in metallic plates using guided waves remains challenging due to the complex interactions between wave propagation and spatially varying thickness distributions. While conventional approaches typically assume uniform thickness reduction, real corrosion processes lead to stochastic, spatially heterogeneous degradation patterns. This introduces significant uncertainty in the interpretation of measured signals and limits the applicability of classical physics-based identification methods. In this study, a behavioral inverse modeling framework is proposed for the identification of global corrosion parameters, namely the mean thickness reduction and the standard deviation of thickness, in plates with randomly varying corrosion geometries. The corrosion morphology is represented as a stochastic field, and guided wave responses are generated through numerical simulations. Signal processing is performed to extract descriptive features, which are subsequently used as inputs to an inverse regression model that maps signal characteristics to corrosion parameters. The performance of the proposed approach is compared with that of a classical time-of-flight-based identification method derived from dispersion curve analysis. The results indicate that the time-of-flight approach provides mediocre estimates of the mean thickness. Furthermore, this approach relies on assumptions derived for plates of uniform thickness, in which the wave velocity can be related to thickness through dispersion characteristics of guided waves. In the presence of spatially varying thickness, this relationship becomes more complex, and using an effective average velocity alone leads to inaccuracies. The proposed inverse model addresses these limitations by learning the mapping directly from signal characteristics and their extracted multiple features, enabling accurate and robust identification of both the mean thickness and its variability across broad ranges thereof. Its efficacy is corroborated through extensive verification studies and supported by rigorous metrics such as relative prediction error and prediction repeatability for stochastically varying corrosion morphologies. The effects of feature selection on the model’s predictive power are also investigated.

A compact printed monopole antenna with multiple ground plane geometries is presented for multi-protocol wireless and sensing applications. The antenna, with physical dimensions of 25 × 20 × 0.8 mm3 and three different ground plane shapes, operates across the C, X, and Ku bands, with the capability to reject interfering bands. It consists of a circular radiating patch, a coplanar waveguide transmission line, and arrow-shaped slots, designed on a Rogers RO4003C substrate. The antenna, with its original ground plane shape, covers three −10 dB operating bands, including 8.2–10.64 GHz, 12.1–24 GHz, 25.54–30 GHz. Although the simulation indicates rejection bands at 10.65–12.1 GHz and 24–25.54 GHz, the measured results show an improved impedance match. After modifying the ground plane, the second configuration operates exclusively within the X band. The third design is optimized for the C and Ku bands and rejects the X band frequencies. Additionally, this work introduces a miniaturized antenna sensor suitable for non-invasive hydration monitoring via sweat analysis. By integrating the sensor into athletes’ shoes, it enables continuous, real-time monitoring during exercise, providing coaches with valuable insights to optimize hydration strategies and training intensity. The antenna sensor demonstrates a significant frequency shift in response to varying sweat concentrations, with a sensitivity of 20.57 MHz/1, allowing effective tracking of hydration levels throughout physical activity.

This paper presents and experimentally validates a multifunctional anisotropic metasurface (MS) that exhibits linear-to-linear (LTL) reflective cross-polarization (RCP) and circular polarization conversion (CPC) responses across multiple frequency ranges. The proposed MS integrates a single-layer design consisting of 45 degrees oriented truncated circular ring and an hourglass-like structure, providing four LTL-RCP conversion bands (2.41 - 2.48, 3.16 - 3.46, 5.80 - 6.59, and 11.64 - 13.75) GHz with polarization conversion ratio (PCR) exceeding 95%, and seven CPC bands (2.33 - 2.38, 2.55 - 2.99, 3.69 - 5.37, 7.13 - 10.92, 13.95 - 14.25, 14.93 - 15.17, and 15.31 - 15.56) GHz with the axial ratio (AR) less than 3 dB, covering the S-, C-, X-, and Ku-bands. By only adjusting the angle of the hourglass structure at 15 degrees within the same structure, LTL-RCP with PCR greater than 95% is realized in the frequency bands from 13.71 to 14.64 GHz, along with <= 3 dB of AR is achieved at four additional CPC bands at (2.20 - 2.36, 2.95 - 8.41, 10.84 - 12.30, and 14.84 - 15.06) GHz. The polarization conversion mechanisms are explained through detailed surface-current distributions analysis and full-wave electromagnetic (EM) simulations. Furthermore, the proposed concept is validated through the measurements of the fabricated prototype consisting of 8 & times; 13 -unit cells. The measurement results are in excellent agreement with the simulation predictions. The proposed MS design demonstrates a simple yet effective solution for achieving tunable multi-band polarization manipulation without structural modification, making it suitable for advanced EM communication applications.

This communication presents an analytical and numerical investigation of the effective permittivity of dielectric slabs perforated with elliptical and capsule-shaped cylindrical microstructures arranged in square and triangular lattice configurations. Based on a quasi-static approximation, the 2-D Mori-Tanaka mixing model is employed to calculate the anisotropic effective permittivity tensor by incorporating direction-dependent depolarization factors for both parallel and perpendicular field orientations. An elliptical-based microstructure is first analyzed, followed by the introduction of a capsule-shaped alternative that simplifies fabrication while maintaining comparable electromagnetic (EM) properties with over 95% accuracy. The EM behavior of both geometries is validated through full-wave unit cell simulations across a broad frequency range (5-48 GHz) and porosity levels. The results demonstrate excellent agreement between analytical predictions and full-wave simulations, particularly up to 70% porosity, beyond which capsule structures exhibit superior stability and modeling consistency even at extreme porosities (82%, 97%). Additionally, the model captures angular field misalignments up to 20 degrees, preserving accurate predictions of permittivity. The calculated effective properties can serve as inputs for EM simulation tools designed for homogeneous media, enabling their application in various practical EM designs.

Surrogate modelling has become increasingly important in microwave engineering. Fast metamodels—particularly behavioral ones—are widely employed to accelerate design tasks such as parametric optimization by replacing costly full-wave electromagnetic (EM) simulations. However, constructing reliable surrogates remains challenging due to the strong nonlinearity of circuit responses and the curse of dimensionality. The difficulty is especially pronounced in design-oriented modeling, where validity must be ensured across wide ranges of parameters. This research introduces an innovative modeling procedure that combines dimensionality reduction with spatial domain restriction to reduce the training data acquisition cost while enhancing predictive accuracy. Dimensionality reduction is achieved using fast global sensitivity analysis, which determines the parameter-space directions with the strongest impact on the system’s frequency characteristics. These vectors define the reduced domain, which is additionally confined by means of principal component analysis (PCA) of pre-screened high-quality designs. As a result, the surrogate is concentrated on the relevant design space subsets, ensuring suitability for practical design tasks while maintaining high accuracy. The proposed methodology has been extensively validated against state-of-the-art benchmarks, demonstrating both competitive precision and significant efficacy gains. Its design readiness has been shown through practical applications, specifically, in EM-driven circuit optimization under varying specification scenarios.