Abstracts:In millimeter-wave (MMW) imaging, the focusing performance of sparse multiple-input-multiple-output (MIMO) arrays is susceptible to various noise interferences during the propagation of electromagnetic waves. In addition, the sparse array imaging based on the traditional compressive sensing (CS) algorithms suffers from low computational efficiency due to the numerous iterations, especially in 3-D imaging scenarios. To address these limitations, an interpretable self-supervised network with a parallel structure is proposed for fast enhanced imaging via sparse cross MIMO arrays. The designed interpolation-free MIMO operators are embedded into a generalized complex-valued framework unfolded by the alternating direction method of multipliers (ADMMs) to form the encoder of reconstruction branch, avoiding large-scale MIMO sensing matrix operations and allowing for the fast 3-D imaging processing of sparse MIMO data. The decoder is constructed to map the data from image domain to signal domain, thereby producing the pseudo-labels for self-supervised learning and eliminating the need for costly data annotation. The enhancement branch employs an optimization module based on the complex-valued total variation (CTV) and $l_{1}$ norms to guide the network training, thus improving the noise immunity and imaging performance of the network. The simulations and experiments show that the proposed network can obtain better-focused 3-D imaging results in a short time.

Abstracts:In this communication, a wideband and wide axial ratio bandwidth (ARBW) metantenna based on magnetoelectric (ME) dipole is proposed for Ka-band satellite communications (SATCOMs). The ME dipole source incorporates slotting in the planar radiator to achieve miniaturization and is integrated with a linear-to-circular polarization converter (LTCPC) operating within the matched band to enable circularly polarized (CP) radiation. To evaluate its performance, a $4 \times 4$ fixed-beam array was designed, fabricated, and tested. The simulation results show an ARBW below 3 dB spanning 26.85–40.5 GHz (40.5%) and an impedance bandwidth covering 26.55–40.1 GHz (40.7%). Measured results agree well with simulations. The radiation characteristics demonstrate high polarization isolation, and the co-polarization type can be switched by rotating the LTCPC. Upon integration with a power divider to form a broadside array, the antenna achieves a stable gain up to 18.9 dBic and supports a 40° beam-scanning angle in phased-array configurations. Owing to its wide operating bandwidth and switchable CP type, this design offers flexible applicability in Ka-band low Earth orbit (LEO) SATCOMs.

Abstracts:Adversarial attack methods against synthetic aperture radar (SAR) automatic target recognition (ATR) can disrupt adversarial reconnaissance and protect valuable targets. In recent years, several adversarial attack algorithms have been proposed and achieved promising results in generating adversarial examples in the digital domain. Extending these attacks from the digital domain to the physical domain remains a critical challenge. In this work, we propose a physically feasible adversarial attack method based on a time-varying metasurface. A novel phase modulation technique is introduced to embed the global perturbation information into the radar echo signal, which will be transformed into adversarial examples by the adversary’s SAR imaging processor. The resulting adversarial examples can mislead the neural network-based SAR ATR system, thereby achieving the intended adversarial attack. The effectiveness of the proposed method is validated through theoretical analysis and simulations, and its feasibility is further assessed under different radar operation bands.

Abstracts:Operational frequency selection is a critical spot for ensuring high-frequency (HF) communication’s optimal quality, thus accurate prediction of usable frequency is essential to HF communication, which has attracted much attention. Therefore, this article proposed a high-precision fusing prediction algorithm based on entropy weight theory to enhance the prediction accuracy and stability of maximum usable frequency (MUF). It integrates the algorithms proposed by the Istituto Nazionale di Geofisica e Vulcanologia (abbreviated to “INGV”) with the Lockwood algorithm through a weighted fusion, resulting in a new prediction algorithm. To assess the reliability of the proposed algorithm, we collected sounding data from the ChangchunJinyang oblique-sounding circuit and the Beijing vertical-sounding station. Compared with the INGV and Lockwood algorithms, the proposed algorithm improves the prediction accuracy by 5.97% and 2.33%, respectively. Specifically, compared to the INGV algorithm, the proposed algorithm reduces the root mean square error (RMSE) by 5.47%, 0.56%, 8.04%, and 9.92% in spring, summer, autumn, and winter, respectively. Similarly, compared to the Lockwood algorithm, the proposed algorithm achieves improvements of 0.82%, 2.20%, 2.83%, and 1.80% in the same four seasons. These results indicate that the proposed method offers notable advantages in MUF prediction, providing an effective channel modeling algorithm for HF communications.

Abstracts:In this article, a method for generating orbital angular momentum (OAM) waves of arbitrary order using point-atom-based holographic mirrors is presented. For efficient modulation, a two-stage complex-amplitude modulation method is proposed based on two-mirror architecture. By utilizing stationary phase method in conjunction with Gerchberg–Saxton (GS) algorithm, the Gaussian beam is converted to the Bessel OAM beam. To achieve high mode purity at arbitrary orders, the impact of the spacing between point-like atoms on beam quality is investigated, providing effective atoms-interval arrangements suitable for different OAM mode orders. Based on this analysis, mirror modulators are designed to generate OAM waves of arbitrary order, achieving mode purities of up to 98% for mode index $l \le 5$ and purities above 95% for $l =10$ and 20. Simulations are performed to evaluate the proposed method, indicating good agreement with theoretical calculations. Subsequently, a prototype for $l =4$ is fabricated and measured. The results demonstrate successful OAM beam generation, with the mode purity reaching 90% over a wide fractional bandwidth of 20.9% spanning from 30 to 37 GHz, with 83.9% energy conversion efficiency (CE). These results demonstrate the effectiveness of the proposed method and highlight its potential for high-power radar imaging and high-speed communication systems.

Abstracts:Bistatic inverse synthetic aperture radar (B-ISAR) has garnered significant attention for its ability to collect target echo data from different observation views and to address monostatic-ISAR device’s inherent constraints. However, the unique time-varying bistatic angle (TBA) is coupled with the rotation of the target; the migration through the resolution cell (MTRC) of scattering points becomes more severe and decreases the B-ISAR image quality. Meanwhile, the geometric distortion caused by the TBA leads to target deformation in the final B-ISAR image. In this article, a signal model for moving targets under the bistatic configuration is developed, explicitly incorporating the TBA. The effects of MTRC and geometric distortion on B-ISAR imaging are thoroughly analyzed. Furthermore, the impact of distortion coefficient estimation errors on geometric correction is investigated, and the tolerance range of TBA estimation errors is quantitatively analyzed. Then, a dual-generalized keystone transform and fractional Fourier transform (DGKT-FrFT) method is employed in the echo domain to eliminate the high-order MTRC. Considering the distortion term, directly applying the fast Fourier transform (FFT) in the azimuth dimension will fail to achieve a high-quality B-ISAR image. To correct the geometric distortion, the distortion coefficient is estimated by location information added nonlinear least square algorithm (LINLSA). The proposed method can succeed in obtaining focused and undistorted images. The experimental results confirm the proposed method’s effectiveness.

Abstracts:In current antenna design practice, full-wave electromagnetic (EM) analysis has established itself a standard method of choice. It enables accurate evaluation at the expense of being CPU-intensive. Consequently, EM-driven design procedures, such as optimization-based parameter refinement or robust design, are time-consuming. Surrogate modeling methods have been widely used to mitigate these issues. Notwithstanding, the construction of reliable behavioral models poses significant challenges by itself, with the major difficulties pertaining to the curse of dimensionality and nonlinearity of antenna characteristics. In this article, we introduce a novel methodology for data-driven modeling of antenna structures. It incorporates two fundamental mechanisms aimed at improving the surrogate’s accuracy and at lowering the computational cost of training dataset generation. The first is dimensionality reduction enabled by a fast global sensitivity analysis (FGSA). GSA is used to establish the key directions in the parameter space, being the main source of antenna response changes. The second is spatial confinement of the model domain based on stochastic prescreening and spectral analysis of the random sample subset selected based on design quality criteria. Comprehensive verification demonstrates competitive accuracy of the surrogates constructed using the proposed method and their superiority over several state-of-the-art modeling techniques. The design usefulness of the developed approach is corroborated by the application case studies, specifically antenna optimization using a variety of design specifications pertaining to target operating bandwidths and material parameters of the substrate. Selected designs are experimentally validated.

Abstracts:Aircraft-like objects exhibit significant electromagnetic scattering in the meter-wave frequency range due to their inherent resonance effects. Traditional scattering cross section (SCS) reduction techniques, such as shape optimization and radar-absorbing materials (RAMs), often require large structural modifications to control meter-wave scattering, which inevitably affect the aerodynamic performance. This article proposes a novel reconfigurable capacitive loading technique based on characteristic mode (CM) cancellation. By strategically placing a small number of capacitive loadings on the surface of aircraft-like objects, dominant scattering modes are induced to cancel each other. The alternating direction method of multiplier (ADMM) is introduced to transform the nonconvex scattering optimization problem into a two-step iterative process, which is incorporated with the CM analysis (CMA) to efficiently determine the near-optimal capacitive loadings. An electronically reconfigurable loading network is then implemented to dynamically control the states of the capacitive loadings for wideband SCS reduction. Numerical results for a simplified aircraft-like object show a maximum monostatic SCS reduction of 44.0 dB and a maximum 10-dB SCS reduction fractional bandwidth of 9.5% over 84.0–120.0 MHz. A scaled prototype is designed, fabricated, and measured, which demonstrates the effectiveness of the proposed approach.

Abstracts:To address the stringent requirements of full coverage and ultrahigh data rates in next-generation mobile communications, it is essential to leverage the coexistence of multiple radio frequency (RF) systems operating in well-separated frequency bands within precisely defined scenarios. In this context, an investigation of frequency- and environment-dependent channel characteristics by exploring the spatial and temporal correlations of multipath channels across different frequency bands and different environments is imperative. This article introduces a structural channel similarity index measure (CSIM) that holistically evaluates multiple multipath parameters between two channels, including amplitude, phase, delay, angle of arrival (AoA), and angle of departure (AoD). Based on extensive field measurement campaigns and ray tracing simulations conducted across both centimeter-wave (cmWave) and millimeter-wave (mmWave) bands in typical indoor and outdoor scenarios, the proposed CSIM is proven to effectively measure similarity from specific dimensions as well as the statistical distributions, and the similarities between channels across different frequencies and different environments are presented. Moreover, the feasibility of out-of-band information-assisted beam search, enabled by cross-band channel similarity, is also validated.

Abstracts:This article introduces a hybrid approach combining ray-tracing (RT) and full-wave numerical analysis to evaluate and optimize reconfigurable intelligent surfaces (RISs) in deterministic channels. Unlike traditional methods relying on statistical models, our approach integrates RT for deterministic channel mapping and finite element method (FEM)-based simulations to assess RIS impacts in realistic, site-specific scenarios. A nondiagonal impedance matrix is formed to link channel transfer functions to RIS tunable load impedances, accounting for mutual coupling between unit cells and enabling precise optimization for enhanced communication performance. A varactor-based RIS operating in the sub-6 GHz frequency band is designed and fabricated, with extensive simulations and real-world measurements conducted in single-input single-output (SISO) and single-input multiple-output (SIMO) scenarios. A measurement-based closed-loop optimization process further refines RIS configurations. Results demonstrate substantial signal strength improvements, with strong agreement between simulations and measurements, confirming the practicality and effectiveness of the proposed method for advanced RIS-enabled communication systems.