Accurate ocean ducting environment sensing is crucial for radio communication. Conventionally, ducting profile inversion relied heavily on radar sea returns. Though feasible, this active approach suffers from an operational drawback, which severely impedes practical deployment. Global positioning system (GPS) has revolutionized ocean monitoring through its inherent global coverage, all-weather capability, and high precision. GPS signals scattered from the sea surface provide unique yet underutilized information on atmospheric environments. Although duct-induced modulations encode key ducting parameters, direct extraction from these signals poses intractable nonlinear inverse problems due to complex signal-refraction interactions. Leveraging the unparalleled capability of deep learning (DL) to approximate highly complex, nonlinear mappings, this study presents a Densely Connected Perception Framework for Atmospheric Ducting Retrieval via GPS Scattered Signals (DenseRefrception). The core breakthrough resides in constructing a sophisticated architecture. This framework innovatively integrates dense connectivity to extract discriminative features from scattered GPS signal power while incorporating multihead attention and fully connected networks (FCNs) to establish high-dimensional nonlinear mappings. Through systematic training of a unified multidimensional duct parameter learning mechanism, it precisely constructs highly nonlinear functional relationships. By transforming passively collected GPS signals into actionable refractivity intelligence, the DenseRefrception represents a shift toward persistent and clandestine monitoring of the oceanic environment, significantly enhancing situational awareness for maritime platforms.

Highlights What are the main findings? What is the implication of the main finding?Highlights What are the main findings? What is the implication of the main finding?Abstract This paper proposes a hybrid Vector Radiative Transfer Shooting (VRT-S)-Bouncing Ray (BR) method, referred to as the VRT-S-BR method, for predicting composite electromagnetic scattering from targets above vegetation-covered rough surfaces. In this proposed framework, the vegetation layer is modeled as a stratified random medium and incorporated into the BR solver through VRT-S-derived amplitude modulation and deterministic phase compensation. Specifically, an offline database of vegetation-induced complex reflection coefficients is first generated using the VRT-S model over a set of incidence angles. During the BR ray-tracing process, these coefficients are used to replace the conventional Fresnel reflection terms on a per-interaction basis, thereby accounting for vegetation-induced attenuation and coherent scattering effects. In addition, a facet-dependent phase compensation scheme is introduced to describe propagation-path variations of individual rays through the vegetation canopy, avoiding the empirical random phase perturbation used in previous hybrid models. The proposed method is validated against field-measured backscattering data over natural grassland, achieving root mean square height (RMSE) values of 1.82 dB and 3.10 dB for horizontal-horizontal (HH) and vertical-vertical (VV) polarizations, respectively. Numerical results further demonstrate the capability of the method to characterize target-vegetation coupled scattering under different percentages of vegetation cover, vegetation heights, terrain backgrounds, and bistatic observation geometries.

In this letter, a method is proposed to obtain the range-Doppler map in cases of antenna-target encounter based on near-field time-domain physical optics (TDPO) method. First, the TDPO method is used to simulate the dynamic scattering echo data during the antenna-target encounter. Then, based on the electromagnetic scattering field expression, the signal processing formula for range-Doppler map simulation is derived. In this letter, modulated Gaussian pulse is used as the transmitting signal, and the corresponding signal processing method is designed, including demodulation, range compression, and slow time fast Fourier transform (FFT). This letter studies the effective combination of the electromagnetic scattering algorithm and signal processing algorithm, and proves its feasibility. The proposed method can be used to quickly generate range-Doppler map data and has significant engineering application value. Finally, some simulation examples are given to verify the feasibility and accuracy of the method.

This study proposes the multicharacteristic sea clutter deep learning (MuSC-DL) unified prediction model to resolve multidimensional backscattering challenges in large-scale marine scenarios. Transcending the constraints of conventional methodologies through DL fusion with multisource heterogeneous parameters, the framework innovatively incorporates spectral knowledge embeddings to holistically represent sea surface morphological physics. Crucially, it achieves multiaspect joint prediction of backscattering properties beyond single-characteristic modeling. Validated by comprehensive sea clutter measurements across large-scale maritime scenarios, the MuSC-DL model achieves superior precision in multidimensional backscattering prediction. It excels not only in single-property estimation but also in joint multidimensional characterization, demonstrating strong generalization capability under diverse marine conditions. Thus, this model provides an efficient computational tool for sea clutter characterization and offers a practical solution for maritime radar surveillance systems.

In traditional radar cross section (RCS) measurements, the target under test is regarded as a point scatterer, which enables the use of traditional comparison methods for target RCS determination. However, for targets with complex structure or large aspect ratios, the power differences between far- and near-end structure become pronounced. Consequently, the calibration process in traditional RCS measurements requires modification to improve the RCS measurement accuracy. In this letter, we first analyze the measurement errors caused by the position of a target on a turntable, based on the scattering center theory and the shooting and bouncing ray algorithm. Then, a modified Levenberg-Marquardt method is employed to determine the optimal position of the target. Finally, the proposed method is verified by both simulations and measurements. Compared with traditional method, the proposed algorithm can reduce the RCS measurement error by approximately 10%.

This paper proposes a partitioned multilayer shooting and bouncing rays (SBR) method based on Spider-Web Isosurface Sampling (SWIS) modeling, enabling efficient and reliable radar cross section (RCS) calculation for plasma-coated targets. Firstly, the SWIS method extracts the spatial distribution of plasma parameters from flow field data and constructs a structured partitioned multilayer geometric model. Subsequently, an octree-accelerated SBR method is employed to simulate multiple reflections and transmissions of electromagnetic (EM) waves within lossy plasma and compute the RCS. Validation against the multilevel fast multipole method (MLFMM) results demonstrate good agreement in accuracy, while significantly reducing computational resource. The effects of flight altitude and velocity on the RCS of the RAM-C III hypersonic vehicle are further analyzed. Numerical results indicate that the proposed method provides an efficient and reliable solution for EM scattering computation of plasma-coated hypersonic targets.

This paper presents a compact four-beam dual-polarized phased array with the high performance front-end module based on system-in-package (SiP) technology. By employing high-temperature co-fired ceramic (HTCC) substrates, the proposed design achieves efficient thermal management and high level of integration within a tile-type architecture. The front-end module based on SiP can simultaneously generate four independent beams with switchable left- and right-hand circular polarizations, providing flexible beam control. To verify the proposed method, a Ku-band 256-element phased array receiver with four beams has been designed and experimentally verified using HTCC and SiP process. Operating in 14-14.5 GHz, the proposed low-profile array demonstrates stable radiation characteristics, beam pointing accuracy and excellent beam consistency across the entire frequency range. The measurement results confirm that the SiP-based phased array maintains efficient thermal management, high polarization purity and robust beam-scanning capability, validating its suitability for mobile satellite communication.

Chaff cloud is one of the effective passive jammers for target covering. The target releases generous of chaff fibers to generate strong radar echoes, thereby preventing radars from locking onto the valid target. Previous studies on the geometric and scattering modeling of the chaff cloud often idealized the chaff fiber as finite-length straight conductors. However, chaff fiber is bending due to the influence of wind and gravity. This bending effect then alters the scattering properties of the chaff cloud, which leads to inaccuracies in tactical effectiveness estimations for chaff clouds. Consequently, this letter describes the bending effect of chaff fibers based on the Euler-Bernoulli beam theory. The Method of Moments (MoM) with pulse basis functions is employed to calculate the scattering of the chaff clouds that contains curved fibers. The letter also explores the relationship between the distribution of impedance matrix components, the monostatic scattering of the chaff cloud, and the curvature of the fibers. The results demonstrate that the bending effect of fibers is a non-negligible factor in evaluating the interference effectiveness of chaff clouds.

The Shooting and Bouncing Ray (SBR) method is widely utillzed for analyzing the Radar Cross Section (RCS) of multiple electrically large targets. However, several bottlenecks are encountered by existing graphics processing unit (GPU)- accelerated SBR solvers, including the dielectric materials, numerical divergence at grazing incidence as ray density increases, and substantial computational waste. To overcome these bottlenecks, the OptiMECA-BRC algorithm is proposed. Within this algorithm, the dielectric material problem is specifically addressed through the development of an OptiX-native dyadic Modified Equivalent Current Approximation (OptiMECA) method. Moreover, numerical divergence at grazing incidence is resolved through the implementation of a near-grazing correction. Additionally, to alleviate computational resource waste in multi-target swept-angle RCS simulations, an adaptive Bitmap-Masked Ray Culling (BRC) scheme is introduced. Finally, the overall computational efficiency is further enhanced by a customized complexvalued hierarchical warp-synchronous parallel reduction scheme for fully on-device far-field accumulation. Numerical validations demonstrate that OptiMECA-BRC reduces GPU memory by up to 66.7% over conventional methods and concurrently delivers up to a 37× speedup compared to Altair FEKO RL-GO while maintaining close agreement.

Range error caused by radar positioning inaccuracies significantly degrades Circular Synthetic Aperture Radar (CSAR) imaging by distorting scatterers' reflectivity distributions. While autofocus methods typically compensate for such errors, they have not established quantitative relationship between acceptable image quality and maximum range error. This paper proposes a novel statistical model linking image degradation to range error limits in CSAR systems. Using the Back-Projection Algorithm (BPA), we derive a direct relationship between reconstructed image quality (measured by Mean Absolute Error, MAE) and the statistical properties of range error. This relationship enables prediction of maximum tolerable range error for specified MAE requirements. The proposed relationship is validated using four targets with different levels of symmetry. For the centrosymmetric targets, the prediction error is less than 7%. Although the symmetry becomes less pronounced, the proposed relationship remains valid, with the prediction error increasing by approximately 10%. The method provides a quantitative design criterion for CSAR trajectory planning, effectively minimizing calibration burdens in practical measurements.

This paper presents a compact multi-beam dual-circularly polarized phased array receiving system operating in the 10.7-12.7 GHz frequency band is designed and implemented, which can generate eight reconfigurable receiving beams with independently configurable polarization modes and scanning directions for each beam. To improve the aperture utilization efficiency of the array and reduce the array size, the proposed phased array architecture adopts a "full-aperture multiplexing" beamforming method, where all beams share the same array aperture. For cost-effective phased array architecture with two-dimensional scalability, the array is divided into several identical receiving subarrays, with the control and power supply modules arranged beneath the array aperture. In addition, a heterogeneous integration scheme is introduced to realize high-density integration of various receiving functional chips, which reduces the overall array footprint by approximately 30% while maintaining the basic performance of the system gain-to-noise-temperature ratio (G/T). Meanwhile, different dielectric substrates are adopted to implement multi-level combining networks, optimizing the trade-off between overall efficiency and cost. To verify the feasibility of the proposed architecture, a prototype with a 16 × 16 array configuration is developed and tested. The measured results show that the array gain reduction is no more than 4 dB at a maximum scanning angle of 60°, and the G/T value of all beams in the boresight direction is not less than 0.9 dB/K at 11.7 GHz. The experimental results validate the effectiveness of the proposed multi-beam dual-circularly polarized phased array architecture in terms of engineering implementation and system performance.

Forecasting sunspot number (SSN) is essential for understanding solar dynamics and mitigating space weather impacts. Previous studies have primarily focused on a single timescale or specific decomposition strategy, whereas the inherently complex and multi-scale nature of solar activity requires a unified predictive framework. In this study, four SSN series, including yearly mean (YMSSN), monthly mean (MMSSN), 13 month smoothed monthly (SMSSN), and daily (DSSN) total sunspot number, spanning from 1818 to 2024 were investigated. To address the multiple timescales SSN prediction challenges, a mode decomposition-long short-term memory (MD-LSTM) framework was proposed. Three MD methods, including empirical mode decomposition, ensemble empirical mode decomposition, and variational mode decomposition (VMD), were applied to decompose each series into intrinsic mode functions (IMFs). This decomposition mitigates non-stationarity effects, enabling clearer temporal pattern extraction. Subsequently, an LSTM network was used to capture long-term dependencies and nonlinearity within each IMF, thereby enhancing temporal modeling. Experiments consistently demonstrated that all MD-LSTM variants outperformed the baseline, confirming the advantage of incorporating decomposition. Among the hybrid frameworks, VMD-LSTM yielding more distinct components and superior performance, with R2 scores often above 0.99. Moreover, the consistent superiority was observed across all four timescales, supporting the applicability of the MD-LSTM framework. Finally, using optimal VMD-LSTM-SMSSN model, the ongoing solar cycle 25 was forecasted to peak at 158.57 in 2024 August, which aligns closely with recent observation. Furthermore, the model forecasts that solar cycle 26 will reach its maximum in 2035 February with a peak of 156.85, demonstrating the framework's utility for long-term prediction.

This paper investigates orbital angular momentum (OAM)-based radar imaging for electrically large realistic targets. Existing OAM imaging studies mainly focus on ideal point scatterers, while the imaging characteristics of realistic targets remain insufficiently explored. To address this gap, scattered echoes of electrically large realistic targets are computed by combining the angular spectrum decomposition method (ASDM) with the physical optics (PO) method, and OAM-based radar imaging is then studied using these echoes. By exploiting the approximate duality between the topological charge and the target's azimuth angle, one-dimensional angular-azimuth imaging is first analyzed. By further incorporating conventional radar imaging methods, two-dimensional range-angular-azimuth imaging, two-dimensional range-cross-range imaging, and three-dimensional image reconstruction are investigated. The polar format algorithm (PFA) with four-nearest-neighbor interpolation is employed to improve imaging quality. The effects of Bessel-function modulation and topological charge on image quality are examined, and the imaging behavior of typical targets, such as a blunt cone, is further analyzed under multiple viewing angles, different signal-to-noise ratios (SNRs), and target-position mismatch conditions. The results demonstrate that vortex waves show promise for target reconstruction while also revealing current limitations. This work therefore serves as a simulation-based bridge between point-scatterer OAM imaging models and realistic-target imaging based on scattered echoes, providing theoretical support for OAM-based radar detection.

A quasi-steady pressure decrease flow-radiation framework was established to investigate the evolution of rocket exhaust plumes under chamber depressurization conditions. In the simulations, the chamber pressure was reduced from 70 atm to 10 atm at several prescribed depressurization rates. Infrared (IR) radiative transfer within the plume was then evaluated using the discrete ordinate method (DOM) combined with a statistical narrow-band (SNB) model. The proposed approach was validated by comparison with experimental plume spectra and reference line-by-line (LBL) solutions. The flow dynamics and infrared thermal radiation characteristics of a typical solid rocket motor (SRM) plume under prescribed chamber-pressure decay were analyzed. The results show an approximately linear dependence between plume infrared radiation and the temporal evolution of chamber pressure. This relationship indicates that the chamber depressurization rate can be used to estimate the infrared thermal radiation of the plume during the engine shutdown process.

A coupled numerical framework is established to investigate exhaust plume evolution under varying chamber depressurization conditions. The chamber pressure is continuously reduced from 70 atm to 10 atm using several prescribed depressurization rates. Infrared radiative transfer within the plume is subsequently evaluated using a discrete ordinate method (DOM) combined with a statistical narrow-band (SNB) model. The proposed approach is validated through comparison with experimental plume spectra and reference line-by-line (LBL) solutions. The simulation analyzes the flow dynamics and infrared thermal radiation of a typical solid rocket motor (SRM) plume during extinguishment. Results reveal a clear linear dependence between plume infrared radiation and the temporal evolution of chamber pressure. This finding indicates that the depressurization rate within the combustion chamber can be effectively utilized to predict the infrared thermal radiation of the plume throughout the entire engine shutdown sequence.

In this work, physical optics (PO) scattered near fields of electrically large target under different excitation sources are calculated by the linear amplitude-based fast physical optics method combined with the adaptive mesh technique (ALFPO). A key innovation is the derivation of phase gradient-based adaptive sampling interval formulas for quadratic patches, applicable to calculation of scattered near and far fields, which addresses the over-sampling issue in PO method. Also, adaptive mesh distribution generated via the phase gradient and local error of scattered field is analyzed in view of high-frequency wave physics. Moreover, the edge grid segmentation (EGS) approach is proposed to smoothen shadowed boundaries, ensuring the more accurate lit region data for calculation. Numerical results demonstrate that both the phase gradient-based ALFPO (PG-ALFPO) and scattered field error-based ALFPO (FE-ALFPO) outperform the linear amplitude-based fast physical optics (LFPO) method and FEKO software. They reduce the number of patches by one or two orders of magnitude while maintaining accuracy. Notably, the PG-ALFPO directly reflects high-frequency physical mechanisms. Both ALFPO methods achieve memory savings at the cost of CPU time, with PG-ALFPO exhibiting nearly twice the efficiency of FE-ALFPO. In short, the ALFPO methods along with EGS approach provide a significant way to solve near-field scattering features of electrically large target.

Optical topologies in the form of Skyrmions have attracted significant interest of late, where their integer Skyrmion number has been shown to be robust to complex media. Here we create the first fractional Skyrmions by structuring light as a vectorial superposition of non-integer orbital angular momentum. We unravel the map structure to reveal a new phenomenon, the abrupt transition jumps in skyrmion number, which serves to reinforce the integer nature of skyrmion topologies. Our experimental demonstration agrees well with simulation, opening a new spectrum of optical topologies to explore, with exciting possibilities in optical communication and sensing.

This letter presents an improved shooting and bouncing ray (SBR) method for fast radar cross section prediction of electrically large target. First, based on the similarity between adjacent ray paths, a facet neighborhood search method is employed to solve the ray-facet intersection tests. A central ray on emission plane is introduced to accurately address mutual occlusion determination. And a nonuniform ray distribution is designed for the emission plane, which considerably reduces the number of ray-facet intersection tests. Second, a forward bounding box is constructed and uniformly divided, which can quickly determine the traversed grids by the ray. Meanwhile, the forward stripping technique is adopted to remove the shadow facets. Invalid nodes are eliminated in the direction of the ray, and the nearest illuminated facet is identified, thereby avoiding unnecessary ray intersection. Third, the high-order ray tracing is achieved by the ray-marching technique. Moreover, the effectiveness of the proposed method is validated through comparison with the multilevel fast multipole algorithm (MLFMA) in the FEKO software. And a comparison with the traditional SBR method demonstrates that our method significantly improves computational efficiency.