A wideband multiple-microstrip dipole antenna with dual polarization is proposed in this letter. The consists of a radiator, cross-shaped slot coupler, pair microstrip baluns, and reflector. When baluns are excited, the coupler would work as four-way equal-split power divider generate four differential signals at ends slotlines. Afterward, be coupled to modified dipoles radiate synthesize slant ±45° linear polarizations. design verified by fabrication testing prototype antenna. Measured...

In this article, we propose an equivalent circuit theory-assisted deep learning approach to accelerate the design of metasurfaces. By combining filter theory and a sophisticated model, designers can achieve efficient metasurface designs. Compared with most existing generative methods that rely on arbitrarily generated training dataset (TDS), proposed method adaptively produce highly relevant low-noise samples under guidance theory, resulting in significantly narrowed target solution space...

Space-time-coding digital metasurface has drawn worldwide attention with the ability to improve communication quality and change direction of electromagnetic (EM) wave propagation in real-time. This article proposes a deep learning-assisted method design space-time-coding element (STCDME) state recognition mapping technique methods. Compared traditional pure EM simulation methods simulate all states, proposed fully considers relationship between different states STCDME accelerate design....

A filtering magnetoelectronic dipole antenna (MEDA) with quasi-elliptic gain response and wide operating band is investigated. Different from the conventional MEDA, studied here fed by fork-shaped microstrip line aperture-coupled excitation. The intrinsic radiation null of MEDA at lower passband edge utilized, a mathematical model first put forward to demonstrate its working mechanism. Four driven stubs connected shorted walls are used excite planar arms. parasitic inductor coupling...

Modern electromagnetic (EM) device design generally relies on extensive iterative optimizations by designers using simulation software (e.g., CST), which is a very time-consuming and tedious process. To relieve human engineers boost productivity, we proposed machine learning (ML) framework to solve the problem of automated for EM tasks. The approach combines advanced reinforcement (RL) algorithms deep neural networks (DNNs) in an attempt simulate decision-making process realize automation...

In this study, we propose a novel approach for automating antenna design through the utilization of domain knowledge-informed reinforcement learning (RL) and imitation (IL). The proposed method allows RL agent to learn from its interactions with environment take actions that maximize reward signal, resulting in optimal policy. Despite attractive advantages, process can be challenging time-consuming, as must explore vast range states, actions, policies. To enhance efficiency process,...

Frequency-selective surfaces (FSSs) refer to planar structures that behave with specific electromagnetic (EM) responses within a frequency range and are widely applied in wireless propagation systems. Given the fact different EM correspond distinguished topologies, conventional inverse design methods of FSSs usually labor-intensive, as they rely on experienced human engineers determine topology then rationally tune its structures. There have been great attempts using optimization algorithms...

Deep learning plays a vital role in the design of electromagnetic (EM) structures. However, current research, single neural network typically supports only one structure and requires complex framework to accommodate multiple designs. This paper proposes using neural-assist for facilitating EM We employ two filling methods control vector length, an identification method ensure accurate prediction results, random auxiliary vectors increase data volume reduce loss. Subsequently, we forward...

Electromagnetic (EM) structures play a significant role in wireless communication, radar detection, medical imaging, and so on. Machine learning (ML) has been increasingly applied to facilitate the design analysis of EM structures. Data acquisition is major bottleneck. Conventional methods blindly sweep geometric parameters on uniform grid acquire corresponding responses via simulation. Acquired data have unstable quality due inconsistent informativeness responses, leading low ratio model...

A transfer-learning-based method for accelerating power-only calibration of phased array antennas by combining the conventional theory with deep learning is presented in this communication. The existing methods either require a significant number measurement cycles or have restrictive phase shifter resolution requirements. proposed uses surrogate model to calibrate all elements one pass without restricting We developed novel feature extraction scheme (FES) that picks out most important power...

This article tackles the generalized synthesis of antenna arrays using two-order deep learning. Existing learning-assisted approaches mainly rely on model training electromagnetic (EM) simulation data, and hence feature limited generalization ability need for a huge amount EM simulations. The proposed learning method uses first-order to learn generic features radiation patterns from data efficiently generated by applying conventional array factors. After that, second-order learns simulations...

A compact and wideband differentially fed dual-polarized antenna with high common-mode suppression is investigated in this paper. square patch crossed-slot consisting of four slant funnel-shaped slots used as the radiator, generating ±45° linear polarizations. Two orthogonal placed baluns, each connected two open-end stubs, act feeding network. Under differential-mode (DM) excitation, resonating mode quarter-wavelength can be excited at low frequency band, respectively, which enhances...

Deep-learning-assisted antenna design methods such as surrogate models have gained significant popularity in recent years due to their potential greatly increase efficiencies by replacing the time-consuming full-wave electromagnetic (EM) simulations. A large number of training data with sufficiently diverse and representative samples (antenna structure parameters, scattering properties, etc.) is mandatory for these ensure good performance. However, traditional modeling relying on manual...

This article introduces a machine learning (ML) framework for the design of space–time-coding digital metasurface elements (STCDMEs), commonly used in reconfigurable intelligent surface (RIS)-based communication. It includes inverse design, forward and automodeling, which quickly achieve multistate electromagnetic (EM) structure designs, e.g., STCDME. The decision tree (DT) model is chosen use with its lightweight, fast response, highly accurate EM small-scale data modeling. In addition, we...

This paper discusses the training of deep neural networks (DNNs) for electromagnetic problems. The main concerns include how to modify EM problems take advantage learning techniques and tailor conventional concepts with domain knowledge, which has been overlooked by most existing DNN based research. A 1×8 patch antenna array adopted as test vehicle investigation, aim use radiation pattern synthesis. It is analyzed via simulation first collect sufficient data sets containing different...

Machine learning (ML) has demonstrated significant potential in accelerating the design of microwave components owing to its great ability approximate projection between geometric parameters and electromagnetic (EM) responses. A well-trained ML model can predict EM responses a component with unseen parameter settings accurately, or determine based on desired constraints matter milliseconds. However, this ML-based process often requires heavy simulation collect large amount training data. To...

In recent years, deep learning-assisted methods for antenna design, such as surrogate models, have gained popularity due to their potential improve design efficiencies by replacing time-consuming electromagnetic simulations. However, these require a large number of diverse and representative training data samples, which can be challenging acquire using traditional manual modeling methods. To address this issue, our study proposes an intelligent image-based parametric approach that leverages...

In this paper, an axially corrugated horn antenna loaded with a profile-shaped dielectric rod operating at Ku/E band is presented. The corrugation section based on conical mainly contributes to the radiation Ku band, and fed by coaxial waveguide operated TE <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ll</sub> mode. A circular constituting inner conductor of excited fundamental mode, which transformed into HE mode inserting in waveguide....

This paper proposes a broadband filtering magnetoelectric dipole antenna (MEDA) with stable gain utilizing U-shaped slots. By loading the slots, resonance is introduced to equivalent magnetic of MEDA, thereby generating radiation null in high-frequency band antenna. Due structural characteristics has its intrinsic low frequency band. Utilizing these two nulls, good out-of-band rejection obtained. The simulated and optimized using simulation software. results indicate that achieves an...

Deep learning-assisted antenna design methods such as surrogate models have gained significant popularity in recent years due to their potential greatly increase efficiencies by replacing the time-consuming full-wave electromagnetic (EM) simulations. However, a large number of training data with sufficiently diverse and representative samples (antenna structure parameters, scattering properties, etc.) is mandatory for these ensure good performance. Traditional modeling relying on manual...