GFGuirong Feng

Guirong Feng

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Guirong Feng has a total of 41 co-authors. Here are the top co-authors by number of shared publications:

  • Xiaowei Shi

    National Key Laboratory of Antennas and Microwave Technology, xidian university, Xian, China

    4 shared publications
  • Jiaze Xu

    School of Electronic Engineering Xi’an University of Posts and Telecommunications Xi’an Shaanxi China

    3 shared publications
  • Lei Chen

    National Key Laboratory of Antenna and Microwave Technology School of Electronic Engineering, Xidian University Xi'an China

    3 shared publications
  • Yabo Lu

    School of Electronic Engineering Xi’an University of Posts and Telecommunications Xi’an Shaanxi China

    3 shared publications
  • Zhizhang Chen

    Dalhousie University

    3 shared publications
  • Zixiang Wu

    School of Electronic Engineering Xi’an University of Posts and Telecommunications Xi’an Shaanxi China

    3 shared publications

Latest Publications

  • Open Access

    Inverse Design of Electromagnetic‐Induced Transparency Metasurface Using Conditional Generative Adversarial Network

    by Guirong Feng, Jiaze Xu, Yabo Lu, Zixiang Wu, Ziheng Zhou, Yueming Gao, Zhizhang (David) Chen

    Traditional metasurface inverse design methods suffer from intensive manual modeling, laborious data collection, and extensive numerical calculations, leading to a compromised design efficiency and accuracy. To solve these drawbacks, this paper proposes a metasurface inverse design method using a conditional generative adversarial network (CGAN). A dataset construction framework integrating modeling, simulation, and response processing is established to ease data samples collection and reduce manual workload. Different from conventional parameterized modeling, the metasurface unit patterns are constructed by using a pixel‐based graphical modeling method based on encoding matrices. Furthermore, Gramian Angle Fields (GAFs) are incorporated to explore the latent physical correlations in electromagnetic responses, effectively improving the quality of generated metasurface structure images. By embedding an optimized U‐Net architecture into the CGAN model to suppress gradient vanishing and improve training stability and resolution, high‐precision mapping from target responses to metasurface structures is realized. An electromagnetically induced transparency (EIT) metasurface validates the proposed method. The results demonstrate that the proposed method achieves lower MSE, fewer training epochs, and less training cost. The simulated transmission spectra of the designed metasurface agree well with the target spectra with an average spectral accuracy above 90%, exhibiting promising application potentials for metasurfaces inverse design.

  • Wideband Circularly Polarized Antenna Design Using Machine Learning Based Hybrid Method

    by Guirong Feng, Jiaze Xu, Zixiang Wu, Yabo Lu, Yueming Gao, Zhizhang David Chen

    A machine learning (ML) based hybrid method integrating XGBoost, convolutional neural network (CNN) and NSGA-Ⅱ optimization algorithm is proposed for design of a wideband circularly polarized antenna. To reduce the dimensionality of antenna structure parameters in the dataset, XGBoost is first employed to select parameters with higher sensitivity to reflection coefficient S11 and 3-dB Axial ratio (AR). The CNN then provides a fast prediction on electromagnetic responses. To leverage the CNN's inherent strength in processing image-type data, Gramian Angular Field (GAF) transformation is adopted to convert the structure parameters into images. Additionally, the CNN integrated with Visual Geometry Group (VGG) and Residual Network (ResNet) blocks enhances the prediction accuracy and model stability. While the NSGA-Ⅱ algorithm optimizes the S11 and 3-dB AR bandwidths by using the trained CNN, which significantly saves the iteration time. A circularly polarized monopole antenna is designed to validate the proposed method. The results demonstrate that the antenna achieves both wide S11 bandwidth of 122.55% (3.04∼12.66 GHz) and a wide 3-dB AR bandwidth of 76.51% (3.01∼6.74 GHz). This method provides an efficient approach for the fast design of antennas with high efficiency.

  • Class-Aware Consistency Learning for Open-Set Semi-Supervised Hyperspectral Image Classification

    by Hao Sun, Renyi Chen, Zihao Yin, Huaxiong Yao, Yaxiong Chen, Wei Xie, Guirong Feng, Xiaoqiang Lu

    Semi-supervised hyperspectral image (HSI) classification methods focus on exploring the spectral and spatial information of unlabeled samples. However, existing methods generally follow the closed-set setting, assuming that unlabeled samples do not contain novel classes, which is hard to hold in practical applications. This paper aims to study semi-supervised HSI classification in the open-set setting, i.e., unlabeled samples fall into novel classes, and proposes a class-aware consistency learning (CACL) method. First, to explore discriminative spectral-spatial features, a position-aware transformer is developed, which effectively models spatial position priors between the center pixel and its neighboring pixels via a symmetric position-aware encoding. Then, to reduce the interference from novel class samples on the model’s discrimination, a prototype-driven consistency learning is proposed, which accurately selects unlabeled samples belonging to known classes via a known class sampler, and efficiently utilizes their spectral-spatial information by modeling consistent predictions across different views. Finally, to further improve the distinguishability between known classes, a prototype contrastive optimization is proposed to decrease the distance between samples from the same class and increase the distances between those from different classes in the feature domain. Furthermore, an adaptive segmentation threshold is designed to accurately predict known classes and reject novel classes. Extensive experiments verify that our CACL outperforms the state-of-the-art methods, achieving the overall accuracy of 81.65%, 88.61%, and 92.88% on the Indian Pines, Salinas, and Pavia University datasets, with 10 labeled samples in each known class. The code is available at https://github.com/rock-in/CACL-main.


  • A Single-Layer Dual-Mode Metasurface Antenna for WBAN

    by Guirong Feng, Yabo Lu, Bin Wang, Jiaze Xu, Zixiang Wu, Hailong Yang

  • A Wideband Omnidirectional Metasurface Antenna Using Characteristic Mode Analysis

    by Guirong Feng, Zixiang Wu, Yabo Lu, Jiaze Xu, Shanzhe Wang, Zhizhang Chen

    A wideband omnidirectional metasurface antenna has been proposed. The characteristic mode analysis (CMA) was applied to explore the modal behaviors of metasurface and guide the design. The proposed metasurface is designed based on a 5×5 square patched array, which is a hybrid metasurface incorporating square patches with 2×2 and 3×3 sub-arrays in a distributive arrangement, significantly improving the bandwidth and current distribution over a wide band. By properly adjusting the sizes of these elements, the bandwidth and gain are further improved. Based on the predicted excitation of metasurface, a probe center-fed circular patch is introduced on the bottom layer, which not only excites the metasurface but also generates an additional mode to compensate the metasurface for attaining stable omnidirectional radiation. An antenna prototype was simulated, fabricated, and tested. Results show that the proposed antenna achieves a wide bandwidth of 70.1%, a peak gain ranging from 2 to 6 dBi, and radiates a good omnidirectional pattern.


  • Open Access

    A New Antenna Capable of Generating Broadband Plane Spiral Orbital Angular Momentum

    by Na Li, Guirong Feng, Ping Li, Xiaowei Shi

    The plane spiral orbital angular momentum (PS-OAM), which is a special form of orbital angular momentum (OAM) beams, solves the problem of OAM beam divergence to some extent. The majority of OAM antennas can only generate vortex beams in a narrow band range, which severely limits their broadband application potential. This study proposed a single antenna that can produce PS-OAM beams in the 8-16 GHz frequency range. The antenna is composed of six Vivaldi antennas placed in a tangential circular array, and the ports of the neighboring Vivaldi antenna elements exhibit phase difference. When fed with equal phase and amplitude, the antenna can generate broadband PS-OAM beams with mode +2. In contrast to current OAM antenna arrays fed with phase-shifting, the suggested antenna is straightforward and simple to feed. The proposed Vivaldi antenna based on UCA may provide a new way for generating broadband OAM beams.


  • Open Access

    A low-profile broadband crossed-dipole antenna with fractal structure and inverted-L plates

    by Guirong Feng, Tianyi Qi, Mingming Li, Jiachen Li, Xin-Wei Wang, Hai-Long Yang

    A low-profile broadband circularly polarized (CP) crossed-dipole antenna is proposed. In this antenna, the dipole uses a fractal-based structure with multiple similar patches for obtaining broadband CP radiation and impedance matching. The incorporation of the crossed-dipole antenna with four triangular parasitic patches improves both the impedance and axial ratio (AR) bandwidths. Four parasitic coupled inverted-L metal plates are loaded on the ground plane to significantly enhance the AR bandwidth and reduce the antenna profile. The measured results are in agreement with the simulations, which demonstrates that the proposed antenna has a low profile of 0.18 λ0, a − 10 dB impedance bandwidth of 91% (2.78–7.42 GHz), a 3 dB AR bandwidth of 81.5% (2.99–7.1 GHz), and a good right hand circular polarization (RHCP) radiation pattern with an average gain of 7.9 dBi over the whole operating band.


  • Open Access

    A Low-profile Broadband Crossed-dipole Antenna with Fractal Structure and Inverted-L Plates

    by Guirong Feng, Tian-Yi Qi, Jiachen Li, Xinwei Wang, Mingming Li, Hailong Yang

    A low-profile broadband circularly polarized (CP) crossed-dipole antenna is proposed. In this antenna, the dipole uses a fractal-based structure with multiple similar patches for obtaining broadband CP radiation and impedance matching. The incorporation of the crossed-dipole antenna with four triangular parasitic patches improves both the impedance and axial ratio (AR) bandwidths. Four parasitic inverted-L metal plates are loaded on the ground plane to significantly enhance the AR bandwidth and reduce the antenna profile. The measured results are in agreement with the simulations, which demonstrates that the proposed antenna has a low profile of 0.18 λ0, a -10 dB impedance bandwidth of 91% (2.78~7.42 GHz), a 3 dB AR bandwidth of 81.5% (2.99~7.1 GHz), and a good right hand circular polarization (RHCP) radiation pattern with an average gain of 7.9 dBi over the whole operating band.

  • A broadband circularly polarized monopole antenna employing parasitic loops and defective ground plane

    by Guirong Feng, Lei Chen, Xiaowei Shi

    A simple printed monopole antenna using parasitic loops and defective ground plane is presented for broadband circular polarized (CP) radiation. For this purpose, two square open loops, which install air gaps at the corner and close to the monopole, are printed on both sides for achieving broadband CP radiation with three minimum axial ratio (AR) points. By using a defective ground plane with a vertical slot and a horizontal slot, additional AR points can be excited to further broaden the AR bandwidth. This design was fabricated, and the measured results show that the antenna with a simple structure achieves both wide impedance bandwidth of 87.2% (2.62‐6.67 GHz) and AR bandwidth of 83.9% (2.6‐6.36 GHz).