Scholarlink

<p>This work introduces a high-performance frequency-reconfigurable leaf-inspired antenna (FRLA) operating in the terahertz (THz) frequency range. The proposed design methodology consists of several stages, including the development of a bio-inspired leaf-shaped antenna, the integration of a graphene-based metasurface reflector for gain enhancement, the implementation of multi-state reconfigurable switching, and the application of machine learning for antenna performance optimization. The leaf-inspired antenna with asymmetric branches is fabricated on a flexible polyimide substrate to achieve increased electrical length while maintaining a compact size of 310&#xa0;μm × 380&#xa0;μm. Frequency reconfigurability is achieved by incorporating eight graphene switches, enabling the antenna to operate in ten different frequency modes ranging from 0.74 THz to 1.37 THz, with an approximate fractional bandwidth variation of 59.7%. To further enhance the directive gain, a metasurface reflector (MSR) consisting of a 5 × 5 hexagonal unit-cell arrangement is strategically integrated beneath the antenna structure. A Gradient Boosting machine learning model was employed to optimize both the distance of the graphene switches from the antenna feed line and the spacing between the FRLA and the MSR. The optimal switch distance was found to be 11.929&#xa0;μm, while the optimal FRLA–MSR separation distance was determined to be 211.364&#xa0;μm. Simulation results demonstrate a maximum predicted gain of 9.06 dBi, which was validated through CST simulations with a realized gain of 9.13 dBi, along with a radiation efficiency of 97%. The proposed design offers a promising solution for adaptive 6G communication and sensing networks. Unlike previously reported designs, the proposed approach uniquely combines bio-inspired geometries, graphene-enabled tunability, metasurface-based gain enhancement, and intelligent machine-learning optimization within a single THz antenna platform, thereby achieving an effective balance between compactness, flexibility, and high performance.</p> Graphical Abstract <p></p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Metasurface-Enhanced Graphene Leaf-Inspired Reconfigurable Antenna for High-gain 6G Terahertz Applications Optimized Using Machine Learning

  • 期刊论文

  • 发表日期：

    2026年7月1日

查看全文
  • Amany A. Megahed,
  • Marwa E. Mousa,
  • A. J. A. Al-Gburi,
  • Rania Hamdy Elabd

摘要

This work introduces a high-performance frequency-reconfigurable leaf-inspired antenna (FRLA) operating in the terahertz (THz) frequency range. The proposed design methodology consists of several stages, including the development of a bio-inspired leaf-shaped antenna, the integration of a graphene-based metasurface reflector for gain enhancement, the implementation of multi-state reconfigurable switching, and the application of machine learning for antenna performance optimization. The leaf-inspired antenna with asymmetric branches is fabricated on a flexible polyimide substrate to achieve increased electrical length while maintaining a compact size of 310 μm × 380 μm. Frequency reconfigurability is achieved by incorporating eight graphene switches, enabling the antenna to operate in ten different frequency modes ranging from 0.74 THz to 1.37 THz, with an approximate fractional bandwidth variation of 59.7%. To further enhance the directive gain, a metasurface reflector (MSR) consisting of a 5 × 5 hexagonal unit-cell arrangement is strategically integrated beneath the antenna structure. A Gradient Boosting machine learning model was employed to optimize both the distance of the graphene switches from the antenna feed line and the spacing between the FRLA and the MSR. The optimal switch distance was found to be 11.929 μm, while the optimal FRLA–MSR separation distance was determined to be 211.364 μm. Simulation results demonstrate a maximum predicted gain of 9.06 dBi, which was validated through CST simulations with a realized gain of 9.13 dBi, along with a radiation efficiency of 97%. The proposed design offers a promising solution for adaptive 6G communication and sensing networks. Unlike previously reported designs, the proposed approach uniquely combines bio-inspired geometries, graphene-enabled tunability, metasurface-based gain enhancement, and intelligent machine-learning optimization within a single THz antenna platform, thereby achieving an effective balance between compactness, flexibility, and high performance.

Graphical Abstract