Reflecting/Absorbing Dual-mode Textile Metasurface with AI-Driven Parametric Studies

This paper presents a textile-based reflecting/absorbing dual-mode metasurface, and emphasizes the role of AI-driven parametric studies in the designing process. The proposed textile metasurface can achieve a reflecting mode with the zero-degree refection phase centre at 2.4 GHz, and an absorbing mode at the same resonance frequency. The absorption and reflection band of the design are centered at the same frequency by applying a state-of-the-art AI-driven antenna design technique, self-adaptive Bayesian neural network surrogate model-assisted differential evolution for antenna optimization (SB-SADEA) method. The proposed design can also achieve polarization insensitivity and a certain level of incident angle insensitivity. The fabricated prototype of the design achieves a maximal absorption rate of 99.8% and maintains an absorption over 90% in the frequency range of 2.39 to 2.43 GHz. Moreover, a textile linear polarized monopole antenna was fabricated and tested along with the reflection metasurface. A 5 dB realized gain enhancement can be achieved with the metasurface applied. Both simulations and measurements verify the effectiveness of the proposed dual-mode textile metasurface design.