Received: 24 June 2023 | Revised: 14 October 2023 | Accepted: 23 November 2023 DOI: 10.1002/mop.33977 RESEARCH ARTICLE Design of resonant metasurface absorber using feed‐forward neural network Abdelghafour Abraray 1,2 | Amit Baghel 1 | Stanislav Maslovski 1,2 1 Instituto de Telecomunicações, Aveiro, Portugal 2 Department of Electronics, Telecommunications and Informatics, University of Aveiro, Aveiro, Portugal Correspondence Abdelghafour Abraray, Instituto de Telecomunicações, Aveiro, Portugal. Email: a.abr@av.it.pt Funding information Fundação para a Ciência e a Tecnologia; Ministério da Ciência, Tecnologia e Ensino Superior; FCT/MCTES, Grant/Award Number: UIDB/50008/ 2020‐UIDP/50008/2020 Abstract In this work, an ultrathin ( ∕ λ λ 50, = 122.45 0 0 mm) design of a frequency‐ selective metamaterial absorber (MTA) is proposed. An absorption of 100% is achieved at the resonant frequency f = 2.45 r GHz with a fractional bandwidth of 1.45%. The MTA unit cell geometry is parametrized and modeled in a full‐ wave electromagnetic simulator. Using a data set generated by varying the absorber unit cell geometry and simulating it in SIMULIA CST Studio Suite, the feed‐forward neural network is able to learn the relationship between the physical structure and its electromagnetic response. The numerical results for the MTA performance have been confirmed by an experiment. An array of 25 MTA elements in a 5 ×5 configuration has been fabricated and tested in an anechoic chamber. The simulated and measured results are in a good agreement. KEYWORDS feed‐forward neural network, metamaterial absorber, metasurface absorber 1 | INTRODUCTION Metamaterial absorbers (MTAs) and metasurface absor- bers (MSAs)—which are 2D versions of MTAs—are essential for applications in many areas of science and technology. At microwave frequencies, both MTAs and MSAs are employed in antiradar coatings to reduce the radar cross‐section (RCS) of various objects, because such coatings suppress the specular reflection of the incident electromagnetic waves. 1,2 The frequency bandwidth of MTAs and MSAs varies depending on their design. There are known physical limitations on the bandwidth of antiradar coatings and analogous structures related to their thickness‐to‐wavelength ratio and the level of reflection. 3 Broadband MSAs can be designed using multilayer structures formed by resistive sheets 4–6 or capacitive sheets. 7 Frequency‐selective absorbers have been reported recently in the literature to reduce RCS at a given frequency. 8 Like other absorbing coatings, frequency‐ selective MSAs suppress specular reflections and reduce RCS, although achieving it within a quasiplanar struc- ture. Such structures can be also used (e.g., as protective sheets) in industrial environments that use powerful microwave signals with a fixed frequency. When dealing with a single‐frequency or narrowband operation, the application of a bulkier MTA may become impractical, as compared with a single‐layer MSA. Utilization of MSAs holds immense potential for enhancing energy management in wireless networks and simplifying the development of highly sensitive sensors. 9 Resonant MTAs are also useful in realizing frequency‐ and angular‐selective thermal emitters and sensors. Relatively recently, structures with high Microw Opt Technol Lett. 2024;66:e33977. wileyonlinelibrary.com/journal/mop © 2023 Wiley Periodicals LLC. | 1 of 7 https://doi.org/10.1002/mop.33977 Abbreviations: ANN, artificial neural network; FFNN, feed‐forward neural network; MTA, metamaterial absorber; RCS, radar cross‐section.