A compact ring-shaped metamaterial antenna with deep learning–enabled gain enhancement for 5G communications
摘要
This article introduces and describes a ring-shaped metamaterial-based combination antenna for usage with 5G technology in the sub-6 GHz range. In this proposed design, the antenna radiating elements serve as radiators on a circular and square ground plane that is created by combining circular and square coupling resonators. Additionally, both square and circular shaped resonators are integrated in the SRR design to improve the antenna’s overall performance within both gain and bandwidth constraints. The proposed SRR has a double-negative behaviour at the configuration’s resonant frequency. This behaviour allows for simultaneous increase in bandwidth and antenna gain while keeping a minimal profile. Important factors like size, gain, bandwidth, and reflection coefficient (S11) are used for evaluating the proposed antenna design. The proposed SRR integration reduces the antenna size measures from 22 × 22 mm to 18 × 18 mm, increased bandwidth from 0.76 GHz to 0.885 GHz, and produced a significant gain improvement from 2.2 dB to 3.3 dB. Further, the metamaterial antenna structure is optimized to enhance the antenna gain using deep learning model and metaheuristic optimization techniques. This optimized antenna prototype has been practically tested and produced a measured gain of 3.95 dB with close agreement between simulated and observed results. The data driven deep learning antenna modelling and structure optimization-based antenna design techniques are helped to improve the gain of the antenna. This antenna data-based design method is relatively inexpensive, simplifies the design process and less time consuming, resulting in improved gain and significantly reduced size, when compared to existing current-generation antenna design methods.