A Reusable Framework Based on Reinforcement Learning to Design Antennas for Curved Surfaces
This work addresses the problem of designing visually unobtrusive and electromagnetically efficient antennas for curved surfaces, which is an incremental improvement for antenna engineers.
This paper proposes a deep reinforcement learning framework to design low-profile antennas for curved surfaces, specifically metal tubes found in public infrastructure. The framework aims to automatically readjust antenna designs efficiently when variables like curvature radius or material properties change, optimizing radiation patterns and reducing visual impact.
The design and implementation of low-profile antennas has been analyzed in past decades from different perspectives while the purpose is to have a small size in the device, and an adequate electromagnetic behavior. This work pursues a methodology to identify small antennas and consequently presents some similarities. Meanwhile, curved surfaces are considered for a certain variety of antennas with reduced size. The so-called deep reinforcement learning technique is used as an assistance against morphological variations that are specifically taken into account in this work. The objective is to identify antennas that can be efficiently mounted on the surface of metal tubes such as those frequently present in public infrastructure (e.g. traffic lights and luminaries). The motivation is to reduce the visual impact and optimize the radiation pattern of the antenna. It is analyzed that if changes in variables such as the radius of curvature, or the electromagnetic properties of the materials appear, an automatic identification of the underlying characteristics of the problem (by means of machine learning techniques) can readjust the design efficiently. The results obtained in this work are analyzed based on variables that are typically used to characterize antennas, such as their impedance and radiation pattern.