Scarecrow‐Shaped Antenna Optimization Using Machine Learning Algorithms

ABSTRACT In this article, scarecrow‐shaped antenna with a Rogers RT6002 substrate with a permittivity of 2.94 and a thickness of 1 mm is presented. It is operating from 3.5 to 12 GHz frequency band. The next generation of wireless communication networks will make extensive use of machine learning (ML). It is anticipated that the growth of various communication‐based applications will improve coverage and spectrum efficiency when compared with traditional systems. A wide range of domains, including antennas, can benefit from the application of ML to generate solutions. Scarecrow‐shaped antenna is optimized using machine learning algorithms decision tree, random forest, XGBoost regression, K‐nearest neighbor (KNN), and light gradient boosting regression (LGBR). The antenna's return loss, gain, and directivity were predicted in this work. The KNN achieved the highest accuracy in the prediction of return loss. Hence, proposed antenna is suitable for flexible wireless communication systems, IoT, 5G, and 6G.