Design and optimization of miniaturized UWB MIMO antenna using ML methods for multiband applications

This article presents a new Ultra-Wide Band (UWB) antenna having miniaturize dimension of 15 x 10 mm 2 . It provides an impedance bandwidth of 109.71% and features a staircase ground structure (SGS) to achieve a wider bandwidth and good impedance match. This design is further developed to include a miniaturized Multiple Input Multiple Output (MIMO) UWB antenna. The MIMO antenna consists of four patch elements with different ground planes and a substrate size of 36 x 36 mm². To achieve a compact design and ensure high isolation, the patch elements are rotated by 90 degrees. The UWB MIMO antenna is optimized using a precise Decision Tree (DT) Machine Learning (ML) algorithm to shorten simulation time. It has a combined elliptical and rectangular patch that shows the required return loss (> 10 dB) and offers high isolation (> 20 dB) through polarization diversity. It has a very low envelope correlation coefficient (ECC) of less than 0.005, TARC values under 10 dB, CCL below 0.4 bits/s/Hz, and provides a Diversity Gain (DG) of more than 9.995 dB across most of the UWB. For UWB applications, there is a strong agreement between the measured and simulated results for the designed antenna. In addition to UWB applications, the antenna also supports Super Wide Band (SWB) applications like C-band, X-band, Ku-band, and K-band.