Machine Learning Optimized Multi-Element Antenna Systems for 5G and IoT Applications

We present a compact four-element Multi-input and multi-output (MIMO) antenna specifically designed for 5G and IoT applications. The design process starts with a single antenna element, which undergoes optimization using a machine learning-based approach, employing the Ensemble Bagged Trees model in combination with the Simulated Annealing (SA) algorithm. The optimized single-element antenna, with dimensions of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$27 \times 29 ~\text{mm}$</tex>, is later integrated into a four-element MIMO system, occupying a total area of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$56 \times 56 ~\text{mm}$</tex>. The designed MIMO antenna offers an ultra-wide operating bandwidth ranging from <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$2.3 \mathbf{G H z}$</tex> to 20.24 GHz. The performance is analyzed using key indicators that measure signal correlation, diversity characteristics, reflection behavior, and overall channel efficiency. The results demonstrate that the proposed design achieves both wideband and compact performance, making it a strong candidate for next-generation 5G and IoT communication systems.