Abstract
This paper presents a computational intelligence-optimized multiband MIMO antenna for next-generation laptops, addressing the demand for compact, high-performance, and multiband wireless connectivity. The proposed 4-port MIMO configuration, with an overall footprint of 94.1 × 28.29 × 0.8 mm(3), is optimized for seamless integration along the laptop's top edge, where minimal width is crucial. It supports operation at 2.45 GHz, 5 GHz, and 6 GHz, enabling compatibility with dual and tri-band Wi-Fi 6E routers. To accelerate the design process and reduce manual iteration, machine learning (ML) algorithms including AdaBoost, SVM, CatBoost, and Decision Trees were employed. A simulation dataset was generated in CST studio by systematically varying critical antenna parameters. This dataset was used to train the ML models, enabling them to learn the nonlinear relationships between geometry and performance metrics such as S-parameters, gain, efficiency, envelope correlation coefficient (ECC), diversity gain. Upon training, the models predicted optimal design parameters for desired performance goals. The resulting antenna exhibited isolation greater than 16 dB, ECC below 0.08, and a measured realized gains of 0.73, 2.1, and 3 dBi across the operating bands. In addition, the channel capacity loss remained under 0.35 bits/s/Hz, confirming strong MIMO performance. This work highlights that incorporating computational intelligence into antenna design not only expedites the development process but also improves system efficiency, providing a scalable and intelligent solution for next-generation multifunctional, high-speed laptop platforms.