Dual-polarized Base Station Antenna Design using Machine Learning-Assisted Optimization Method

A
$\pm 45^{\circ}$
dual-polarized base station antenna with a simple structure is proposed, and the method of machine learning assisted optimization(MLAO) is introduced. The time of full-wave simulation is reduced by establishing a Gaussian process regression(GPR) model, and the genetic algorithm(GA) is used to predict and iterate the optimal solution. Finally, a dual-polarized base station antenna with a gain greater than 9.5dBi, a horizontal half-power beam width(HPBW) of 65.69°, and a coverage of 2496 - 2690MHz was designed.