Broadband Millimeter-Wave SIW Cavity-Backed Slot Antenna for 5G Applications Using Machine-Learning-Assisted Optimization Method

A novel broadband substrate integrated waveguide cavity-backed slot antenna for 5G millimeter-wave wireless applications is proposed. This new antenna topology with constrained footprint achieves great bandwidth enhancement compared with previously proposed designs, by introducing multi-mode operation using loaded cross-shaped slot and unbalanced metallic vias. Moreover, a new machine-learning-assisted optimization method with additional feature is proposed to optimize antenna parameters. The above mentioned design-optimization procedure yields an antenna with measured fractional impedance bandwidth of 23.7%. The measured maximum gain of the antenna is 7.4 dBi within the entire bandwidth.