Machine Learning Assisted Optimization of Dielectric Resonator based mm-Wave MIMO Antenna for 5G Communication Systems

In this paper, a dual-port dielectric resonator (DR) multiple-input multiple-output (MIMO) antenna along with a suspended metasurface is designed and analyzed. The metasurface is used to improve the isolation value between the port of the proposed antenna. Moreover, different machine learning (ML) algorithms (Decision Tree, Deep Neural Network, K-nearest Neighbor, Random Forest, and Extreme Gradient Boosting) are utilized for optimization purposes. The performance of different ML algorithms are compared in term of measured parameters. The prototype of the proposed antenna design is fabricated for verification purposes. The measured result confirms that the proposed antenna operates from 26.24 to 27.94 GHz, with an isolation level of more than 45 dB between the ports. The gain of the proposed antenna is around 5.0 dBi in the operating band. Due to all these features, the proposed antenna can be efficiently utilized for a 5G communication system.