Optimized 5G-MMW Compact Yagi-Uda Antenna Based on Machine Learning Methodology

The fifth generation (5G) of the mobile communication should provide a faster latency rate, wider Bandwidth (BW), and higher Gain (G) in comparison with older systems, (e.g. fourth generation (4G)). For 5G applications, the millimeter wave (MMW) antennas seem to be a suitable choice due to their small size. Owing to a large number of design parameters, designing an optimum antenna that can satisfy the 5G conditions is a very challenging task. In the meanwhile, using machine learning (ML) approaches to find the optimum design is an appropriate solution. Surrogate-based optimization (SBO) can handle the high computational cost of ML approaches, especially when the number of design parameters is large. The microstrip Yagi-Uda antennas play an important role in 5G communication systems due to their high BW and high G.