A Statistical Method for the Project and Analysis of a Yagi Antenna Using the HFSS and the Unscented Transform

This work investigates the use of statistical sensibility as a design tool for Yagi Uda antennas using the HFSS (High Frequency Structure Simulator) software and the Unscented Transform (UT). The study is based on correlation, covariance and regression analysis of input and output parameters. The test scenario used a Yagi antenna with four elements with a half wavelength feeder. The data were obtained by means of simulation using the HFSS software. The correlation study used the UT among the involved variables in the process, basically, distances between the elements and the impact of these in the output parameters: VSWR and the Total Gain. The idea is to seek which combinations allow better antenna performance, while being robust to small variations of the input parameters.