Statistical analysis of antenna robustness

Abstract

This work presents a methodology for the evaluation of antenna susceptibility to various types of random physical damage. A Monte Carlo statistical method is proposed to estimate antenna parameters for different damage levels. A virtual experiment is used in order to avoid lengthy experimental trials. Antenna parameters are modeled for each randomly generated damage pattern. Statistical post processing with the Bootstrap method evaluates the parameters acquired from the antenna modeling software. To illustrate the methodology a comparative study of the Input Return Loss for the damage inflicted upon a planar dipole and a bow-tie antenna is made. The median value is estimated to describe the "typical" behavior of the antenna. With taken assumptions, Input Return Loss for the Bow-tie antenna is found more sensitive to partial damage.