Abstract
Design optimization of antennas is challenging, in most cases it involves electromagnetic (EM) simulations. High-fidelity simulation is computationally expensive and, therefore, employing EM solver in an optimization loop may be prohibitive. In this paper, we discuss optimization approaches that exploit coarse-discretization EM simulations as a low-fidelity model, and allow us to locate the optimum design of the antenna at a low computational cost. Some practical issues, including the importance and selection of the model fidelity are also discussed.