Antenna design through variable-fidelity simulation-driven optimization

In most cases, the adjustment of antenna geometry parameters requires repetitive electromagnetic (EM) simulations. As accurate, high-fidelity EM simulations are computationally expensive, an automated optimization process by embedding the EM solver directly into the optimization algorithm may be prohibitive. In this paper, a simple yet efficient simulation-driven design methodology of antenna structures is discussed and demonstrated. Our approach exploits a family of EM-based models of increasing discretization density that are sequentially optimized with the optimal design of the lower-fidelity model being the initial design for the higher-fidelity one. The design is further refined using a suitable response surface approximation model. We show that-using this technique-an optimized antenna design can be obtained at a low computational cost. Two examples are provided: a wideband microstrip antenna and a double ring antenna.