Numerically efficient design optimization of a printed 2.45 GHz Yagi antenna

Simulation-driven design of a 2.45 GHz quasi-Yagi antenna is presented that exploits a coarse-discretization antenna model and surrogate-based optimization algorithm. The design objectives include maximum directivity of the principal polarization, input impedance, side lobe level, front-to-back ratio, and direction of maximal radiation. The computation cost of the design process is low because most of the operations are performed on the coarse-discretization model, whereas the high-fidelity simulation is only used for setting up and updating the surrogate model, as well as for design verification.