Cost-Efficient Design Optimization of Compact Patch Antennas With Improved Bandwidth

In this letter, a surrogate-assisted optimization procedure for fast design of compact patch antennas with enhanced bandwidth is presented. The procedure aims at addressing a fundamental challenge of the design of antenna structures with complex topologies, which is simultaneous adjustment of numerous geometry parameters. The latter is necessary in order to find a truly optimum design and cannot be executed-at the level of high-fidelity electromagnetic (EM)-simulation models-using conventional numerical optimization procedures due to prohibitive computational cost. Here, we employ coarse-discretization EM simulations as faster representation of the antenna under design (referred to as a low-fidelity model). The low-fidelity model is enhanced using a combination of frequency scaling and response correction techniques. Frequency scaling is particularly suitable for reducing misalignment between EM models of various fidelities in case of narrowband structures. Iterative correction-prediction loop is capable to identify an optimum design at a reasonably low computational cost while adjusting all relevant geometry parameters at the same time. For the sake of illustration, we consider a modified patch antenna based on transversal signal-interference feed and additional slots in the radiator. Upon optimization, the antenna provided 29% bandwidth while maintaining a compact footprint of 645 mm
²
. Experimental validation is also provided.