Numerical and experimental generated data during project https://doi.org/10.1109/TAP.2024.3370296

The dataset was generated using a procedure for efficient global optimization (EGO) of multiband antennas, where the surrogate is repeatedly built and refined using custom-defined response features. The infill criteria are based on minimizing a surrogate-evaluated objective function, whereas the underlying optimization engine is the particle swarm optimization algorithm (PSO). Comprehensive benchmarking demonstrates the superiority of the presented approach over surrogate-assisted methods handling antenna frequency responses, as well as direct nature-inspired optimization.