Surrogate-Model-Aided Optimization of the Bandwidth for a Planar UWB Inverted-F Antenna

A new efficient methodology to speed up the optimization of the bandwidth of antenna designs is proposed in this paper. First, the reflection coefficient S
₁₁
(f) of the considered antenna is compactly represented by Fourier series. Second, the coefficients of the Fourier series are modeled by using the Least-Angle-Regression-Sampling-based Polynomial Chaos Expansion (LARS-PCE). Lastly, the derived surrogate model serves as an extremely efficient "engine" which can substitute the time-consuming full-wave simulators for the optima seeking. The proposed methodology has been validated on a planar Ultra-Wideband (UWB) Inverted-F Antenna (IFA), whose impedance bandwidth has been increased by more than 18 % compared to the original design and the computational cost is limited.