Synthesis of patch antennas loaded by inhomogeneous substrates via a combined spectral domain: Genetic algorithm approach

Abstract

The optimized design of patch antennas with inhomogeneous\ud loading materials is presented in this paper. The analysis and a first coarse\ud design of the antennas is applied to a numerical tool based on the method\ud of moments (MoM) and on a full-wave formulation developed in the spectral\ud domain that exploits both rigorous and asymptotic closed-form expressions\ud of the Green’s function for a stratified integrated structure with inhomogeneous\ud substrates. In order to match the specific prescribed\ud requirements for the antennas, then, the optimization of the patch shape\ud (fine design) is performed by a dedicated subroutine implementing the genetic\ud algorithm (GA) search and integrated in the aforementioned MoM\ud numerical tool. Some numerical results showing the capabilities of the code\ud are finally presented

Design of an Active Integrated Antenna for a Pcmcia Card

“…This iteration stage is not viable through a trial and error approach, because the time required for each MoM full-wave simulation is very long and the amount of the employed memory is very extensive. To overcome this limitation, a GA search has been integrated in the MoM code as in [13], employing the following fitness function to be minimized:…”

“…Applying this optimization tool, the structure is simulated only once (i.e., at the first iteration, as reported in the previous point), while the next configurations, obtained varying the geometrical dimensions, do not need to be simulated again through an extensive MoM simulation (see [13] and references therein). The electrical and radiating features of the optimized antenna are reported in Figs.…”

“…The starting iteration stage of this procedure is given by the set of the geometrical parameters obtained when the radiating element (with a PEGP) and the two-layer HIGP are designed independently in order to work in the required frequency band. Then, a classic GA [13] is used to adjust the geometrical parameters so that the antenna matches the desired requirements.…”

“…Therefore, a proper optimization technique, based on the Genetic Algorithms (GAs) [13], has been employed to optimize the overall antenna module (radiating element + HIGP). The analysis tool is based on the Spectral Domain (SD) full-wave analysis [14] together with a Method of Moment (MoM) numerical technique [15].…”