Optimization of Multilayer Frequency Selective Surfaces using the Binary Particle Swarm Algorithm

Multilayer frequency selective surfaces (FSS) structures can be applied in many research areas, as they can satisfy more restrictive behaviors and obtain spectral responses wider than the single layer ones. In this work, the frequency response of a dual layer FSS is optimized by means of a binary version of particle swarm algorithm (BPSO) that determines the best physical parameters of the structure -period, dielectric material and the two metallic geometries-comparing its response with some restrictions given in terms of frequency masks. Furthermore, a cost function used to weigh up any solution provided by the algorithm is defined taking into account the transmission and reflection coefficients obtained with a multilayer FSS analysis method. This approach combines a rigorous numerical procedure, the Conjugate Gradient Fast Fourier Transform (CG-FFT), to obtain the frequency characterization of each level, with a cascade algorithm, the discrete plane wave spectrum (DPWS), to calculate the full structure behavior. Two different masks have been used to demonstrate the usefulness of the approach proposed.