Parallel and bio-inspired computing applied to analyze microwave and photonic metamaterial strucutures

Author(s)

Carlos Henrique da Silva Santos

• Document type: Doctoral Thesis
• Press: Campinas, SP.
• Institution: Universidade Estadual de Campinas (UNICAMP). Faculdade de Engenharia Elétrica e de Computação
• Defense date: 2010-05-31

Examining board members

Hugo Enrique Hernández Figueroa; Vitaly Félix Rodríguez Esquerre; Cosme Eustaquio Rubio Mercedes; Romis Ribeiro de Faissol Attux; Gustavo Fraidenraich

• Advisor: Hugo Enrique Hernández Figueroa

Abstract

The objectives of this work are to develop efficient (runtime and convergence) and flexible (multiplatform) bio-inspired algorithms (Genetic Algorithm, Evolution Strategy and Artificial Immune System), for the optimization of electromagnetic devices either in frequency and time domain. In addition, to satisfy the computational requirements for these optimizations a parallel version of these bio-inspired algorithms and a three-dimensional finite difference time domain (3D FDTD) were developed capable to be processed in cluster computers. The successful performance of these parallel resources was confirmed by their careful assessment through key tests available in the specialized literature. This permitted the proposal and simulation of novel and complex electromagnetic devices based on metamaterial structures (AU)