Use of TSK type fuzzy system based fitness function approximation for efficient optimization of low-profile wideband diversity PIFA by PSO

This paper presents a computer aided design (CAD) framework for efficient optimization of antennas. Generally complicated antenna structure is directly optimized using evolutionary algorithm (EA) where each fitness function evaluation requires a full-wave electromagnetic (EM) simulation and the whole design process requires thousands of full-wave simulations. In the proposed method, a Takagi-Sugeno-Kang (TSK) type fuzzy system is used to approximate the fitness function from fewer full-wave EM simulations. This approximated fitness function is used to optimize an antenna structure using particle swarm optimization (PSO) algorithm. In this paper, a low-profile selection-combining diversity printed-inverted-F-antenna (PIFA) structure is proposed and its geometry is optimized in the proposed CAD framework. It is found that the PSO in conjunction with TSK type fuzzy system forms an effective CAD framework which requires 100 times less number of full-wave EM simulations than the number of EM simulations required for direct PSO based optimization of the proposed diversity PIFA.