Particle swarm optimization for performance enhancement of electromagnetic band gap resonator antennas

This thesis investigates the use of particle swarm optimization (PSO) algorithm for performance improvement of the electromagnetic band gap resonator antennas (ERAs) with a significantly less computational cost compared to other methods. Our recent investigations into aperture field phase distribution of ERAs indicates that the phase distribution is quite non-uniform for conventional ERAs, resulting in poor quality of radiation patterns and below-optimal directivity. A PSO algorithm is implemented to design two phase correcting structures (PCSs), one made from all dialectric heights and the other made from triple printed layers. A performance comparison between the proposed PSO algorithm and the embedded optimizer in CST Microwave Studio (CST MWS) revealed that the proposed PSO algorithm reached the final design fourteen times faster than the built-in PSO optimizer in CST MWS. The all-dialectric PCS designed by PSO has significantly enhanced the phase uniformity of the ERA, improving radiation patterns by 4.6 dB in peak directivity. Following this approach, a printed PCS was designed by the PSO algorithm to achieve a planar structure. The use of PSO resulted in creating the printed PCS by 2480 unit-cell simulations. On the other hand, to design such a PCS by parametric sweep, 729,000 unit-cell simulations are required. The printed PCS greatly outperforms its dialectric counterpart in terms of phase uniformity and radiation patterns. Aperture efficiency increased by almost ten times and the peak directivity also improved by 9.6 dB compared to the ERA without PCS. In the end, a low-profile bandpass frequency selective surface (FSS) was designed using the PSO algorithm to verify the versatility of the proposed algorithm in designing various electromagnetic devices. Such FSS can also be designed by the parametric sweep at a significantly higher computational cost.