Technique for the design of matched antennas using method of moment-based greedy best-first search

As the number of wireless devices continues to grow, technical requirements and design constraints are approaching the limits of antenna performance and feature integration. Complex multi-band antennas that must satisfy stringent volumetric constraints are pushing the use of optimization tools to the top of the antenna designer's toolbox. Beyond parametric studies, global optimization algorithms have provided an array of insights into both theory and practical applications. In this thesis, a method of moments (MoM) based greedy best-first search (GBFS) technique for antenna optimization is presented. This novel approach using GBFS is a simple method to design antennas for minimizing the input reflection coefficient at one or multiple frequencies. By meshing the antenna region using MoM and applying a GBFS algorithm, antennas can be "grown" to optimize for certain performance specifications. The optimization technique results in designs that are contiguous pieces of metal, without "islands", making the results easier to analyze. This technique is applied to three design examples: a phone model optimized at a single frequency, a small planar monopole optimized for dual-band operation in the GSM-850 and PCS-1900 bands, and a miniaturized microstrip patch antenna. A sample small dual-band planar monopole antenna is fabricated and measured to validate the design process. This greedy search approach can also be used to make improvements to an already existing antenna design