Optimization of linear antenna arrays using intelligent algorithms for secondary lobe level control

Abstract

Currently, evolutionary computation has developed several techniques based on the emulation of natural evolutionary processes to solve complex problems and are used in situations with large and non-linear search spaces, where other classical methods are not able to find solutions in a reasonable time. Among the most recognized techniques are genetic algorithms and swarm algorithms. Taking into account the various evolutionary optimization techniques, this thesis aims to compare the performance of at least three types of evolutionary algorithms in optimizing the feedforward coefficients of a linear array to determine a desired primary to secondary lobe level. Chapter 2 describes the state of the art of linear antenna arrays, emphasizing the synthesis for obtaining radiation patterns, as well as the design of waveguide slot arrays (SWAA), which was the design used for the simulations of the different evolutionary optimization techniques. In addition, a general description of the genetic and swarm-based algorithms is given, showing the advantages of each of them. Chapter 3 shows the state of the art of similar projects that have been carried out based on linear arrays, yagi-Uda antennas and waveguide slot arrays (SWAA). Describing the project on which this thesis was based, which was the design, optimization, and simulation of a twolayer centrally-fed waveguide slot array antenna with ten slots (2×10) operating at 12 GHz. The antenna achieves a bandwidth of 4% (440 MHz) with a return loss of -14 dB. The design process prioritized meeting bandwidth requirements and achieving a sidelobe level (SLL) of 20 dB through least-square optimization of excitation coefficients. Chapter 4 describes the implementation of the algorithms used in the study: Genetic Algorithm, Particle Swarm Optimization (PSO), Differential Evolution (DE), and Limitedmemory Broyden, Fletcher, Goldfarb & Shanno - Box Constraints (L-BFGS-B), providing an overview of the implemented functions, design decisions, and the elements used. Chapter 5 presents the results obtained in the project, describing the initial problem to optimize and the results of the conducted simulations. Finally, Chapter 6 presents the conclusions.