Evolving Antennas From Building Blocks Using Evolutionary Algorithms

Evolutionary algorithms use concepts of biological evolution in order to find solutions to difficult optimization problems. This work presents a genetic algorithm to evolve arbitrary antenna designs towards the goal of either matching a desired radiation pattern or maximizing sensitivity in user-specified directions. The algorithm constructs antennas by combining geometric shapes like building blocks. We explore the results of evolutions towards both fitness metrics in single-frequency (300 MHz) and broadband (200 MHz to 800 MHz) applications. This work was initially inspired from astrophysics experiments that are reliant on sensitive detectors, and future work will include metrics from astrophysical simulation software as the measure of fitness.