Computer Vision and Simulation Tools For Three-Dimensional Random Antenna Arrays

Abstract

This research project aims to develop and improve auxiliary tools which aid in ongoing research on beamforming with random antenna arrays. One of the challenges of conducting research on random arrays is the tedious nature of preparing electromagnetic simulations to test beamforming algorithms. This project presents an automation framework written in Python which will expedite the setup of simulations and reduce the room for error during this process. A computer vision system designed to locate the feedline of the patch antennas used in this array is used in the random array lab setup. Refinements to the light filtering software and the position finding software are presented, and a machine learning approach for distinguishing between different antennas is explored.