Document Type

Honors Thesis

Date of Award

Spring 2024

Abstract

Military systems require monitoring a wide frequency range of the RF spectrum to detect and monitor enemy targets. A vital component of that is calculating the direction of arrival of enemy radars. The use of compressive sensing architectures like the Nyquist Folding Receiver (NyFR) allow for computationally cheap capturing of wideband signals; however, these techniques make traditional direction of arrival (DOA) estimation techniques difficult. Given the advantages of compressive sensing, this research aims to apply machine learning (ML) techniques to multi-arm spiral antennas to estimate the incoming DOA. Software like MATLAB was used to simulate the multi-arm spiral antenna and NyFR, allowing training of an ML algorithm using freely available tools like PyTorch.

Advisor

Gerald Fudge

Share

COinS