Date of Award

3-1-2018

Document Type

Thesis

Degree Name

Master of Science in Computer Science

Department

Department of Electrical and Computer Engineering

First Advisor

Kenneth M. Hopkinson, PhD.

Abstract

Decreased budgets have pushed the United States Air Force towards using existing systems in new ways. The use of unmanned aerial vehicle swarms is one example of reuse of existing systems. One problem with the increased utilization of these swarms is the congestion of the electromagnetic spectrum. Software-defined or cognitive radios have been proposed as a basis for a potential robust communications solution. The present research aims to develop and test a genetic algorithm-based cognitive engine to begin looking at real-time engines that could be used in future swarms. Here, latency is the optimization objective of primary importance. In testing the engine, particular items of interest include the number of solutions evaluated in a given bound and the engine's reliability in yielding acceptable network performance. Initial experiments indicate the engine can consider significant portions of the search space within a relatively small bound and that the engine is efficient at finding highly fit solutions. Future work for this research includes evaluating how well high fitness correlates to acceptable performance and testing the engine with additional noise floors.

AFIT Designator

AFIT-ENG-MS-18-M-032

DTIC Accession Number

AD1056151

Download

Share

COinS