Downloads & Free Reading Options - Results

Interference on the Global Positioning System (GPS) infrastructure poses a threat to the nation's security and economy as systems become more dependent on the technology. The pervasiveness of GPS interference methods such as jamming and spoofing present multiple opportunities for adversaries to infiltrate and inject false data on systems as diverse as military, banking, shipping, ecommerce, transportation and other critical economic sectors. The study of GPS spoofing detection methods requires innovative and novel schemes to meet the challenge posed. With the increasing processing power of computer systems, artificial intelligence methods have become a prime candidate for application to the detection and reporting of these cyber threats. This thesis studied the application of machine learning and data analytics to identify false data injection attempts on military GPS. The study combined live and simulated GPS message traffic data to train and test machine learning algorithms to identify the threats. Applying both unsupervised and supervised learning methods to the dataset helped advance the study of the GPS spoofing problem and proved to be effective tools to monitor GPS traffic while serving as another layer of security to the GPS infrastructure.