Date of Award
Master of Science (MS)
College of Science and Mathematics
Thesis Sponsor/Dissertation Chair/Project Chair
Existing network detection techniques rely on SSIDs, network patterns or MAC addresses of genuine wireless devices to identify malicious attacks on the network. However, these device characteristics can be manipulated posing a security threat to information integrity, lowering detection accuracy, and weakening device protection. This research study focuses on empirical analysis to elaborate the relationship between received signal strength (RSSI) and distance; investigates methods to detect rogue devices and access points on Wi-Fi networks using network traffic analysis and fingerprint identification methods. In this paper, we conducted three experiments to evaluate the performance of RSSI and clock skews as features to detect rogue devices for indoor and outdoor locations. Results from the experiments suggest different devices connected to the same access point can be detected (p < 0.05) using RSSI values. However, the magnitude of the difference was not consistent as devices were placed further from the same access point. Therefore, an optimal distance for maximizing the detection rate requires further examination. The random forest classifier provided the best performance with a mean accuracy of 79% across all distances. Our experiment on clock skew shows improved accuracy in using beacon timestamps to detect rogue APs on the network.
Chege, Daniel, "Time of Flight and Fingerprinting Based Methods for Wireless Rogue Device Detection" (2021). Theses, Dissertations and Culminating Projects. 733.