Date of Award
3-2020
Document Type
Thesis
Degree Name
Master of Science in Cyber Operations
Department
Department of Electrical and Computer Engineering
First Advisor
Barry E. Mullins, PhD
Abstract
The proposed security agent, Internet of Things Active Management Unit (IoTAMU), provides confidentiality of IoT networks via the following capabilities: (1) authentication, (2) firewall, (3) encryption, and (4) spoofing. To test the spoofer's effect, an Identical Device Model Classifier (IDMC) is developed, which measures the similarities of the observed network signatures of each pair of devices, and recognize identical model devices. The IDMC performs well in baseline network settings without the spoofer, achieving 100% precision, recall, and specificity at high threshold (SS>0.9). When the spoofer is enabled, none of the identical pairs are identified at high threshold, and up to 66% identical pairs are identified at lower thresholds (SS>0.8, 0.7). Overall, the spoofer is able to sufficiently modify the observed network signatures of each device; the observed differences between each pair increase overall (p-value = 0.01132) at 120 spoofed samples, making it more difficult to identify similar devices. Finally, the experiments in this work show the spoofer has a negligible effect on network congestion.
AFIT Designator
AFIT-ENG-MS-20-M-053
DTIC Accession Number
AD1104615
Recommended Citation
Park, Youngjun, "Development and Evaluation of a Security Agent for Internet of Things" (2020). Theses and Dissertations. 3621.
https://scholar.afit.edu/etd/3621