Network Traffic Anomaly Detection On A Navy Network
Abstract
Navy watchstanders are ill-equipped to monitor network status in real time, to include an inability to identify network anomalies and potential risks on-the-fly. This leads to a lack of situational awareness and ultimately an inability to determine the current network risk level. An existing unsupervised machine learning technique is identified and leveraged to enable the detection of anomalous DNS network traffic on ashore-based unclassified Navy network. The research conducted by the team outlines an architecture that could be extended to produce a capability to provide the watchstander a near real-time metric of a subset of the risk that the network is experiencing by classifying DNS traffic anomalies
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 2020
- Accession Number
- AD1114636
Entities
People
- Greg T Bunder
- Michael J Laws
Organizations
- Naval Postgraduate School