Department of Computer Science
We propose networkmetrics, a new data-driven approach for monitoring, troubleshooting and understanding communication networks using multivariate analysis. Networkmetric models are powerful machine-learning tools to interpret and interact with data collected from a network. In this paper, we illustrate the application of Multivariate Big Data Analysis (MBDA), a recently proposed networkmetric method with application to Big Data sets. We use MBDA for the detection and troubleshooting of network problems in a campus-wide Wi-Fi network. Data includes a seven-year trace (from 2012 to 2018) of the network’s most recent activity, with approximately 3,000 distinct access points, 40,000 authenticated users, and 600,000 distinct Wi-Fi stations. This is the longest and largest Wi-Fi trace known to date. To analyze this data, we propose learning and visualization procedures that extend MBDA. These procedures result in a methodology that allows network analysts to identify problems and diagnose and troubleshoot them, optimizing the network performance. In the paper, we go through the entire workflow of the approach, illustrating its application in detail and discussing processing times for parallel hardware.
José Camacho, Rasmus Bro, and David Kotz. Networkmetrics unraveled: MBDA in Action. Technical Report number 1907.02677, arXiv, July 2019.
Dartmouth Digital Commons Citation
Camacho, José; Bro, Rasmus; and Kotz, David, "Networkmetrics unraveled: MBDA in Action" (2019). Other Faculty Materials. 5.