Statistical Network Surveillance
Daniel R. Jeske, Nathaniel T. Stevens, James David Wilson, Alexander G. Tartakovsky · Wiley StatsRef: Statistics Reference Online · 2018
Abstract The termnetwork surveillanceis defined, and the statistical methodologies that can be used as tools for network surveillance are discussed. Three examples of network surveillance are used to illustrate the diversity of contexts where the term can apply. Common to all network surveillance applications is a desire to identify the anomalous patterns in time‐varying network characteristics, such as traffic profiles, communication patterns, and fraud. As in traditional statistical process monitoring, the goal is to detect the anomalies with minimal delay while also controlling false alarms. A significant challenge associated with designing a network surveillance algorithm is how to handle monitoring streams that exhibit nonstationary behavior and have complex correlation structures. Even deciding what characteristic(s) to monitor can be a nontrivial decision.