VIPNANO: Monitoring of Virtual Private Cloud Networks for Automated Anomaly Detection

DSpace Repositorium (Manakin basiert)


Dateien:

Zitierfähiger Link (URI): http://hdl.handle.net/10900/163783
http://nbn-resolving.org/urn:nbn:de:bsz:21-dspace-1637831
http://dx.doi.org/10.15496/publikation-105113
Dokumentart: Wissenschaftlicher Artikel
Erscheinungsdatum: 2025-04-03
Sprache: Englisch
Fakultät: 7 Mathematisch-Naturwissenschaftliche Fakultät
Fachbereich: Informatik
DDC-Klassifikation: 004 - Informatik
Zur Langanzeige

Abstract:

Anomaly detection in enterprise networks is crucial for cybersecurity, system monitoring, and identifying outages. Despite extensive academic research, practical deployment of proposed mechanisms remains rare. The VIPNANO project investigates key shortcomings in academic approaches, focusing on two major obstacles: (1) reliance on unrealistic datasets that fail to reflect real-world complexity, and (2) overly complex machine learning models with impractical computational overhead. Additionally, we highlight a critical gap – the lack of rigorous real-world validation. Through systematic analysis, we emphasize the need to prioritize realistic data, scalability, and verifiable solutions to bridge the gap between theory and deployment.

Das Dokument erscheint in: