Anomaly Detection Using Digital Signature of Network Segment Aiming to Help Network Management
M.L. Proença Júnior, B.B. Zarpelão, L.S. Mendes · Journal of Communication and Information Systems · 2008
The rapid and accurate identification and diagnosis of anomalies are a fundamental step for management of today’s high speed and multiservice networks. This paper presents a model for anomaly detection based on the application of BLGBA model to characterize the traffic, on three levels of sensibility alarms and on the correlation of multiples SNMP objects. The obtained results validate the experiment and show significant improvement in networks management. The main contributions of this work are: (i) case studies for traffic characterization of network servers using BLGBA model and DSNS; (ii) a model for anomaly detection; (iii) several tests of the model using real data in four network servers.