Improving network anomaly detection via selective flow-based sampling
Authors:
- 1 March 2008
- journal article
- Published by Institution of Engineering and Technology (IET) in IET Communications
- Vol. 2 (3) , 399-409
Abstract
Sampling has become an essential component of scalable Internet traffic monitoring and anomaly detection. A new flow-based sampling technique that focuses on the selection of small flows, which are usually the source of malicious traffic, is introduced and analysed. The proposed approach provides a flexible framework for preferential flow sampling that can effectively balance the tradeoff between the volume of the processed information and the anomaly detection accuracy. The performance evaluation of the impact of selective flow-based sampling on the anomaly detection process is achieved through the adoption and application of a sequential non-parametric change-point anomaly detection method on realistic data that have been collected from a real operational university campus network. The corresponding numerical results demonstrate that the proposed approach achieves to improve anomaly detection effectiveness and at the same time reduces the number of selected flows.Keywords
This publication has 13 references indexed in Scilit:
- Impact of Packet Sampling on Portscan DetectionIEEE Journal on Selected Areas in Communications, 2006
- Inverting sampled trafficIEEE/ACM Transactions on Networking, 2006
- Estimating flow distributions from sampled flow statisticsIEEE/ACM Transactions on Networking, 2005
- Change-point monitoring for the detection of DoS attacksIEEE Transactions on Dependable and Secure Computing, 2004
- Adaptive packet sampling for accurate and scalable flow measurement2004
- Multivariate statistical analysis of audit trails for host-based intrusion detectionIEEE Transactions on Computers, 2002
- A signal analysis of network traffic anomalies2002
- Characteristics of network traffic flow anomalies2001
- Effective traffic measurement using ntopIEEE Communications Magazine, 2000
- The NLANR Network Analysis InfrastructureIEEE Communications Magazine, 2000