An Evaluation of the Effectiveness of Measurement-based Anomaly Detection Techniques

A number of recent studies have proposed measurement based approaches to network traffic analysis. These techniques treat traffic volume and traffic header data as signals or images in order to make analysis feasible. We use trace-driven experiments and compare the performance of different strategies. Our evaluations on real traces reveal differences in the effectiveness of different traffic header data as potential signals for traffic analysis in terms of their detection rates and false alarm rates. Our results show that address distributions and number of flows are better signals than traffic volume for anomaly detection.