Traffic Anomaly Detection

This book presents an overview of traffic anomaly detection analysis, allowing you to monitor security aspects of multimedia services. The author's approach is based on the analysis of time aggregation adjacent periods of the traffic. As traffic varies throughout the day, it is essential to consider the concrete traffic period in which the anomaly occurs. This book presents the algorithms proposed specifically for this analysis and an empirical comparative analysis of those methods and settle a new information theory based technique, named "typical day analysis".

A new information-theory based technique for traffic anomaly detection (typical day analysis)Introductory chapters to anomaly detection methods including control charts, tests of goodness-of-fit Mutual InformationContains comparative analysis of traffic anomaly detection methods

1. Theoretical anomaly detection methods. Set of algorithms proposed for this analysis: the most used SCC (CUSUM), the two main tests of goodness-of-fit and Mutual Information. 2. Finding the optimal aggregation period for a time series of Internet traffic 3. Comparative analysis of traffic anomaly detection methods 4. Proposal of a new information-theory based technique (typical day analysis) 5. Conclusions