Traffic Anomaly Detection by Antonio Cuadra-Sánchez, Javier Aracil

By Antonio Cuadra-Sánchez, Javier Aracil

This publication provides an outline of site visitors anomaly detection research, permitting you to watch safeguard facets of multimedia providers. The author's method is predicated at the research of time aggregation adjoining classes of the site visitors. As site visitors varies in the course of the day, it really is necessary to ponder the concrete site visitors interval within which the paradox happens. This ebook provides the algorithms proposed in particular for this research and an empirical comparative research of these tools and settle a brand new details thought dependent process, named "typical day analysis".

  • A new information-theory established method for site visitors anomaly detection (typical day analysis)
  • Introductory chapters to anomaly detection tools together with keep watch over charts, exams of goodness-of-fit Mutual Information
  • Contains comparative research of site visitors anomaly detection methods

Show description

Read or Download Traffic Anomaly Detection PDF

Best industrial engineering books

Molecular Techniques in the Microbial Ecology of Fermented Foods (Food Microbiology and Food Safety)

With the appliance of recent analytical ideas, the sector of meals fermentation has grown in recent times. This e-book offers the newest details and correct advances at the microbial ecology of fermented meals and the applying of molecular tools. This e-book serves as a consultant for college students and researchers at the such a lot complicated innovations to spot micro organism and is helping in opting for the main acceptable instruments to review fermented foodstuff from a microbiological standpoint.

Managing Maintenance Shutdowns and Outages

Making plans a shutdown? decrease your expenditures by means of examining and utilizing this exact reference! initially designed as a radical direction after which honed over a number of years, coping with upkeep Shutdowns and Outages organizes the main points of a upkeep shutdown right into a attainable shape by means of introducing all of the steps wanted for a winning shutdown or outage.

Chemistry and technology of polyols for polyurethanes

Polyurethanes are some of the most dynamic teams of polymers, they locate use in approximately each element of contemporary lifestyles, in functions equivalent to furnishings, bedding, seating and software panels for automobiles, shoe soles, thermoinsulation, carpet backings, packaging, adhesives, sealants, binders and as coatings.

Additional info for Traffic Anomaly Detection

Example text

3. Traffic contaminated evenhandedly and weighed. 1. Statistical control charts The state of the art shows that there is no single SCC algorithm that stands out from the others, but the performance of SCC depends on the underlying traffic characteristics [MAR 11, MAT 11, CAR 12, OPR 13, BUL 12]. g. EWMA) would lead to equivalent results. 38 Traffic Anomaly Detection Control charts perform measurements obtained as variations from the expected value by the standard deviation. In particular, CUSUM (CUMulative SUM) is a sequential analysis algorithm that allows us to monitor sudden changes in continuous processes, such as changes in traffic.

They present a fault diagnosis procedure based on discriminant analysis and MI. In order to obtain good classification performances, a selection of important features is done with a newly developed algorithm based on the MI between variables. Nevertheless, they do not use the MI theory as a criterion for fault detection. In the same way, in [AMI 11], the authors present MI-based feature selection for intrusion detection systems. The usefulness of MI for selecting the most relevant features in a given classification task has been proven in other fields [DRU 14].

5) differs from the first scenario in that the increased traffic in those periods is not the same but 40% higher than the maximum average inside each period (75 KBytes/s at 4am, 425 KBytes/s at 12am and 350 KBytes/s at 5pm). MI detects the last two periods (20 and 40%), CUSUM detects the intermediate period (100% increase) and χ2 detects the first period (30% higher). 6), we have notably decreased the traffic 40% lower than the minimum of the entire series for the three periods evenhandedly (1 KByte/s).

Download PDF sample

Rated 4.92 of 5 – based on 12 votes