Non-Gaussian and Long Memory Statistical Characterizations for Internet Traffic with Anomalies

Scherrer, Antoine; Larrieu, Nicolas; Owezarski, Philippe; Borgnat, Pierre; Abry, Patrice

doi:10.1109/tdsc.2007.12

Cited by 125 publications

(97 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The anomalies can be observed from single links or network-wide data. Standard references include [3] [1] [9] [11], with some notable recent work as [13] [5] [4] [16]. Dimensionality reduction of aggregated traffic data has also received recent attention, and techniques like sketches [9] [13] [5] and principal components analysis [11] are very promising for online anomaly detection.…”

Section: Related Workmentioning

confidence: 99%

0day Anomaly Detection Made Possible Thanks to Machine Learning

Owezarski

Mazel

Labit

2010

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

Abstract. This paper proposes new cognitive algorithms and mechanisms for detecting 0day attacks targeting the Internet and its communication performances and behavior. For this purpose, this work relies on the use of machine learning techniques able to issue autonomously traffic models and new attack signatures when new attacks are detected, characterized and classified as such. The ultimate goal deals with being able to instantaneously deploy new defense strategies when a new 0day attack is encountered, thanks to an autonomous cognitive system. The algorithms and mechanisms are validated through extensive experiments taking advantage of real traffic traces captured on the Renater network as well as on a WIDE transpacific link between Japan and the USA.

show abstract

Section: Related Workmentioning

confidence: 99%

0day Anomaly Detection Made Possible Thanks to Machine Learning

Owezarski

Mazel

Labit

2010

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

show abstract

“…We have developed, for several common laws (exponential, gamma, χ 2 , etc. ), an exact method of synthesis described in [28]. We can thus produce synthetic long-range dependent sample paths that can be used, as it will be shown in Section 4, in order to evaluate the impact of lrd on the performance of an on-chip network.…”

Section: Synthesis Of Long-range Dependent Processesmentioning

confidence: 99%

Long-range dependence and on-chip processor traffic

Scherrer

Fraboulet

Risset

2009

Microprocessors and Microsystems

Self Cite

View full text Add to dashboard Cite

Long-range dependence is a property of stochastic processes that has an important impact on network performance, especially on the buffer usage in routers. We analyze the presence of long-range dependence in on-chip processor traffic and we study the impact of long-range dependence on networks-on-chip. We propose to investigate the presence of long-range dependence in communication traces of processor ips at the cycle-accurate level. We also study the impact of long-range dependence on a real network-on-chip using the SocLib simulation environment and traffic generators of our own. Our experiments show that long-range dependence is not an ubiquitous property of on-chip processor traffic and that its impact on the network-on-chip is highly correlated with the low level communication protocol used.

show abstract

“…However, these results have been recently called to be far unrealistic [18]. Finally, and only for the sake of further highlighting diversity of findings, we cite [22] which recently recommended the Gamma distribution as the "best choice on average" for a comprehensive set of recorded traffic traces.…”

Section: A General Model For Measurement-based Admission Controlmentioning

confidence: 99%

“…Observant readers might have noticed that (22) poses an overflow model rather than a loss model. Consequently, the algorithm does not differentiate between a one packet and 100-packet loss but does consider both cases as an equal overflow.…”

Section: Control Accuracy Evaluationmentioning

confidence: 99%

Non-parametric and Self-tuning Measurement-Based Admission Control

Bohnert

Monteiro

Koucheryavy

et al. 2007

NETWORKING 2007. Ad Hoc and Sensor Networks, Wireless Networks, Next Generation Internet

View full text Add to dashboard Cite

Abstract. The Measurement-based Admission Control algorithm presented in this paper has been devised to overcome three shortcomings. Firstly, its configuration parameter strictly correspond to standard QoS definitions in Service Level Agreements, namely packet loss probability. While this is featured by alternative designs too, the second issue, that of considerable performance fluctuations under varying traffic conditions, is a rather general problem. Applying a non-parametric approach the presented algorithm's estimation model is free from assumptions and simulations assert consistent performance for a set of various conditions. Finally, the third improvement is certainly most appealing, the algorithm's independence from fine-tuning. In fact, the optimal value of the performance determining parameter is being estimated from the measured sample. This makes the algorithm highly adaptive as well as autonomous and simulations confirm near optimal performance and accuracy.

show abstract

Non-Gaussian and Long Memory Statistical Characterizations for Internet Traffic with Anomalies

Cited by 125 publications

References 39 publications

0day Anomaly Detection Made Possible Thanks to Machine Learning

0day Anomaly Detection Made Possible Thanks to Machine Learning

Long-range dependence and on-chip processor traffic

Non-parametric and Self-tuning Measurement-Based Admission Control

Contact Info

Product

Resources

About