A Longitudinal Study of Small Time Scaling Behavior of Internet Traffic

Select |


Mahanti, Anirban; Gupta, Himanshu; Ribeiro, Vinay


Conference Material

IFIP Networking 200X (Networking 200X)

Chennai, INDIA

pp. 83-95

We carry out a longitudinal study of evolution of small time- scaling behavior of Internet trac on MAWI dataset spanning 8 years. MAWI dataset contains a number of anomalies which conict in the correct identi?cation of scaling behavior, and hence to mitigate these eects, we use a sketch based procedure for robust estimation of Hurst exponent. We ?rst show the importance of robust estimation procedure while studying small time-scaling behavior of Internet trac. We further study the evolution of following properties concerning the origins of small time-scaling behavior: (1) Scaling at IP level is independent of ow ar- rivals and (2) Dense ows are primary correlation-causing factor in small time-scales. Traditionally these properties have been shown to hold by using a semi-experiments based methodology. We next show that due to network anomalies, semi-experiments can result in misleading inferences. Hence we propose and motivate the use of “robust semi-experiments? i.e. a semi-experiment coupled with the use of a robust estimation procedure for inferring scaling behavior. By making use of robust semi-experiments we ?nd the above properties to be invariant across the entire MAWI dataset. Our other results consist in showing that dense ows are car- rying a larger fraction of aggregate trac for recent traces and hence recent traces show larger short range correlations vis-a-vis earlier traces.




Mahanti, Anirban; Gupta, Himanshu; Ribeiro, Vinay. A Longitudinal Study of Small Time Scaling Behavior of Internet Traffic.[Conference Material]. Springer; 2011-05-10. <a href="http://hdl.handle.net/102.100.100/104399?index=1" target="_blank">http://hdl.handle.net/102.100.100/104399?index=1</a>

Loading citation data...

Citation counts
(Requires subscription to view)