Sample Based Estimation of Network Traffic Flow Characteristics.

Sampling;Design;Network;Statistics and Numeric Data;Science;Statistics

Yang, LiliStoev, Stilian Atanasov ;

University of Michigan

关键词: Sampling;    Design;    Network;    Statistics and Numeric Data;    Science;    Statistics;   

Others  :  https://deepblue.lib.umich.edu/bitstream/handle/2027.42/62382/yanglili_1.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

Understanding the characteristics of traffic flows is crucial for allocating the necessary resources (bandwidth) to accommodate usersdemand. In this dissertation research, the problem of nonparametric and Moment-Based estimations of network flow characteristics basedon sampled flow data from single-stage Bernoulli sampling, and two-stage sampling will be addressed. An Expectation-maximization(EM) algorithm is used for the flow length distribution, which in addition provides an estimate for the number of active flows. Theestimation of the flow sizes (in bytes) is accomplished through a regression model. A variation of this approach, particularly suitedfor mixture distributions that appear in real network traces, is also considered. Lastly, estimation of traffic characteristic acrossthe network and sampling allocation problem is studied. The proposed approaches are illustrated and compared on a number of synthetic andreal data sets.