Presentation 2001/6/14
Nonparametric Effective Bandwidth Estimation for a Variable-Length-Packet Flow
Shigeo Shioda, Daisuke Satoh,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We develop a new method for estimating "effective bandwidth" of a variable-length-packet flow. The effective bandwidth is the minimum bandwidth that should be allocated to a packet flow for making the buffer overflow probability below a certain quality-of-service objective, Our method needs only a small set of measurements (buffer-occupancy and traffic volume measurements) for estimation and does not rely on any specific traffic model. We analytically show that our proposal has several desirable features (for example, being a kind of non-biased estimator) for IP-network traffic engineering. We provide numerical results to show that the estimation by our method is much more accurate than that of the similar existing proposal.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Internet / traffic / Effective Bandwidth / Heavy Traffic Approximation / Large Deviation Principle / QoS
Paper # IN2001-20
Date of Issue

Conference Information
Committee IN
Conference Date 2001/6/14(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Vice Chair

Paper Information
Registration To Information Networks (IN)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Nonparametric Effective Bandwidth Estimation for a Variable-Length-Packet Flow
Sub Title (in English)
Keyword(1) Internet
Keyword(2) traffic
Keyword(3) Effective Bandwidth
Keyword(4) Heavy Traffic Approximation
Keyword(5) Large Deviation Principle
Keyword(6) QoS
1st Author's Name Shigeo Shioda
1st Author's Affiliation Department of Urban Environment Systems, Chiba University()
2nd Author's Name Daisuke Satoh
2nd Author's Affiliation NTT Service Integration Labratories
Date 2001/6/14
Paper # IN2001-20
Volume (vol) vol.101
Number (no) 122
Page pp.pp.-
#Pages 8
Date of Issue