Statistical bandwidth sharing: a study of congestion at flow level

By Slim Ben Fredj, Thomas Bonald, Alexandre Proutière, G. Régnié, James W. Roberts. Conference on Applications, Technologies, Architectures, and Protocols for Computer Communication, 2001

459 citations recorded by the source corpus.

In this paper we study the statistics of the realized throughput of elastic document transfers, accounting for the way network bandwidth is shared dynamically between the randomly varying number of concurrent flows. We first discuss the way TCP realizes statistical bandwidth sharing, illustrating essential properties by means of packet level simulations. Mathematical flow level models based on the theory of stochastic networks are then proposed to explain the observed behavior. A notable result is that first order performance (e.g., mean throughput) is insensitive with respect both to the flow size distribution and the flow arrival process, as long as "sessions" arrive according to a Poisson process. Perceived performance is shown to depend most significantly on whether demand at flow level is less than or greater than available capacity. The models provide a key to understanding the effectiveness of techniques for congestion management and service differentiation.

Subfields: Computer Networks and Communications, Electrical and Electronic Engineering, Management Information Systems

View Paper: https://www.semanticscholar.org/paper/c8fe3dcf5d36dfc099ef732a1ab3b432e4750a2e

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Networks and Communications

In Electrical and Electronic Engineering

In Management Information Systems

Citations in the corpus

A scalable, commodity data center network architecture (2008, unknown): “Several studies have indicated that the distribution of transfer times and burst lengths of Internet trafﬁc is long-tailed [14], and characterized by few large long-lived ﬂows (responsible for most of the bandwidth) and many small short-lived ones [16].”
Impact of fairness on Internet performance (2001, methodology): “R E F E R E N C E S [1] S.”
Wireless downlink data channels: user performance and cell dimensioning (2003, unknown): “Admission control has already been proposed for data services in a wireline context to preserve network efficiency in overload [ 5 ].”
Less Is More: Trading a Little Bandwidth for Ultra-Low Latency in the Data Center (2012, methodology): “The M/G/1-PS model provides a good approximation for large flows for which TCP has time to converge to the fair bandwidth allocation [20].”
Sizing router buffers (2004, background): “It has been pointed out [ 23 ] that in a network with low latency, fast access links and no limit on the TCP window size, there would be very few concurrent flows.”