ABSTRACT
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.
- 1.E.Altman,C.Barakat,and K.Avrachenkov.A stochastic model of tcp/ip with stationary random losses.In Proc.of ACM SIGCOMM '00 2000. Google ScholarDigital Library
- 2.F.Baskett,K.M.Chandy R.R.Muntz,and F.G. Pallacios.Open,closed and mixed networks of queues with di .erent classes of customers.Journal of ACM 22 pages 248 260,1975. Google ScholarDigital Library
- 3.S.Ben Fredj,S.Oueslati-Boulahia,and J.W.Roberts. Measurement-based admission control for elastic tra .c.In Proceedings of ITC17 September 2001.Google Scholar
- 4.A.Berger and Y.Kogan.Dimensioning bandwidth for elastic tra .c in high-speed data networks. IEEE/ACM Trans Networks 8(5):643 654,October 2000. Google ScholarDigital Library
- 5.D.Bertsekas and R.Gallager.Data Networks . Prentice-Hall International,1992. Google ScholarDigital Library
- 6.T.Bonald and L.Massouli~.Impact of fairness on internet performance.In Proceedings of ACM SIGMETRICS '01 2001. Google ScholarDigital Library
- 7.T.Bonald,A.Prouti~re,G.R~gni~,and J.W.Roberts.Insensitivity results in statistical bandwidth sharing.In Proceedings of ITC17 September 2001.Google Scholar
- 8.T.Bu and D.Towsley.Fixed point approximations for TCP behavior in an aqm network.In Proceedings of ACM SIGMETRICS '01 2001. Google ScholarDigital Library
- 9.J.W.Cohen.The multiple phase service network with generalized processor sharing.In Acta Informatica pages 245 284,1979.Google Scholar
- 10.M.Crovella and A.Bestravos.Self-similarity in world wide web tra .c:Evidence and possible cause.In Proceedings of ACM SIGMETRICS '96 1996. Google ScholarDigital Library
- 11.G.De Veciana,T.J.Lee,and T.Konstantopoulos. Stability and performance analysis of networks supporting services with rate control -could the internet be unstable?In Proceedings of IEEE INFOCOM '99 1999.Google Scholar
- 12.D.P.Heyman,T.V.Lakshman,and A.L.Neidhardt. New method for analyzing feedback protocols with applications to engineering web tra .c over the internet.In Proceedings of ACM SIGMETRICS '97 1997. Google ScholarDigital Library
- 13.G.Fayolle,A.de la Fortelle,J-M.Lasgouttes, L.Massouli~,and J.W.Roberts.Best-e .ort networks: modeling and performance analysis via large networks asymptotics.In Proceedings of IEEE INFOCOM '01 2001.Google Scholar
- 14.G.Fayolle,I.Mitrani,and R.Iasnogorodski.Sharing a processor among many classes.Journal of the ACM 27:519 -532,1980. Google ScholarDigital Library
- 15.A.Feldman.Characteristics of TCP connection arrivals.K.Park,W.Willinger,editors,Self-similar network tra .c and performance evaluation J.Wiley &Sons,2000.Google Scholar
- 16.A.Feldman,A.Gilbert,P.Huang,and W.Willinger. Data networks as cascades:Explaining the multifractal nature of internet WAN tra .c.In Proceedings of ACM SIGCOMM '98 1998. Google ScholarDigital Library
- 17.A.Feldman,A.Gilbert,P.Huang,and W.Willinger. Dynamics of IP tra .c:A study of the role of variability and the impact of control.In Proceedings of ACM SIGCOMM '99 1999. Google ScholarDigital Library
- 18.A.Jean-Marie and P.Robert.On the transient behavior of the processor sharing queue.Queueing Systems Theory and Applications 17:129 -136,1994.Google Scholar
- 19.F.P.Kelly.Reversibility and Stochastic Networks .J. Wiley &Sons,1979.Google Scholar
- 20.F.P.Kelly A.Maulloo,and D.Tan.Rate control for communication networks:shadow prices,proportional fairness and stability.Journal of the Operational Research Society 49,1998.Google ScholarCross Ref
- 21.A.A.Kherani and A.Kumar.Performance analysis of TCP with nonpersistent sessions.Preprint available at http://ece.iisc.ernet.in/anurag/Anurag Kumar.html, 2000.Google Scholar
- 22.L.Kleinrock.Queueing Systems,Volume 2 .J.Wiley &Sons,1975.Google Scholar
- 23.L.Massouli~ and J.W.Roberts.Arguments in favor of admission control for TCP .ows.In P.Key and D.Smith,editors,Teletra .c Engineering in a Competitive Worl ,Proceedings of ITC 16 ,pages 33 -44.Elsevier,1999.Google Scholar
- 24.L.Massouli~ and J.W.Roberts.Bandwidth sharing: objectives and algorithms.In Proceedings of IEEE INFOCOM '99 1999.Google Scholar
- 25.L.Massouli~ and J.W.Roberts.Bandwidth sharing and admission control for elastic tra .c. Telecommunication Systems 15:185 -201,2000.Google Scholar
- 26.J.Mo and J.Walrand.Fair end-to-end window-based congestion control.In Proceedings of SPIE '98 International Symposium on Voice,Video and Data Communications 1998.Google Scholar
- 27.C.J.Nuzman,I.Saniee,W.Sweldens,and A.Weiss. A compound model of tcp arrivals.In Proceedings of ITC Seminar on IP Tra .c Modeling Monterey 2000.Google Scholar
- 28.V.Paxson and S.Floyd.Wide-area tra .c:The failure of poisson modeling.IEEE/ACM Trans. Networking 3(3):226 -255,1995. Google ScholarDigital Library
- 29.J.W.Roberts and S.Oueslati-Boulahia.Quality of service by .ow-aware networking.Phil.Trans.R. Soc.London A 358:2197 -2207,2000.Google Scholar
- 30.M.Vojnovic,J.-Y.Le Boudec,and C.Boutremans. Global fairness of additive-increase and multiplicative-decrease with heterogeneous roundtrip times.In Proceedings of IEEE INFOCOM '00 2000.Google Scholar
Index Terms
- Statistical bandwidth sharing: a study of congestion at flow level
Recommendations
Statistical bandwidth sharing: a study of congestion at flow level
Proceedings of the 2001 SIGCOMM conferenceIn 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 ...
Fair bandwidth sharing using Shapley value
EATIS '08: Proceedings of the 2008 Euro American Conference on Telematics and Information SystemsIn the Internet, applications with very distinct requirements share the same network resources. In order to avoid QoS degradation, the resource manager component from QoS mechanism should distribute the available bandwidth in efficient and fair way. ...
Comments