How good is stationary analysis for the transient phenomena of connection admission in ATM?

  • Christoph Herrmann
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 977)


How to decide about the admission of a new connection in ATM with exploiting statistical multiplexing is still an open question. Based on analytical queueing models describing the behaviour of a single line, the effective bandwidth of an individual source has been defined as the service rate necessary for guaranteeing a required loss probability (for a given buffer size) [18]. By a stationary analysis, it was shown for some traffic sources that the effective bandwidth is additive, i.e. the sum of the effective bandwidths of two individual sources results in the effective bandwidth of their superposition [18]. However, such results are only reasonable ”if sources come and go relatively slowly compared to the way congestion changes” [21]. In the strict sense, they are only valid, if — after the admission of a new connection — the existing traffic does not change for a very long time. Such an assumption is obviously not realistic, since there will be considerable fluctuation within the existing traffic of a link. This paper applies the method of the unfinished work of [19, 10] for obtaining transient perstream Quality of Service (QOS) parameters for a (correlated) superposition of a new connection and the existing traffic of a given link. Numerical results show that transient loss probabilities can be significantly higher than the stationary ones. This illustrates that QOS will be — at least temporarily — worse than expected through the effective bandwidth estimate, which is based on stationary computations.


Discrete-time finite buffer queue per-stream loss probabilities deterministic service time transient analysis for CAC 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    C. Blondia. A discrete-time batch Markovian arrival process as B-ISDN traffic model. Belgian Journal of Operations Research, Statistics and Computer Science, Vol. 32, No. 3,4, pp. 3–23, 1993.Google Scholar
  2. 2.
    C. Blondia and O. Casals. Statistical multiplexing of VBR sources: A matrix-analytic approach. Perf. Eval., Vol. 16, No. 1–3, pp. 5–20, Nov. 1992.CrossRefGoogle Scholar
  3. 3.
    E. Cinlar. Introduction to Stochastic Processes. Prentice Hall, New Jersey, 1st edition, 1975.Google Scholar
  4. 4.
    W. Ding and P. Decker. Waiting time distribution of a discrete SSMP/G/1 queue and its implications in ATM systems. In 7th ITC Spec. Sem., paper 9.2, Morristown, USA, Oct. 9–11 1990.Google Scholar
  5. 5.
    A. I. Elwalid, D. Mitra. Effective Bandwidth of General Markovian Traffic Sources and Admission Control of High Speed Networks. In IEEE INFOCOM93, San Francisco, pp. 256–265, March 31–April 2, 1993.Google Scholar
  6. 6.
    L. Fahrmeir, H. Kaufmann, F. Ost. Stochastische Processe. Hanser, 1980.Google Scholar
  7. 7.
    C. Herrmann. Correlation Effect on Per-stream QOS Parameters of ATM Traffic Superpositions relevant to Connection Admission Control. In IEEE ICC'93, pp. 1027–1031, Geneva, May 23–26, 1993.Google Scholar
  8. 8.
    C. Herrmann. A Performance Model for Statistical Multiplexing of correlated ATM Traffic Superpositions. In 7. ITG/GI Fachtagung: MMB'93, paper 4A.2, Aachen, Sept. 21.–23., 1993, Springer-Verlag.Google Scholar
  9. 9.
    C. Herrmann. Discrete-time Models for Connection Admission Control (CAC) in ATM. In 2nd IFIP Workshop on Performance Modelling and Evaluation of ATM Networks, Bradford, UK, July 4–6, 1994.Google Scholar
  10. 10.
    C. Herrmann. Local Correlation Properties of some Discrete-time Finite Queues with Semi-Markovian Input. Archiv f. Elektronik und übertragungstechnik, Vol. 48, No. 5, pp. 244–255, Sept. 1994.Google Scholar
  11. 11.
    C. Herrmann. Stochastische Modelle für ATM-Konzepte. PhD thesis, Verlag der Augustinus Buchhandlung, Aachen, Germany, ISBN 3-860-73380-X, Jan. 1995.Google Scholar
  12. 12.
    J. J. Hunter. Mathematical Techniques of Applied Probability, Vol. 2, Discrete Time Models: Techniques and Applications. Academic Press, New York, 1983.Google Scholar
  13. 13.
    F.P. Kelly. Effective bandwidths at multi-class queues. Queueing Systems, Vol. 9, pp. 5–15, 1991.CrossRefGoogle Scholar
  14. 14.
    D.M. Lucantoni. New Results on the single Server Queue with a Batch Markovian Arrival Process. Stochastic Models, Communications in Statistics, Vol. 7, No. 1, pp. 1–46, 1991.Google Scholar
  15. 15.
    M. Murata et al. Analysis of a Discrete-Time Single-Server Queue with Bursty Inputs for Traffic Control in ATM Networks. IEEE J-SAC, Vol. 8, No. 3, pp. 447–458, April 1990.Google Scholar
  16. 16.
    Marcel F. Neuts. Structured Stochastic Matrices of M/G/1 Type and their Applications. Marcel Dekker, New York, 1989.Google Scholar
  17. 17.
    R. Pyke. Markov Renewal processes with finitely many states. Ann. Math. Stat., Vol. 32, pp. 1243–1259, 1961.Google Scholar
  18. 18.
    R. Guerin, H. Ahmadi, M. Naghshineh. Equivalent Capacity and Its Application to Bandwidth Allocation in High-Speed Networks. IEEE J. Sel. Areas Comm., Vol. 9, No. 7, pp. 968–981, 1991.CrossRefGoogle Scholar
  19. 19.
    P. Tran-Gia, Hamid Ahmadi. Analysis of a discrete-time G [X] /D/1-S queueing system with applications in packet-switching systems. In IEEE INFOCOM'88, paper 9A.1.10, March 27–31, 1988.Google Scholar
  20. 20.
    W. Whitt. Approximating a point process by a renewal process, I.: two basic methods. Operations Research, Vol. 30, No. 1, pp. 125–147, 1982.Google Scholar
  21. 21.
    W. Whitt. Tail probabilities with statistical multiplexing and effective bandwidth in multi-class queues. Telecommunication Systems, Vol. 2, No. 1, pp. 71–107, 1993.CrossRefGoogle Scholar
  22. 22.
    J. Zhang. Performance Study of Markov Modulated Fluid Flow Models with Priority Traffic. In IEEE INFOCOM'93, San Francisco, March, 28–April, 1 1993.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1995

Authors and Affiliations

  1. 1.Communication NetworksAachen University of TechnologyAachenGermany

Personalised recommendations