Annals of Operations Research

, Volume 49, Issue 1, pp 241–278 | Cite as

A universal building block for the approximate analysis of a shared buffer ATM switch architecture

  • Demetres D. Kouvatsos
  • Spiros G. Denazis


A universal analytic approximation is proposed for the performance analysis of a general queueing model of a shared buffer ATM switch architecture with bursty arrivals. The forms of the joint, aggregate and marginal state probabilities are characterised via entropy maximisation. As an application, a continuous-time maximum entropy (ME) solution is implemented at equilibrium by assuming that the arrival process to each port of the ATM switch is modelled by a Compound Poisson Process (CPP) with geometrically distributed batches. Consequently, efficientz-transform-type recursive expressions of low computational cost are derived. Validation tests against simulation show that the ME approximation is credible with a very good error-level. Moreover, performance bounds for the mean queue length and cell-loss probability at each output port are experimentally defined over those generated by Interrupted Poisson Processes (IPPs) having the same first two interarrival-time moments.


Poisson Process Performance Bound Maximum Entropy Queue Length Validation Test 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    F.A. Tobagi, Fast packet switch architectures for broadband integrated services networks, Proc. IEEE 78 (1990) 1133–1167.Google Scholar
  2. [2]
    M. Devault, J.-Y. Cochennec and M. Servel, The “Prelude” ATD experiment: Assessments and future prospects, IEEE J. Sel. Areas Comm. SAC-6(1988) 1528–1537.Google Scholar
  3. [3]
    P. Boyer, M.R. Lehnert and P.J. Kuehn, Queueing in an ATM basic switch element, Technical Report CNET-123-030-CD-CC, CNET, France (1988).Google Scholar
  4. [4]
    H. Kuwahara, N. Endo, M. Ogino and T. Kozaki, A shared buffer memory switch for an ATM exchange,Int. Conf. on Communications, Boston, MA (June 1989) pp. 441–448.Google Scholar
  5. [5]
    H. Lee, K. Kook, C.S. Rim, K. Jun and S.-K. Lim, A limited shared output buffer switch for ATM,4th Int. Conf. on Data Communication Systems and Their Performance, Barcelona (1990) pp. 163–179.Google Scholar
  6. [6]
    A.E. Eckberg and T.-C. Hou, Effects of output buffer sharing on buffer requirements in an ATDM packet switch,INFOCOM '88 (March 1988) pp. 459–466.Google Scholar
  7. [7]
    G.H. Petit and E.M. Desmet, Performance evaluation of shared buffer multiserver output queue switches used in ATM,7th ITC Specialist Seminar, New Jersey (1990) paper 7.1.Google Scholar
  8. [8]
    I. Iliadis, Performance of a packet switch with shared buffer and input queueing, in:Teletraffic and Datatraffic in a Period of Change, ITC-13, eds. A. Jensen and V.B. Iversen (North-Holland, 1991) pp. 911–916.Google Scholar
  9. [9]
    P.G. Harrison and A. de C. Pinto, Blocking in asynchronous, buffered Banyan networks,Proc. Int. Seminar, Kyoto, Japan (December 1991).Google Scholar
  10. [10]
    H. Yamashita, H.G. Perros and S. Hong, Performance modelling of a shared buffer ATM switch architecture, in:Teletraffic and Datatraffic in a Period of Change, eds. A. Jensen and V.B. Iversen (North-Holland, 1991).Google Scholar
  11. [11]
    S. Hong, H.G. Perros and H. Yamashita, A discrete-time queueing model of the shared buffer ATM switch with bursty arrivals, Research Report, Computer Science Dept., North Carolina State University (1992).Google Scholar
  12. [12]
    H.G. Perros, private communication (1992).Google Scholar
  13. [13]
    E.T. Jaynes, Information theory and statistical mechanics, Phys. Rev. 106 (1957) 620–630.Google Scholar
  14. [14]
    D.D. Kouvatsos, A maximum entropy analysis of theG/G/1 queue at equilibrium, J. Oper. Res. Soc. 39 (1989) 183–200.Google Scholar
  15. [15]
    D.D. Kouvatsos and N.P. Xenios, MEM for arbitrary queueing networks with multiple general servers and repetitive-service blocking, Perf. Eval. 10 (1989) 169–195.Google Scholar
  16. [16]
    H.G. Perros and R. Onvural, On the superposition of arrival processes for voice and data,4th Int. Conf. on Data Communication Systems and Their Performance, Barcelona (June 1990) pp. 341–357.Google Scholar
  17. [17]
    F. Kamoun and L. Kleinrock, Analysis of shared finite storage in a computer network node environment under general traffic conditions, IEEE Trans. Comm. COM-28 (1980).Google Scholar
  18. [18]
    A.C. Williams and R.A. Bhandiwad, A generating function approach to queueing network analysis of multiprogrammed computers, Networks 6 (1976) 1–22.Google Scholar
  19. [19]
    D.D. Kouvatsos, Maximum entropy and theG/G/1/N queue, Acta Inf. 23 (1986) 545–565.Google Scholar
  20. [20]
    C. Sauer, Configuration of computing systems: An approach using queueing network models, Ph.D. Thesis, University of Texas (1975).Google Scholar
  21. [21]
    S. Nojo and H. Watanabe, A new stage method getting arbitrary coefficient of variation by two stages,Trans. IEICE '70 (1987) 33–36.Google Scholar
  22. [22]
    M. Veran and D. Potier, QNAP-2: A portable environment for queueing network modelling, in:Modelling Techniques and Tools for Performance Analysis (North-Holland, 1985) pp. 25–63.Google Scholar
  23. [23]
    K.M. Chandy, U. Herzog and L. Woo, Approximate analysis of general queueing networks, IBM J. Res. Develop. 19 (1975).Google Scholar
  24. [24]
    F.P. Kelly,Reversibility and Stochastic Networks (Wiley, 1979).Google Scholar
  25. [25]
    D.D. Kouvatsos and N.M. Tabet-Aouel, GGeo-type approximations for general discrete-time queueing systems, IFIP Trans. C-15, Special issue on Modelling and Performance Evaluation of ATM Technology (1993) 469–483.Google Scholar
  26. [26]
    D.D. Kouvatsos, Entropy maximisation and queueing network models, Ann. Oper. Res. 48 (1994) 63–126.Google Scholar

Copyright information

© J.C. Baltzer AG, Science Publishers 1994

Authors and Affiliations

  • Demetres D. Kouvatsos
    • 1
  • Spiros G. Denazis
    • 1
  1. 1.Computer Systems Modelling Research GroupUniversity of BradfordBradfordEngland

Personalised recommendations