Annals of Operations Research

, Volume 48, Issue 1, pp 63–126

Entropy maximisation and queueing network models

  • Demetres D. Kouvatsos
Exact Analysis

Abstract

Over recent years it has become increasingly evident that “classical” queueing theory cannot easily handle complex queueing systems and networks with many interacting elements. As a consequence, alternative ideas and tools, analogous to those applied in the field of Statistical Mechanics, have been proposed in the literature. In this context, the principles of Maximum Entropy (ME) and Minimum Relative Entropy (MRE), a generalisation, provide consistent methods of inference for characterising the form of an unknown but true probability distribution, based on information expressed in terms of known to exist true expected values or when, in addition, there exists a prior estimate of the unknown distribution. This paper traces the progress achieved so far towards the creation of ME and MRE product-form approximations and related algorithms for the performance analysis of general Queueing Network Models (QNMs) and indicates potential research extensions in the area.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    F. Baskett, K.M. Chandy, R.R. Muntz and F.G. Palacios, Open, closed and mixed networks with different classes of customers, J.ACM 22 (1975) 248–260.Google Scholar
  2. [2]
    M. Reiser and H. Kobayashi, Queueing networks with multiple closed chains: theory and computational algorithms, IBM J. Res. Dev. 19 (1975) 283–294.Google Scholar
  3. [3]
    R. Marie, An approximate analytical method for general queueing networks, IEEE Trans. Software Eng. SE-5 (1979) 530–538.Google Scholar
  4. [4]
    M. Reiser and H. Kobayashi, Accuracy of the diffusion approximation for some queueing systems, IBM J. Res. Dev. 18 (1974) 110–124.Google Scholar
  5. [5]
    E. Gelenbe and G. Pujolle, The behaviour of a single queue in a general queueing network, Acta Info. 7 (1976) 123–160.Google Scholar
  6. [6]
    P.J. Courtois,Decomposability: Queueing and Computer Systems Applications (Academic Press, New York, 1977).Google Scholar
  7. [7]
    K.M. Chandy, U. Herzog and L. Woo, Approximate analysis of general queueing networks, IBM J. Res. Dev. 19 (1975) 43–49.Google Scholar
  8. [8]
    K.C. Sevcik, A.I. Levy, S.K. Tripathi and J.L. Zahorjan, Improving approximation of aggregated queueing network subsystems, inComputer Performance, eds. K.M. Chandy and M. Reiser (North-Holland, New York, 1977) pp. 1–22.Google Scholar
  9. [9]
    F.P. Kelly,Reversibility and Stochastic Networks (Wiley, New York, 1979).Google Scholar
  10. [10]
    B. Bondi and Y.M. Chuang, A new MVA-based approximation for closed queueing networks with a preemptive priority server, Perf. Eval. 8 (1988) 195–221.Google Scholar
  11. [11]
    R.M. Bryant, A.E. Krzesinski, M.S. Laksmi and K.M. Chandy, The MVA priority approximation, T.O.C.S. 2 (1984) 335–359.Google Scholar
  12. [12]
    D. Neuse and K.M. Chandy, HAM: The heuristic algorithm method for solving general closed queueing models of computer systems, Perf. Eval. Rev. 11 (Proc. ACM SIGMETRICS Conf.) (1982) 195–212.Google Scholar
  13. [13]
    J. Zahorjan, E.D. Lazowska and K.C. Sevick, The use of approximations in production performance evaluation software,Proc. Int. Workshop on Modelling Techniques and Performance Evaluation, Paris (1987).Google Scholar
  14. [14]
    T. Altiok and H.G. Perros, Approximate analysis of arbitrary configurations of queueing networks with blocking, Ann. Oper. Res. 9 (1987) 481–509.Google Scholar
  15. [15]
    N.M. Van Dijk, A simple bounding methodology for non-product-form queueing networks with blocking, in:Queueing Networks with Blocking, eds. H.G. Perros and T. Altiok (North-Holland, 1980) pp. 3–18.Google Scholar
  16. [16]
    I.F. Akyildiz and H. Von Brand, Exact solutions for open, closed and mixed queueing networks with rejecton blocking, Theoret. Comp. Sci. 64 (1989) 203–219.Google Scholar
  17. [17]
    I.F. Akyildiz and H. Von Brand, Central server models with multiple job classes, state dependent routing and rejection blocking, IEEE Trans. Software Eng., to appear.Google Scholar
  18. [18]
    H.G. Perros, Approximation algorithms for open queueing networks with blocking, in:Stochastic Analysis of Computer and Communication Systems, ed. H. Takagi (North-Holland, 1990) pp. 451–494.Google Scholar
  19. [19]
    R.O. Onvural, Survey of closed queueing networks with blocking, ACM Comp. Surveys 22 (1990) 83–121.Google Scholar
  20. [20]
    R.O. Onvural and H.G. Perros, Approximate analysis of multi-class tandem queueing networks with Coxian parameters and finite buffers, in:Performance 1990, eds. P.J.B. King et al. (North-Holland, 1990) pp. 131–141.Google Scholar
  21. [21]
    V.E. Benes,Mathematical Theory of Connecting Networks and Telephone Traffic (Academic Press, New York, 1965).Google Scholar
  22. [22]
    A.E. Ferdinand, A statistical mechanical approach to systems analysis, IBM J. Res. Dev. 14 (1970) 539–547.Google Scholar
  23. [23]
    E. Pinsky and Y. Yemini, A statistical mechanics of some interconnection networks, in:Performance '84, ed. E. Gelenbe (North-Holland, 1984) pp. 147–158.Google Scholar
  24. [24]
    E. Pinsky and Y. Yemini, The canonical approximation in performance analysis, in:Computer Networking and Performance Evaluation, eds. T. Hasegawa et al. (North-Holland, 1986) pp. 125–137.Google Scholar
  25. [25]
    J.E. Shore, Information theoretic approximations forM/G/1 queueing systems, Acta Info. 17 (1982) 43–61.Google Scholar
  26. [26]
    M.A. El-Affendi and D.D. Kouvatsos, A maximum entropy analysis of theM/G/1 andG/M/1 queueing systems at equilibrium, Acta Info. 19 (1983) 339–355.Google Scholar
  27. [27]
    D.D. Kouvatsos, Maximum entropy methods for general queueing networks, in:Modelling Techniques and Tools for Performance Analysis, ed. D. Potier (North-Holland, 1985) pp. 589–609.Google Scholar
  28. [28]
    D.D. Kouvatsos, A universal maximum entropy algorithm for the analysis of general closed networks, in:Computing Networking and Performance Evaluation, eds. T. Hasegawa et al. (North-Holland, 1986) pp. 113–124.Google Scholar
  29. [29]
    E.T. Jaynes, Information theory and statistical mechanics, Phys. Rev. 106 (1957) 620–630.Google Scholar
  30. [30]
    E.T. Jaynes, Information theory and statistical mechanics II, Phys. Rev. 108 (1957) 171–190.Google Scholar
  31. [31]
    E.T. Jaynes, Prior probabilities, IEEE Trans. Syst. Sci. Cybern. SSC-4 (1968) 227–241.Google Scholar
  32. [32]
    C.E. Shannon, A mathematical theory of communication, Bell Syst. Tech. J. 27 (1948) 379–423, 623–656.Google Scholar
  33. [33]
    M. Tribus,Rational Description, Decisions and Designs (Pergamon, New York, 1969).Google Scholar
  34. [34]
    J.E. Shore and R.W. Johnson, Axiomatic derivation of the principle of ME and the principle of minimum cross-entropy, IEEE Trans. Info. Theory IT-26 (1980) 26–37.Google Scholar
  35. [35]
    J.E. Shore and R.W. Johnson, Properties of cross-entropy minimisation, IEEE Trans. Info. Theory IT-27 (1981) 472–482.Google Scholar
  36. [36]
    Y. Bard, A model of shared DASD and multipathing, Commun. ACM 23 (1980) 564–572.Google Scholar
  37. [37]
    C. Bobrowski, The principle of maximum entropy in some computer systems modelling problems, MSc Thesis, University of Toronto (1983).Google Scholar
  38. [38]
    P.J. Denning and J.P. Buzen, The operational analysis of queueing network models, Comp. Surveys 10 (1978) 225–261.Google Scholar
  39. [39]
    J.E. Shore, Derivation of equilibrium and time-dependent solutions ofM/M/∞/N andM/M/∞ queueing systems using entropy maximisation,Proc. 1978 National Computer Conf., AFIPS (1978) pp. 483–487.Google Scholar
  40. [40]
    J. Cantor, A. Ephremides and D. Horton, Information theoretic analysis for a general queueing system at equilibrium with application to queues in tandem, Acta Info. 23 (1986) 657–678.Google Scholar
  41. [41]
    J.L. Lutton, E. Bonomi and M.R. Feix, Information theory and statistical mechanics approach to solving a queueing problem encountered in telephone exchanges, Research Report PA/1199, Division ATR — Centre National d'Etudes des Télécommunications, France (1984).Google Scholar
  42. [42]
    S. Guiasu, Maximum entropy condition in queueing theory, J. Oper. Res. Soc. 37 (1986) 293–301.Google Scholar
  43. [43]
    V. Rego and W. Szpankowski, The presence of exponentiality in entropy maximisedM/GI/1 queues, Comp. Oper. Res. 16 (1989) 441–449.Google Scholar
  44. [44]
    D.D. Kouvatsos, A maximum entropy analysis of theG/G/1 queue at equilibrium, J. Oper. Res. Soc. 39 (1989) 183–200.Google Scholar
  45. [45]
    D.D. Kouvatsos, Maximum entropy and theG/G/1/N queue, Acta Info. 23 (1986) 545–565.Google Scholar
  46. [46]
    I. Arizono, Y. Cui and H. Ohta, An analysis ofM/M/s queueing systems based on the maximum entropy principle, J. Oper. Res. Soc. 42 (1991) 69–73.Google Scholar
  47. [47]
    D.D. Kouvatsos and J. Almond, Maximum entropy two-station cyclic queues with multiple general servers, Acta Info. 26 (1988) 241–267.Google Scholar
  48. [48]
    J.S. Wu and W.C. Chan, Maximum entropy analysis of multiple-server queueing systems, J. Oper. Res. Soc. 40 (1989) 815–825.Google Scholar
  49. [49]
    C. Strelen, Piecewise approximation of densities applying the principle of maximum entropy: Waiting times inG/G/1-systems, in:Modelling Techniques and Tools for Computer Performance Evaluation, eds. R. Puigjaner and D. Potier (Plenum, 1989) pp. 421–438.Google Scholar
  50. [50]
    D.D. Kouvatsos and N.M. Tabet-Aouel, A maximum entropy priority approximation for a stableG/G/1 queue, Acta Info. 27 (1989) 247–286.Google Scholar
  51. [51]
    N.M. Tabet-Aouel and D.D. Kouvatsos, On an approximation to the mean response teimes of priority classes in a stableG/G/c/PR queue, J. Oper. Res. Soc. 43 (1992) 227–239.Google Scholar
  52. [52]
    D.D. Kouvatsos and N.M. Tabet-Aouel, An ME-based approximation for multi-server queues with preemptive priority, Euro. J. Oper. Res. (1993), to appear.Google Scholar
  53. [53]
    D.D. Kouvatsos and P.H. Georgatsos, General queueing networks with SQJ routing,Proc. 5th Performance Engineering Workshop, Edinburgh (1988).Google Scholar
  54. [54]
    G. Falin, G.I. Martin and J. Artaleo, Information theoretic approximations for theM/G/1 retrial queue, Acta Info. (1993), to appear.Google Scholar
  55. [55]
    D.D. Kouvatsos, P.H. Georgatsos and N.M. Tabet-Aouel, A universal maximum entropy algorithm for general multiple class open networks with mixed service disciplines, in:Modelling Techniques and Tools for Computer Performance Evaluation, eds. R. Puigjaner and D. Potier (Plenum, 1989) pp. 397–419.Google Scholar
  56. [56]
    D.D. Kouvatsos and N.M. Tabet-Aouel, Product-form approximations for an extended class of general closed queueing networks, in:Performance '90, eds. P.J.B. King et al. (North-Holland, 1990) pp. 301–315.Google Scholar
  57. [57]
    P.H. Georgatsos, Modelling and analysis of computer communication networks with random or semidynamic routing, PhD Thesis, Computer Systems Modelling Research Group, University of Bradford (1989).Google Scholar
  58. [58]
    N.M. Tabet-Aouel, General queueing networks with priorities, PhD Thesis, Computer Systems Modelling Research Group, University of Bradford (1989).Google Scholar
  59. [59]
    J. Almond, Generalised analytic queueing network models, PhD Thesis, Computer Systems Modelling Research Group, University of Bradford (1988).Google Scholar
  60. [60]
    D.D. Kouvatsos and N.M. Tabet-Aouel, Approximate solutions for networks of queues with multiple server stations under PR discipline,Proc. 8th UK Performance Engineering Workshop, eds. P. Harrison and T. Field, Imperial College, London (1992).Google Scholar
  61. [61]
    D.D. Kouvatsos and N.P. Xenios, Maximum entropy analysis of general queueing netwoks with blocking, in:Queueing Networks with Blocking, eds. H.G. Perros and T. Altiok (North-Holland, Amsterdam, 1989) pp. 281–309.Google Scholar
  62. [62]
    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
  63. [63]
    D.D. Kouvatsos, S.G. Denazis and P.H. Georgatsos, MEM for arbitrary exponential open networks with blocking and multiple job classes,Proc. 7th Performance Engineering Workshop, eds. J. Hillston et al. (Springer, 1991) pp. 163–178.Google Scholar
  64. [64]
    D.D. Kouvatsos and S.G. Denazis, Entropy maximised queueing networks with blocking and multiple job classes, Perf. Eval. 17 (1993) 189–205.Google Scholar
  65. [65]
    N.P. Xenios, General queueing networks with blocking, PhD Thesis, Computer Systems Modelling Research Group, University of Bradford (1989).Google Scholar
  66. [66]
    D.D. Kouvatsos and P.J. Tomaras, Multilevel aggregation of central server models: A minimum relative entropy approach, Int. J. Syst. Sci. 23 (1992) 713–739.Google Scholar
  67. [67]
    P.J. Tomaras and D.D. Kouvatsos, MRE hierarchical decomposition of general queueing network models, Acta Info. 28 (1991) 265–295.Google Scholar
  68. [68]
    P.J. Tomaras, Decomposition of general queueing network models, PhD Thesis, University of Bradford (1989).Google Scholar
  69. [69]
    D.D. Kouvatsos and P.H. Georgatsos, Queueing models of packet-switched networks with locally adaptive routing, in:Teletraffic and Datatraffic in Period of Change, ITC-13, eds. A. Jensen and V.B. Iversen (North-Holland, 1991) pp. 303–308.Google Scholar
  70. [70]
    D.D. Kouvatsos and A.K. Simpson, Two new methods for analysing complex I/O subsystems, Technical Report, University of Bradford (February 1992).Google Scholar
  71. [71]
    A.K. Simpson, Delay models for complex disk I/O subsystems, PhD Thesis, University of Bradford (1993).Google Scholar
  72. [72]
    D.D. Kouvatsos and A. Skliros, General QNMs with variable concurrent programming structures and synchronisation,Proc. 7th UK Performance Engineering Workshop, eds. J. Hillston et al. (Springer, 1991) pp. 56–72.Google Scholar
  73. [73]
    A. Skliros, Modelling computer systems with job concurrency and synchronisation, PhD Thesis, University of Bradford (1991).Google Scholar
  74. [74]
    D.D. Kouvatsos and A.T. Othman, Optimal flow control of aG/G/c finite capacity queue, J. Oper. Res. Soc. 40 (1989) 659–670.Google Scholar
  75. [75]
    A.T. Othman, Performance analysis and control of computer communication network models, PhD Thesis, University of Bradford, Computer Systems Modelling Research Group (1988).Google Scholar
  76. [76]
    D.D. Kouvatsos and A.T. Othman, Optimal flow control of end-to-end packet-switched network with random routing, IEE Proc. 136 (1989) 90–99.Google Scholar
  77. [77]
    R.W. Johnson, Determining probability distributions by maximum entropy and minimum cross-entropy, in:APL 1979 Conf. Proc. (1979) pp. 24–29.Google Scholar
  78. [78]
    C. Sauer, Configuration of computing systems: An approach using queueing network models, PhD Thesis, University of Texas (1975).Google Scholar
  79. [79]
    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
  80. [80]
    W. Whitt, On approximations for queues, III: Mixtures of exponential distributions, AT&T BLTJ 63 (1984) 163–175.Google Scholar
  81. [81]
    M. Reiser and S.S. Lavenberg, Mean-value analysis of closed multi-chain queueing networks, J. ACM 27 (1980) 313–322.Google Scholar
  82. [82]
    R. Walstra, Iterative analysis of networks of queues, PhD Thesis, Technical Report CSE1-166, University of Toronto (1984).Google Scholar
  83. [83]
    M. Veran and D. Potier, QNAP-2: A portable environment for queueing network modelling, in:Modelling Techniques and Tools for Performance Analysis, ed. D. Potier (North-Holland, 1985) pp. 25–63.Google Scholar
  84. [84]
    A.B. Bondi, Modelling the effect of local area networks contention on the performance of host computers,Perf. '84, ed. E. Gelenbe (North-Holland, 1984) pp. 303–318.Google Scholar
  85. [85]
    I. Mitrani and P.J.B. King, Multiprocessor systems with preemptive priorities, Perf. Eval. 1 (1981) 118–125.Google Scholar
  86. [86]
    D.D. Kouvatsos and S.G. Denazis, Comments on and tuning to: MEM for arbitrary queueing networks with repetitive-service blocking, Technical Report CS-18-91, University of Bradford (May 1991).Google Scholar
  87. [87]
    D.D. Kouvatsos and S.G. Denazis, An efficient convolution procedure for general product-form approximations, Technical Report CS-18-92, University of Bradford (February 1992).Google Scholar
  88. [88]
    J. Labetoulle and G. Pujolle, Isolation method in a network of queues, IEEE Trans. Software Eng. SE-6 (1980) 373–381.Google Scholar
  89. [89]
    D.D. Kouvatsos and J. Almond, Maximum entropy two-station cyclic queues with blocking, Technical Report CS-18-88, University of Bradford (March 1988).Google Scholar
  90. [90]
    D.D. Kouvatsos, J. Almond and J. Sztrik, TheGE-type building block for general queueing networks with before or after service blocking, Technical Report CS-23-91, University of Bradford (September 1991).Google Scholar
  91. [91]
    D.D. Kouvatsos and S.G. Denazis, Queueing network models with before service blocking, Technical Report CS-20-92, University of Bradford (February 1992).Google Scholar
  92. [92]
    J.E. Shore, Cross-entropy minimisation given fully decomposable subset and aggregate constraints, IEEE Trans. Inf. Theory IT-28 (1982) 956–961.Google Scholar
  93. [93]
    H. Vantilborgh, Exact aggregation in exponential queueing networks, ACM J. 24 (1978) 620–629.Google Scholar
  94. [94]
    S.K. Tripathi, On approximate solution techniques for queueing network models of computer systems, PhD Thesis, University of Toronto (1979).Google Scholar
  95. [95]
    E. Koenigsberg, On jockeying in queues, Manag. Sci. 12 (1966) 412.Google Scholar
  96. [96]
    R.L. Disney and W.E. Mitchell, A solution for queues with instantaneous jockeying and other customer rules, Naval Res. Log. Quarterly 17 (1970) 315–325.Google Scholar
  97. [97]
    Y. Bard, Estimation of state probabilities using the maximum entropy principle, IBM J. Res. Dev. 24 (1980) 563–569.Google Scholar
  98. [98]
    E.D. Lazowska, J. Zahorjan, G. Scott-Graham and K.C. Sevcik,Quantitiative Systems Performance (Prentice-Hall, 1984).Google Scholar
  99. [99]
    P. Heidelberger and K.S. Trivedi, Analytic queueing models for programs with internal concurrency, IEEE Trans. Comp. C-32 (1983) 73–82.Google Scholar
  100. [100]
    D. Towsley, K.M. Chandy and J.C. Browne, Models for parallel processing with programs: Applications to CPU:I/O and I/O:I/O overlap, Commun. ACM 21 (1978) 821–831.Google Scholar
  101. [101]
    S.S. Lam and M. Reiser, Congestion control of store-and-forward networks by input buffer limits — An analysis, IEEE Trans. Commun. COM-27 (1979) 127–133.Google Scholar
  102. [102]
    A.A. Lazar, Optimal flow control of a class of queueing networks in equilibrium, IEEE Trans. Auto. Control AC-28 (1983) 1001–1007.Google Scholar
  103. [103]
    F.A. Tobagi, Fast packet switch architectures for broadband integrated service digital networks, Proc. IEEE 78 (1990) 133–167.Google Scholar
  104. [104]
    G. Pujolle, Discrete-time queueing systems for data networks performance evaluation, in:Queueing, Performance and Control in ATM, eds. J.W. Cohen and C.D. Park (North-Holland, 1991) pp. 239–244.Google Scholar
  105. [105]
    D.D. Kouvatsos and N.M. Tabet-Aouel,GGeo-type approximations for general discrete-time queueing systems,Proc. Modelling and Performance Evaluation of ATM Technology, eds. H.G. Perros et al., La Martinique (1993) pp. 11.1.1.–11.1.12.Google Scholar
  106. [106]
    H.G. Perrors, R. Puignaner and G. Pujolle, Performance issues in ATM networks: An annotated bibliography, Research Report, North Carolina State University (1990).Google Scholar
  107. [107]
    D.D. Kouvatsos, N.M. Tabet-Aouel and S.G. Denazis, ME-based approximations for general discrete-time queueing models, Technical Report CS-41-92, University of Bradford (November 1992).Google Scholar
  108. [108]
    J. Sztrik and D.D. Kouvatsos, Asymptotic analysis of a heterogeneous multi-processor system in a randomly changing environment, IEEE Trans. Software Eng. SE-17 (1991) 1069–1075.Google Scholar

Copyright information

© J.C. Baltzer AG, Science Publishers 1994

Authors and Affiliations

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

Personalised recommendations