Entropy maximisation and queueing network models
 Demetres D. Kouvatsos
 … show all 1 hide
Rent the article at a discount
Rent now* Final gross prices may vary according to local VAT.
Get AccessAbstract
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 productform approximations and related algorithms for the performance analysis of general Queueing Network Models (QNMs) and indicates potential research extensions in the area.
 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.
 M. Reiser and H. Kobayashi, Queueing networks with multiple closed chains: theory and computational algorithms, IBM J. Res. Dev. 19 (1975) 283–294.
 R. Marie, An approximate analytical method for general queueing networks, IEEE Trans. Software Eng. SE5 (1979) 530–538.
 M. Reiser and H. Kobayashi, Accuracy of the diffusion approximation for some queueing systems, IBM J. Res. Dev. 18 (1974) 110–124.
 E. Gelenbe and G. Pujolle, The behaviour of a single queue in a general queueing network, Acta Info. 7 (1976) 123–160.
 P.J. Courtois,Decomposability: Queueing and Computer Systems Applications (Academic Press, New York, 1977).
 K.M. Chandy, U. Herzog and L. Woo, Approximate analysis of general queueing networks, IBM J. Res. Dev. 19 (1975) 43–49.
 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 (NorthHolland, New York, 1977) pp. 1–22.
 F.P. Kelly,Reversibility and Stochastic Networks (Wiley, New York, 1979).
 B. Bondi and Y.M. Chuang, A new MVAbased approximation for closed queueing networks with a preemptive priority server, Perf. Eval. 8 (1988) 195–221.
 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.
 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.
 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).
 T. Altiok and H.G. Perros, Approximate analysis of arbitrary configurations of queueing networks with blocking, Ann. Oper. Res. 9 (1987) 481–509.
 N.M. Van Dijk, A simple bounding methodology for nonproductform queueing networks with blocking, in:Queueing Networks with Blocking, eds. H.G. Perros and T. Altiok (NorthHolland, 1980) pp. 3–18.
 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.
 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.
 H.G. Perros, Approximation algorithms for open queueing networks with blocking, in:Stochastic Analysis of Computer and Communication Systems, ed. H. Takagi (NorthHolland, 1990) pp. 451–494.
 R.O. Onvural, Survey of closed queueing networks with blocking, ACM Comp. Surveys 22 (1990) 83–121.
 R.O. Onvural and H.G. Perros, Approximate analysis of multiclass tandem queueing networks with Coxian parameters and finite buffers, in:Performance 1990, eds. P.J.B. King et al. (NorthHolland, 1990) pp. 131–141.
 V.E. Benes,Mathematical Theory of Connecting Networks and Telephone Traffic (Academic Press, New York, 1965).
 A.E. Ferdinand, A statistical mechanical approach to systems analysis, IBM J. Res. Dev. 14 (1970) 539–547.
 E. Pinsky and Y. Yemini, A statistical mechanics of some interconnection networks, in:Performance '84, ed. E. Gelenbe (NorthHolland, 1984) pp. 147–158.
 E. Pinsky and Y. Yemini, The canonical approximation in performance analysis, in:Computer Networking and Performance Evaluation, eds. T. Hasegawa et al. (NorthHolland, 1986) pp. 125–137.
 J.E. Shore, Information theoretic approximations forM/G/1 queueing systems, Acta Info. 17 (1982) 43–61.
 M.A. ElAffendi 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.
 D.D. Kouvatsos, Maximum entropy methods for general queueing networks, in:Modelling Techniques and Tools for Performance Analysis, ed. D. Potier (NorthHolland, 1985) pp. 589–609.
 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. (NorthHolland, 1986) pp. 113–124.
 E.T. Jaynes, Information theory and statistical mechanics, Phys. Rev. 106 (1957) 620–630.
 E.T. Jaynes, Information theory and statistical mechanics II, Phys. Rev. 108 (1957) 171–190.
 E.T. Jaynes, Prior probabilities, IEEE Trans. Syst. Sci. Cybern. SSC4 (1968) 227–241.
 C.E. Shannon, A mathematical theory of communication, Bell Syst. Tech. J. 27 (1948) 379–423, 623–656.
 M. Tribus,Rational Description, Decisions and Designs (Pergamon, New York, 1969).
 J.E. Shore and R.W. Johnson, Axiomatic derivation of the principle of ME and the principle of minimum crossentropy, IEEE Trans. Info. Theory IT26 (1980) 26–37.
 J.E. Shore and R.W. Johnson, Properties of crossentropy minimisation, IEEE Trans. Info. Theory IT27 (1981) 472–482.
 Y. Bard, A model of shared DASD and multipathing, Commun. ACM 23 (1980) 564–572.
 C. Bobrowski, The principle of maximum entropy in some computer systems modelling problems, MSc Thesis, University of Toronto (1983).
 P.J. Denning and J.P. Buzen, The operational analysis of queueing network models, Comp. Surveys 10 (1978) 225–261.
 J.E. Shore, Derivation of equilibrium and timedependent solutions ofM/M/∞/N andM/M/∞ queueing systems using entropy maximisation,Proc. 1978 National Computer Conf., AFIPS (1978) pp. 483–487.
 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.
 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).
 S. Guiasu, Maximum entropy condition in queueing theory, J. Oper. Res. Soc. 37 (1986) 293–301.
 V. Rego and W. Szpankowski, The presence of exponentiality in entropy maximisedM/GI/1 queues, Comp. Oper. Res. 16 (1989) 441–449.
 D.D. Kouvatsos, A maximum entropy analysis of theG/G/1 queue at equilibrium, J. Oper. Res. Soc. 39 (1989) 183–200.
 D.D. Kouvatsos, Maximum entropy and theG/G/1/N queue, Acta Info. 23 (1986) 545–565.
 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.
 D.D. Kouvatsos and J. Almond, Maximum entropy twostation cyclic queues with multiple general servers, Acta Info. 26 (1988) 241–267.
 J.S. Wu and W.C. Chan, Maximum entropy analysis of multipleserver queueing systems, J. Oper. Res. Soc. 40 (1989) 815–825.
 C. Strelen, Piecewise approximation of densities applying the principle of maximum entropy: Waiting times inG/G/1systems, in:Modelling Techniques and Tools for Computer Performance Evaluation, eds. R. Puigjaner and D. Potier (Plenum, 1989) pp. 421–438.
 D.D. Kouvatsos and N.M. TabetAouel, A maximum entropy priority approximation for a stableG/G/1 queue, Acta Info. 27 (1989) 247–286.
 N.M. TabetAouel 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.
 D.D. Kouvatsos and N.M. TabetAouel, An MEbased approximation for multiserver queues with preemptive priority, Euro. J. Oper. Res. (1993), to appear.
 D.D. Kouvatsos and P.H. Georgatsos, General queueing networks with SQJ routing,Proc. 5th Performance Engineering Workshop, Edinburgh (1988).
 G. Falin, G.I. Martin and J. Artaleo, Information theoretic approximations for theM/G/1 retrial queue, Acta Info. (1993), to appear.
 D.D. Kouvatsos, P.H. Georgatsos and N.M. TabetAouel, 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.
 D.D. Kouvatsos and N.M. TabetAouel, Productform approximations for an extended class of general closed queueing networks, in:Performance '90, eds. P.J.B. King et al. (NorthHolland, 1990) pp. 301–315.
 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).
 N.M. TabetAouel, General queueing networks with priorities, PhD Thesis, Computer Systems Modelling Research Group, University of Bradford (1989).
 J. Almond, Generalised analytic queueing network models, PhD Thesis, Computer Systems Modelling Research Group, University of Bradford (1988).
 D.D. Kouvatsos and N.M. TabetAouel, 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).
 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 (NorthHolland, Amsterdam, 1989) pp. 281–309.
 D.D. Kouvatsos and N.P. Xenios, MEM for arbitrary queueing networks with multiple general servers and repetitiveservice blocking, Perf. Eval. 10 (1989) 169–195.
 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.
 D.D. Kouvatsos and S.G. Denazis, Entropy maximised queueing networks with blocking and multiple job classes, Perf. Eval. 17 (1993) 189–205.
 N.P. Xenios, General queueing networks with blocking, PhD Thesis, Computer Systems Modelling Research Group, University of Bradford (1989).
 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.
 P.J. Tomaras and D.D. Kouvatsos, MRE hierarchical decomposition of general queueing network models, Acta Info. 28 (1991) 265–295.
 P.J. Tomaras, Decomposition of general queueing network models, PhD Thesis, University of Bradford (1989).
 D.D. Kouvatsos and P.H. Georgatsos, Queueing models of packetswitched networks with locally adaptive routing, in:Teletraffic and Datatraffic in Period of Change, ITC13, eds. A. Jensen and V.B. Iversen (NorthHolland, 1991) pp. 303–308.
 D.D. Kouvatsos and A.K. Simpson, Two new methods for analysing complex I/O subsystems, Technical Report, University of Bradford (February 1992).
 A.K. Simpson, Delay models for complex disk I/O subsystems, PhD Thesis, University of Bradford (1993).
 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.
 A. Skliros, Modelling computer systems with job concurrency and synchronisation, PhD Thesis, University of Bradford (1991).
 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.
 A.T. Othman, Performance analysis and control of computer communication network models, PhD Thesis, University of Bradford, Computer Systems Modelling Research Group (1988).
 D.D. Kouvatsos and A.T. Othman, Optimal flow control of endtoend packetswitched network with random routing, IEE Proc. 136 (1989) 90–99.
 R.W. Johnson, Determining probability distributions by maximum entropy and minimum crossentropy, in:APL 1979 Conf. Proc. (1979) pp. 24–29.
 C. Sauer, Configuration of computing systems: An approach using queueing network models, PhD Thesis, University of Texas (1975).
 S. Nojo and H. Watanabe, A new stage method getting arbitrary coefficient of variation by two stages, Trans. IEICE 70 (1987) 33–36.
 W. Whitt, On approximations for queues, III: Mixtures of exponential distributions, AT&T BLTJ 63 (1984) 163–175.
 M. Reiser and S.S. Lavenberg, Meanvalue analysis of closed multichain queueing networks, J. ACM 27 (1980) 313–322.
 R. Walstra, Iterative analysis of networks of queues, PhD Thesis, Technical Report CSE1166, University of Toronto (1984).
 M. Veran and D. Potier, QNAP2: A portable environment for queueing network modelling, in:Modelling Techniques and Tools for Performance Analysis, ed. D. Potier (NorthHolland, 1985) pp. 25–63.
 A.B. Bondi, Modelling the effect of local area networks contention on the performance of host computers,Perf. '84, ed. E. Gelenbe (NorthHolland, 1984) pp. 303–318.
 I. Mitrani and P.J.B. King, Multiprocessor systems with preemptive priorities, Perf. Eval. 1 (1981) 118–125.
 D.D. Kouvatsos and S.G. Denazis, Comments on and tuning to: MEM for arbitrary queueing networks with repetitiveservice blocking, Technical Report CS1891, University of Bradford (May 1991).
 D.D. Kouvatsos and S.G. Denazis, An efficient convolution procedure for general productform approximations, Technical Report CS1892, University of Bradford (February 1992).
 J. Labetoulle and G. Pujolle, Isolation method in a network of queues, IEEE Trans. Software Eng. SE6 (1980) 373–381.
 D.D. Kouvatsos and J. Almond, Maximum entropy twostation cyclic queues with blocking, Technical Report CS1888, University of Bradford (March 1988).
 D.D. Kouvatsos, J. Almond and J. Sztrik, TheGEtype building block for general queueing networks with before or after service blocking, Technical Report CS2391, University of Bradford (September 1991).
 D.D. Kouvatsos and S.G. Denazis, Queueing network models with before service blocking, Technical Report CS2092, University of Bradford (February 1992).
 J.E. Shore, Crossentropy minimisation given fully decomposable subset and aggregate constraints, IEEE Trans. Inf. Theory IT28 (1982) 956–961.
 H. Vantilborgh, Exact aggregation in exponential queueing networks, ACM J. 24 (1978) 620–629.
 S.K. Tripathi, On approximate solution techniques for queueing network models of computer systems, PhD Thesis, University of Toronto (1979).
 E. Koenigsberg, On jockeying in queues, Manag. Sci. 12 (1966) 412.
 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.
 Y. Bard, Estimation of state probabilities using the maximum entropy principle, IBM J. Res. Dev. 24 (1980) 563–569.
 E.D. Lazowska, J. Zahorjan, G. ScottGraham and K.C. Sevcik,Quantitiative Systems Performance (PrenticeHall, 1984).
 P. Heidelberger and K.S. Trivedi, Analytic queueing models for programs with internal concurrency, IEEE Trans. Comp. C32 (1983) 73–82.
 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.
 S.S. Lam and M. Reiser, Congestion control of storeandforward networks by input buffer limits — An analysis, IEEE Trans. Commun. COM27 (1979) 127–133.
 A.A. Lazar, Optimal flow control of a class of queueing networks in equilibrium, IEEE Trans. Auto. Control AC28 (1983) 1001–1007.
 F.A. Tobagi, Fast packet switch architectures for broadband integrated service digital networks, Proc. IEEE 78 (1990) 133–167.
 G. Pujolle, Discretetime queueing systems for data networks performance evaluation, in:Queueing, Performance and Control in ATM, eds. J.W. Cohen and C.D. Park (NorthHolland, 1991) pp. 239–244.
 D.D. Kouvatsos and N.M. TabetAouel,GGeotype approximations for general discretetime 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.
 H.G. Perrors, R. Puignaner and G. Pujolle, Performance issues in ATM networks: An annotated bibliography, Research Report, North Carolina State University (1990).
 D.D. Kouvatsos, N.M. TabetAouel and S.G. Denazis, MEbased approximations for general discretetime queueing models, Technical Report CS4192, University of Bradford (November 1992).
 J. Sztrik and D.D. Kouvatsos, Asymptotic analysis of a heterogeneous multiprocessor system in a randomly changing environment, IEEE Trans. Software Eng. SE17 (1991) 1069–1075.
 Title
 Entropy maximisation and queueing network models
 Journal

Annals of Operations Research
Volume 48, Issue 1 , pp 63126
 Cover Date
 19940201
 DOI
 10.1007/BF02023095
 Print ISSN
 02545330
 Online ISSN
 15729338
 Publisher
 Baltzer Science Publishers, Baarn/Kluwer Academic Publishers
 Additional Links
 Topics
 Industry Sectors
 Authors

 Demetres D. Kouvatsos ^{(1)}
 Author Affiliations

 1. Computer Systems Modelling Research Group, University of Bradford, BD7 1DP, Bradford, West Yorkshire, England