Problems of Information Transmission

, Volume 55, Issue 2, pp 174–200 | Cite as

The Geometry of Big Queues

  • A. A. PuhalskiiEmail author
Communication Network Theory


We use Hamilton equations to identify most likely scenarios of long queues being formed in ergodic Jackson networks. Since the associated Hamiltonians are discontinuous and piecewise Lipschitz, one has to invoke methods of nonsmooth analysis. Time reversal of the Hamilton equations yields fluid equations for the dual network. Accordingly, the optimal trajectories are time reversals of the fluid trajectories of the dual network. Those trajectories are shown to belong to domains that satisfy a certain condition of being “essential.” As an illustration, we consider a two-station Jackson network. In addition, we prove certain properties of substochastic matrices, which may be of interest in their own right.

Key words

queueing theory Jackson networks large deviations large deviation principle optimal trajectories Hamilton equations dual Markov processes fluid dynamics 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.



The author is grateful to S.A. Pirogov and A.N. Rybko for helpful discussions and advice on improving the presentation.


  1. 1.
    Wentzell, A.D. and Freidlin, M.I., Fluktuatsii v dinamicheskikh sistemakh pod deistviem malykh sluchainykh vozmushchenii, Moscow: Nauka, 1979. Tanslated under the title Random Perturbations of Dynamical Systems, New York: Springer, 1984.Google Scholar
  2. 2.
    Shwartz, A. and Weiss, A., Large Deviations for Performance Analysis: Queues, Communications, and Computing, London: Chapman & Hall, 1995.zbMATHGoogle Scholar
  3. 3.
    Kleinrock, L., Queueing Systems, vol. 1: Theory, New York: Wiley, 1975. Translated under the title Teoriya massovogo obsluzhivaniya, Moscow: Mashinostroenie, 1979.zbMATHGoogle Scholar
  4. 4.
    Jacod, J. and Shiryaev, A.N., Limit Theorems for Stochastic Processes, Berlin: Springer, 1987. Translated under the title Predel’nye teoremy dlya sluchainykh protsessov, Moscow: Fizmatlit, 1994.CrossRefzbMATHGoogle Scholar
  5. 5.
    Ethier, S.N. and Kurtz, T.G., Markov Processes: Characterization and Convergence, New York: Wiley, 1986.CrossRefzbMATHGoogle Scholar
  6. 6.
    Puhalskii, A.A., The Action Functional for the Jackson Network, Markov Process. Related Fields, 2007, vol. 13, no. 1, pp. 99–136.MathSciNetzbMATHGoogle Scholar
  7. 7.
    Atar, R. and Dupuis, P., Large Deviations and Queueing Networks: Methods for Rate Function Identification, Stochastic Process. Appl., 1999, vol. 84, no. 2, pp. 255–296.MathSciNetCrossRefzbMATHGoogle Scholar
  8. 8.
    Ignatiouk-Robert, I., Large Deviations of Jackson Networks, Ann. Appl. Probab., 2000, vol. 10, no. 3, pp. 962–1001.MathSciNetCrossRefzbMATHGoogle Scholar
  9. 9.
    Collingwood, J., Path Properties of Rare Events, PhD Thesis, Univ. of Ottawa, Canada, 2015.Google Scholar
  10. 10.
    Bouchet, F., Laurie, J., and Zaboronski, O., Langevin Dynamics, Large Deviations and Instantons for the Quasi-geostrophic Model and Two-Dimensional Euler Equations, J. Stat. Phys., 2014, vol. 156, no. 6, pp. 1066–1092.MathSciNetCrossRefzbMATHGoogle Scholar
  11. 11.
    Clarke, F.H., Optimization and Nonsmooth Analysis, New York: Wiley, 1983. Translated under the title Optimizatsiya i negladkii analiz, Moscow: Mir, 1988.zbMATHGoogle Scholar
  12. 12.
    Rockafellar, R.T., Convex Analysis, Princeton Math. Series, vol. 28, Princeton: Princeton Univ. Press, 1970. Translated under the title Vypuklyi analiz, Moscow: Mir, 1973.CrossRefzbMATHGoogle Scholar
  13. 13.
    Bramson, M., Stability of Queueing Networks (Ecole d’ÉtédeProbabilités de Saint-Flour XXXVI-2006), Lect. Notes Math., vol. 1950, Berlin: Springer, 2008.zbMATHGoogle Scholar
  14. 14.
    Anantharam, V., Heidelberger, R., and Tsoukas, P., Analysis of Rare Events in Continuous Time Markov Chains via Time Reversal and Fluid Approximation, IBM Res. Rep. RC 16280, Yorktown Heights, NY, 1990.Google Scholar
  15. 15.
    Shwartz, A. and Weiss, A., Induced Rare Events: Analysis via Large Deviations and Time Reversal, Adv. in Appl. Probab., 1993, vol. 25, no. 3, pp. 667–689.MathSciNetCrossRefzbMATHGoogle Scholar
  16. 16.
    Majewski, K. and Ramanan, K., How Large Queue Lengths Build Up in a Jackson Network, Preprint, 2008.Google Scholar
  17. 17.
    Vinter, R.B., Optimal Control, Boston: Birkhäuser, 2000.zbMATHGoogle Scholar
  18. 18.
    Bazaraa, M.S. and Shetty, C.M., Nonlinear Programming: Theory and Algorithms, New York: Wiley, 1979. Translated under the title Nelineinoe programmirovanie. Teoriya i algoritmy, Moscow: Mir, 1982.zbMATHGoogle Scholar

Copyright information

© Pleiades Publishing, Inc. 2019

Authors and Affiliations

  1. 1.Kharkevich Institute for Information Transmission ProblemsRussian Academy of SciencesMoscowRussia

Personalised recommendations