Queueing Systems

, Volume 21, Issue 3–4, pp 293–335 | Cite as

Sample path methods in the control of queues

  • Zhen Liu
  • Philippe Nain
  • Don Towsley


Sample path methods are now among the most used techniques in the control of queueing systems. However, due to the lack of mathematical formalism, they may appear to be non-rigorous and even sometimes mysterious. The goal of this paper is threefold: to provide a general mathematical setting, to survey the most popular sample path methods including forward induction, backward induction and interchange arguments, and to illustrate our approach through the study of a number of classical scheduling and routing optimization problems arising in queueing theory.


Control scheduling mathematical formalism sample path arguments queueing system discrete event system stochastic comparison 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    F. Baccelli, Z. Liu and D. Towsley, Extremal scheduling of parallel processing with and without real-time constraints, J. ACM 40 (1993) 1209–1237.Google Scholar
  2. [2]
    J.S. Baras, D.-J. Ma and A.M. Makowski,K competing queues with geometric requirements and linear costs: the μc-rule is always optimal. Syst. Contr. Lett. 6 (1985) 173–180.Google Scholar
  3. [3]
    C. Buyukkoc, P. Varaiya and J. Walrand, The μic-rule revisited, Adv. Appl. Prob. 17 (1985) 237–238.Google Scholar
  4. [4]
    C.S. Chang, A new ordering for stochastic majorization: theory and applications, Adv. Appl. Prob. 24 (1992) 604–634.Google Scholar
  5. [5]
    P.W. Day, Rearrangement inequalities, Can. J. Monthly 24 (1972) 930–943.Google Scholar
  6. [6]
    A. Ephremides, P. Varaiya and J. Walrand, A simple dynamic routing problem, IEEE Trans. Autom. Contr. 25 (1980).Google Scholar
  7. [7]
    W. Feller,An Introduction to Probability Theory and Its Applications, Vol. 2, 2nd Ed. (Wiley, 1971).Google Scholar
  8. [8]
    S.G. Foss, Approximation of multichannel queueing systems (in Russian), Sibirski Mat. Zh. 21 (1980) 132–140 [transl: Siberian Math. J. 21 (1980) 851–857].Google Scholar
  9. [9]
    S.G. Foss, Comparison of servicing strategies in multichannel queueing systems (in Russian), Sibirski Math. Zh. 22 (1981) 190–197 [transi.: Siberian Math. J. 22 (1981) 141–147].Google Scholar
  10. [10]
    S.G. Foss, Queues with customers of several types, in:Limit Theorems and Related Problems, ed. A.A. Borovokov (Optimization Software, 1984) pp. 348–377.Google Scholar
  11. [11]
    S.G. Foss, Comparison of service disciplines inG/GI/m queues, INRIA Technical Report No. 1097 (1989).Google Scholar
  12. [12]
    S. Ghahramani and R.W. Wolff, A new proof of finite moment conditions for GI/G/1 busy periods, Queueing Systems 4 (1989) 171–178.Google Scholar
  13. [13]
    T. Hirayama and M. Kijima, An extremal property of FIFO discipline in G/IFR/1 queues, Adv. Appl. Prob. 21 (1989) 481–484.Google Scholar
  14. [14]
    A. Hordijk and G. Koole, On the optimality of the generalized shortest queue policy, Prob. Eng. Inf. Sci. 4 (1991) 477–487.Google Scholar
  15. [15]
    G.P. Klimov, Time-sharing service systems I, Th. Prob. Appl. 19 (1974) 532–551.Google Scholar
  16. [16]
    A. Hordijk and G. Koole, The μ-rule is not optimal in the second node of the tandem queue: a counterexample, Adv. Appl. Prob. 24 (1992) 234–237.Google Scholar
  17. [17]
    Z. Liu, P. Nain and D. Towsley, On optimal polling policies, Queueing Systems 11 (1992) 59–83.Google Scholar
  18. [18]
    Z. Liu and R. Righter, Optimal load balancing on distributed homogeneous unreliable processors, submitted for publication.Google Scholar
  19. [19]
    Z. Liu and D. Towsley, Effects of service disciplines inG/G/s queueing systems, Ann. Oper. Res. 48 (1994) 401–429.Google Scholar
  20. [20]
    Z. Liu and D. Towsley, Stochastic scheduling in in-forest networks, Adv. Appl. Prob. 26 (1994) 222–241.Google Scholar
  21. [21]
    Z. Liu and D. Towsley, Optimality of the Round Robin routing policy. J. Appl. Prob. 31 (1994) 466–475.Google Scholar
  22. [22]
    A.W. Marshall and I. Olkin,Inequalities: Theory of Majorization and Its Applications (Academic Press, 1979).Google Scholar
  23. [23]
    R. Menich and R.F. Serfozo, Optimality of routing and servicing in dependent parallel processing systems, Queueing Systems 9 (1991) 403–418.Google Scholar
  24. [24]
    P. Nain, Interchange arguments for classical scheduling problems in queues, Syst. Contr. Lett. 12 (1989) 177–184.Google Scholar
  25. [25]
    P. Nain, P. Tsoucas and J. Walrand, Interchange arguments in stochastic scheduling, J. Appl. Prob. 27 (1989) 815–826.Google Scholar
  26. [26]
    P. Nain and D. Towsley, Optimal scheduling in a machine with stochastic varying processing rate, IEEE Trans. Autom. Contr. 39 (1994) 1853–1855.Google Scholar
  27. [27]
    S.M. Ross,Applied Probability Models with Optimization Applications (Holden-Day, San Francisco, 1970).Google Scholar
  28. [28]
    M. Schäl, Conditions for optimality in dynamic programming and for the limit of n-stage optimal policies to be optimal, Z. Wahrscheinlichkeitsth. 32 (1975) 179–196.Google Scholar
  29. [29]
    J.G. Shanthikumar and U. Sumita, Convex ordering of sojourn times in single-server queues: extremal properties of FIFO and LIFO service disciplines, J. Appl. Prob. 24 (1987) 737–748.Google Scholar
  30. [30]
    P.D. Sparaggis, C.G. Cassandras and D. Towsley, On the duality between routing and scheduling systems with finite buffer space, IEEE Trans. Autom. Contr. 38 (1993) 1440–1446.Google Scholar
  31. [31]
    P.D. Sparaggis, D. Towsley and C.G. Cassandras, Sample path criteria for weak majorization, Adv. Appl. Prob. 26 (1994) 155–171.Google Scholar
  32. [32]
    P.D. Sparaggis, C.G. Cassandras and D. Towsley, Routing with limited state information in queueing systems with blocking, IEEE Trans. Autom. Contr. 39 (1994) 1492–1497.Google Scholar
  33. [33]
    D. Stoyan,Comparison Methods for Queues and Other Stochastic Models (Wiley, Berlin, 1983).Google Scholar
  34. [34]
    D. Towsley and F. Baccelli, Comparison of service disciplines in tandem queueing networks with delay dependent customer behavior, Oper. Res. Lett. 10 (1991) 49–55.Google Scholar
  35. [35]
    D. Towsley, S. Fdida and H. Santoso, Flow control protocols for interconnected high speed networks,Proc. NATO Workshop on High Speed Networks, Sophia-Antipolis, France (1990).Google Scholar
  36. [36]
    D. Towsley and P. Sparaggis, Optimal routing in systems with ILR service time distributions, Technical Report TR93-13, Department of Computer Science, Univ. of Massachusetts (1993).Google Scholar
  37. [37]
    D. Towsley, P. Sparaggis and C.G. Cassandras, Optimal routing and buffer allocation for a class of finite capacity queueing systems, IEEE Trans. Autom. Contr. 37 (1992) 1446–1451.Google Scholar
  38. [38]
    P.P. Varaiya, J. Walrand and C. Buyukkoc, Extensions of the multiarmed bandit problem: the discounted case, IEEE Trans. Autom. Contr. 30 (1985) 426–439.Google Scholar
  39. [39]
    O.A. Vasicek, An inequality for the variance of waiting time under a general queueing discipline, Oper. Res. 25 (1977) 879–884.Google Scholar
  40. [40]
    J. Walrand,An Introduction to Queueing Networks (Prentice Hall, Englewood Cliff, NJ, 1988).Google Scholar
  41. [41]
    R.R. Weber, On the optimal assignment of customers to parallel queue, J. Appl. Prob. 15 (1978) 406–413.Google Scholar
  42. [42]
    G. Weiss, Branching bandit processes, Prob. Eng. Inf. Sci. 2 (1988) 269–278.Google Scholar
  43. [43]
    W. Whitt, Deciding which queue to join, Oper. Res. 34 (1986) 55–62.Google Scholar
  44. [44]
    W. Winston, Optimality of the shortest line discipline, J. Appl. Prob. 14 (1977) 181–189.Google Scholar

Copyright information

© J.C. Baltzer AG, Science Publishers 1995

Authors and Affiliations

  • Zhen Liu
    • 1
  • Philippe Nain
    • 1
  • Don Towsley
    • 2
  1. 1.INRIASophia Antipolis CedexFrance
  2. 2.Department of Computer ScienceUniversity of MassachusettsAmherstUSA

Personalised recommendations