Transient analysis of quasi-birth-death processes

The concept of Complete Level Crossing Information is shown to lead to a matrix geometric solution for the transform of the distribution of the number in queue at each time t in queues modeled as Quasi-Birth-Death (QBD) processes. Specifically, if (t) is the vector of probabilities of state occupancy at time t for the states on level n of a QBD-process given the process starts from level 1 at , and if is the Laplace transform of , then there exists a square matrix W(s) such that: We also show how to determine the boundary vectors and , for this recursion To illustrate this method, a QBD process which models a CSMA/CD network is analyzed in detail. The mean number of busy users in the network and the tail distribution, both of which are functions of time, are computed from the closed-form expressions for the transform domain recursion. These results are significant since they should greatly expand the class of queues for which transient analysis is tractable.

[1]  R. Bellman.  Introduction To Matrix Analysis By Richard Bellman, New York, 1960 , 1960 .

[2]  M. Ackroyd.  M/M/1 Transient State Occupancy Probabilities Via the Discrete Fourier Transform , 1982, IEEE Trans. Commun..

[3]  Vincent Hodgson,et al.  The Single Server Queue. , 1972 .

[4]  Amedeo R. Odoni,et al.  An Empirical Investigation of the Transient Behavior of Stationary Queueing Systems , 1983, Oper. Res..

[5]  Erhan Çinlar,et al.  Introduction to stochastic processes , 1974 .

[6]  Kai Lai Chung,et al.  Markov Chains with Stationary Transition Probabilities , 1961 .

[7]  George Yankovsky,et al.  A handbook of methods of approximate Fourier transformation and inversion of the Laplace transformation , 1977 .

[8]  Pierce E. Cantrell,et al.  Comparison of generalized Q- function algorithms , 1987, IEEE Trans. Inf. Theory.

[9]  C. Loan,et al.  Nineteen Dubious Ways to Compute the Exponential of a Matrix , 1978 .

[10]  Bruce Hajek,et al.  Birth-and-death processes on the integers with phases and general boundaries , 1982, Journal of Applied Probability.

[11]  Marcel F. Neuts,et al.  Matrix-Geometric Solutions in Stochastic Models , 1981 .

[12]  Edward J. Coyle,et al.  A Matrix Representation of CSMA/CD Networks , 1985, IEEE Trans. Commun..

[13]  Edward J. Coyle,et al.  State space expansions and the limiting behavior of quasi-birth-and-death processes , 1989, Advances in Applied Probability.

[14]  E. Coyle,et al.  Finite Population CSMA/CD Networks , 1983, IEEE Trans. Commun..

[15]  W. Grassmann.  Transient solutions in Markovian queues : An algorithm for finding them and determining their waiting-time distributions , 1977 .

[16]  Pierce E. Cantrell.  Computation of the Transient M/M/1 Queue cdf, pdf, and Mean with Generalized Q-Functions , 1986, IEEE Trans. Commun..

[17]  G. Honig,et al.  A method for the numerical inversion of Laplace transforms , 1984 .

[18]  T. Stern.  Approximations of Queue Dynamics and Their Application to Adaptive Routing in Computer Communication Networks , 1979, IEEE Trans. Commun..