Stochastic models of systems

From MaRDI portal





This text considers general random systems evolving in a random medium and puts special emphasis on the question how these systems (which may be a priori very large and involve a huge number of parameters) may be simplified such that probability theory becomes applicable to the description of the random system. The authors restrict themselves to the treatment of Markovian models of the random medium in order to avoid complications. The random system is hence determined by two processes: a switched one that describes the evolution of the system and a switching one that describes the medium. The simplification of the system is achieved by a so-called phase merging algorithm which is obtained from the Markov process by restricting it to the set of equivalence classes with respect to a suited equivalence relation on the phase space. After an elementary explanation of the approach and an introduction to renewal processes in Chapter~1, the theory of Markov renewal processes and Markov jump processes is presented in Chapter~2, including regenerative processes and aspects of ergodicity. It is assumed that the processes take values in an arbitrary measurable space whose singletons are measurable. Phase merging algorithms are considered in Chapter~3, after an introduction to reducible-invertible operators, and the martingale characterization of Markov processes. In Chapter~4, the random evolution of ergodic Markov processes is treated, and diffusion approximation results and ergodic theorems are discussed. Finally, Chapter~5 specializes to the cases of Markov queueing systems and networks and presents algorithms of diffusion approximation in these cases. The emphasis is made on developing the mathematical framework for the models such that their simplification by phase merging and diffusion approximation may be embedded in the theory of stochastic processes. Hence, the proofs for limit assertions for long time or diffusion approximation are short since they may be cited from the theory. The explanations are elementary and easy to follow, complicated technicalities are absent. The reference list contains 20 items only, half of them being written in Russian.




Cited in
(82)








This page was built for publication: Stochastic models of systems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4506433)