arXiv Analytics

Sign in

arXiv:1002.0796 [math.PR]AbstractReferencesReviewsResources

Metastability in communication networks

D. Tibi

Published 2010-02-03, updated 2010-08-01Version 2

Two models of loss networks, introduced by Gibbens et al. and by Antunes et al., are known to exhibit a mean field limiting regime with several stable equilibria. These models are reexamined in the light of Freidlin and Wentzell's large deviation approach of randomly perturbed dynamical systems. Assuming that some of their results still hold under slightly relaxed conditions, the metastability property is derived for both systems. A Lyapunov function exhibited by Antunes et al. is next identified with the quasipotential associated with a slightly modified, asymptotically reversible, Markovian perturbation of the same dynamical system. Another interpretation, in terms of entropy dissipation, of this Lyapunov function is finally given. The argument extends to another, similar but closed model.

Related articles: Most relevant | Search more
arXiv:2005.07288 [math.PR] (Published 2020-05-14)
Limit theorems for a random walk with memory perturbed by a dynamical system
arXiv:1011.3414 [math.PR] (Published 2010-11-15)
Small Random Perturbations of a Dynamical System with Blow-up
arXiv:1911.01748 [math.PR] (Published 2019-11-05)
$L^2$ hypocoercivity, deviation bounds, hitting times and Lyapunov functions