School Seminars and Colloquia

A quantitative differential equation approximation for a routing model

Stochastic Processes Seminar

by Malwina Luczak


Institution: School of Mathematics, Queen Mary, University of London
Date: Thu 24th January 2013
Time: 2:15 PM
Location: Old Geology 1

Abstract: We consider a Markovian load balancing model on a fully-connected network, where calls have Poisson arrivals and exponential durations. The endpoints of each call are uniform over all the links of the network. Each call is routed either along the link connecting its endpoints, or, if the direct route is unavailable, along a two-link path between them, via an intermediate node. We use an explicit and simple coupling to show a strong concentration of measure property, and deduce that the evolution of the process may be approximated by a differential equation. The technique is likely to be useful in other settings.