Queueing theory without limits: transient and asymptotic analysis

October 17-19, 2007

EURANDOM, Eindhoven, The Netherlands

Programme & Abstracts | Speakers  & Participants | Registration | Practical Information

The organizers are very pleased to announce that the following persons have agreed to be

keynote speakers

Søren Asmussen (Aarhus University, Denmark)
Sir John Kingman (former director Isaac Newton Institute, UK)
William Massey (Princeton University, USA)

tutorial speakers

Michael Drmota (TU Vienna, Austria)
Gennady Samorodnitsky (Cornell University, USA)

There will be 20 speakers in total (see the list below).

Scope

Queueing theory studies systems which require customers or objects to wait. It is among its most important objectives to assess overall system performance, as well as the evolution of this performance over time. However, exact expressions for performance metrics often cannot be found. In such cases, one can resort to approximate or asymptotic methods. This workshop is devoted to the study of queues and stochastic networks, with special attention to both transient (time-dependent) analysis and asymptotic methods.

Transient analysis:

Most queueing systems are studied under the (simplifying) assumption that the system has reached some steady state. However, in various situations it is the time-dependent behavior of systems that is of utmost importance. A key question is: How fast converges the system to equilibrium? This question may be answered using a variety of techniques from probability theory, combinatorics, complex analysis, and numerical analysis.

Asymptotic analysis:

Traditionally, performance metrics are studied by deriving exact expressions for generating functions or Laplace transforms. Inversion of these transforms then gives the distributions of the performance metrics. In many cases, though, inversion is impossible, which makes it hard to obtain these performance characteristics in explicit form. The workshop will focus on situations where asymptotic theory might come to the rescue to get information about the distributions of the performance metrics.

Long-term goal:

The goal of the workshop is to bring together young researchers interested in the field of stochastic processes and queueing theory. The modeling of stochastic networks is a truly multi-disciplinary field and tends to attract researchers with various backgrounds. As a result of this, the community of researchers in the area is rather scattered and it would potentially be extremely fruitful to bring together different groups of researchers. The emphasis of the workshop will be on mathematical/probabilistic aspects.

This workshop is intended to be the first in a series named YEQT meetings, organized within the framework of EURANDOM's program 'Queueing and Performance Analysis'. YEQT stands for Young European Queueing Theorists. The main goal of these meetings would be to bring together young researchers in various related fields, and the scope of the meetings shall vary. It will complement the successful YEP (Young European Probablists) meetings organized by the probability section and YES (Young European Statisticians) meetings organized by the statistics section of EURANDOM.