English

Extreme values for the waiting time in large fork-join queues

Probability 2023-09-18 v1 Performance

Abstract

We prove that the scaled maximum steady-state waiting time and the scaled maximum steady-state queue length among NN GI/GI/1GI/GI/1-queues in the NN-server fork-join queue, converge to a normally distributed random variable as NN\to\infty. The maximum steady-state waiting time in this queueing system scales around 1γlogN\frac{1}{\gamma}\log N, where γ\gamma is determined by the cumulant generating function Λ\Lambda of the service distribution and solves the Cram\'er-Lundberg equation with stochastic service times and deterministic inter-arrival times. This value 1γlogN\frac{1}{\gamma}\log N is reached at a certain hitting time. The number of arrivals until that hitting time satisfies the central limit theorem, with standard deviation σAΛ(γ)γ\frac{\sigma_A}{\sqrt{\Lambda'(\gamma)\gamma}}. By using distributional Little's law, we can extend this result to the maximum queue length. Finally, we extend these results to a fork-join queue with different classes of servers.

Keywords

Cite

@article{arxiv.2309.08373,
  title  = {Extreme values for the waiting time in large fork-join queues},
  author = {Dennis Schol and Maria Vlasiou and Bert Zwart},
  journal= {arXiv preprint arXiv:2309.08373},
  year   = {2023}
}
R2 v1 2026-06-28T12:22:35.270Z