@techreport{5ccc13eca00e432a9cc15a698b0f8275,
title = "Extreme values for the waiting time in large fork-join queues",
abstract = "We prove that the scaled maximum steady-state waiting time and the scaled maximum steady-state queue length among \$N\$ \$GI/GI/1\$-queues in the \$N\$-server fork-join queue, converge to a normally distributed random variable as \$N\textbackslash{}to\textbackslash{}infty\$. The maximum steady-state waiting time in this queueing system scales around \$\textbackslash{}frac\{1\}\{\textbackslash{}gamma\}\textbackslash{}log N\$, where \$\textbackslash{}gamma\$ is determined by the cumulant generating function \$\textbackslash{}Lambda\$ of the service distribution and solves the Cram\textbackslash{}'er-Lundberg equation with stochastic service times and deterministic inter-arrival times. This value \$\textbackslash{}frac\{1\}\{\textbackslash{}gamma\}\textbackslash{}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 \$\textbackslash{}frac\{\textbackslash{}sigma\_A\}\{\textbackslash{}sqrt\{\textbackslash{}Lambda'(\textbackslash{}gamma)\textbackslash{}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 = "math.PR, cs.PF",
author = "Dennis Schol and Maria Vlasiou and Bert Zwart",
year = "2023",
month = sep,
day = "15",
doi = "10.48550/arXiv.2309.08373",
language = "English",
publisher = "ArXiv.org",
type = "WorkingPaper",
institution = "ArXiv.org",
}