Uncovering the topology of an infinite-server queueing network from population data
By: Hritika Gupta , Michel Mandjes , Liron Ravner and more
Potential Business Impact:
Helps understand how many people are waiting.
This paper studies statistical inference in a network of infinite-server queues, with the aim of estimating the underlying parameters (routing matrix, arrival rates, parameters pertaining to the service times) using observations of the network population vector at Poisson time points. We propose a method-of-moments estimator and establish its consistency. The method relies on deriving the covariance structure of different nodes at different sampling epochs. Numerical experiments demonstrate that the method yields accurate estimates, even in settings with a large number of parameters. Two model variants are considered: one that assumes a known parametric form for the service-time distributions, and a model-free version that does not require such assumptions.
Similar Papers
Analyzing homogenous and heterogeneous multi-server queues via neural networks
Performance
Predicts how many people will be waiting.
Online Distributed Queue Length Estimation
Data Structures and Algorithms
Lets computers guess how many things are waiting.
Geometric lower bounds for the steady-state occupancy of processing networks with limited connectivity
Probability
Makes computer networks handle more tasks faster.