Analysis of polling systems
Deducing queueing from transactional data: the queue inference engine, revisited
Operations Research - Supplement to Operations Research: stochastic processes
| Hi-index | 0.00 |
Queue inferencing algorithms are used to derive estinates of queue lengths and/or customer waiting times from a priori information about the customer arrival process and the observed sequence of times at which each customer enters and leaves service. In this paper, we extend these techniques by decoupling the arrival time constraints from the customer departure times, which allows us to handle additional features like server vacations. We then show how these techniques can be used to monitor a single station in a polling system, or in a shared medium Local Area Network such as Ethernet, Token Ring and FDDI. Using these results, passive, non-intrusive network monitoring tools could be developed to estimate waiting times and queue lengths for any host on the network by observing only the packet departure times from the nodes.