Introduction to queueing theory (2nd ed)
Introduction to queueing theory (2nd ed)
Asymptotic expansion for large closed queuing networks
Journal of the ACM (JACM)
A network of priority queues in heavy traffic: one bottleneck station
Queueing Systems: Theory and Applications
Discrete flow networks: bottleneck analysis and fluid approximations
Mathematics of Operations Research
On Harris recurrence in continuous time
Mathematics of Operations Research
Strong approximations for time-dependent queues
Mathematics of Operations Research
Strong approximations for Markovian service networks
Queueing Systems: Theory and Applications
A large closed queueing network with autonomous service and bottleneck
Queueing Systems: Theory and Applications
Asymptotic analysis of closed queueing networks with bottlenecks
Proceedings of the IFIP WG 7.3 International Conference on Performance of Distributed Systems and Integrated Communication Networks
On Existence of Limiting Distribution for Time-Nonhomogeneous Countable Markov Process
Queueing Systems: Theory and Applications
Queueing Systems: Theory and Applications
Continuity theorems for the M/M/1/n queueing system
Queueing Systems: Theory and Applications
Hi-index | 0.00 |
The paper provides the up- and down-crossing method to study the asymptotic behavior of queue-length and waiting time in closed Jackson-type queueing networks. These queueing networks consist of central node (hub) and k single-server satellite stations. The case of infinite server hub with exponentially distributed service times is considered in the first section to demonstrate the up- and down-crossing approach to such kind of problems and help to understand the readers the main idea of the method. The main results of the paper are related to the case of single-server hub with generally distributed service times depending on queue-length. Assuming that the first k−1 satellite nodes operate in light usage regime, we consider three cases concerning the kth satellite node. They are the light usage regime and limiting cases for the moderate usage regime and heavy usage regime. The results related to light usage regime show that, as the number of customers in network increases to infinity, the network is decomposed to independent single-server queueing systems. In the limiting cases of moderate usage regime, the diffusion approximations of queue-length and waiting time processes are obtained. In the case of heavy usage regime it is shown that the joint limiting non-stationary queue-lengths distribution at the first k−1 satellite nodes is represented in the product form and coincides with the product of stationary GI/M/1 queue-length distributions with parameters depending on time.