Data networks
A maximum entropy priority approximation for a stable G/G/1 Queue
Acta Informatica
MEM for arbitrary queueing networks with multiple general servers and repetitive-service blocking
Performance Evaluation - Queueing networks with finite capacity queues
On the self-similar nature of Ethernet traffic (extended version)
IEEE/ACM Transactions on Networking (TON)
LCN '02 Proceedings of the 27th Annual IEEE Conference on Local Computer Networks
Approximate Analysis of Arbitrary QNMs with Space and Service Priorities
Proceedings of the IFIP TC6 WG6.3/WG6.4 Fifth International Workshop on Performance Modelling and Evaluation of ATM Networks: Performance Analysis of ATM Networks
Entropy maximisation and open queueing networks with priorities and blocking
Performance Evaluation
Time Scale Analysis of an ATM Queueing System With Long-Range Dependent Traffic
INFOCOM '97 Proceedings of the INFOCOM '97. Sixteenth Annual Joint Conference of the IEEE Computer and Communications Societies. Driving the Information Revolution
A Study of Flow-based Traffic Admission Control Algorithm in the ATM-based MPLS Network
ICOIN '01 Proceedings of the The 15th International Conference on Information Networking
Analysis and Modeling of Voice over IP Traffic in the Real Network
IEICE - Transactions on Information and Systems
Computer and Communication Networks
Computer and Communication Networks
Priority queues with batch Poisson arrivals
Operations Research Letters
Performance analysis of priority queueing systems in discrete time
Network performance engineering
Hi-index | 0.01 |
M^X/G/1 non-preemptive priority queue with bulk arrivals has already been analysed in many papers. Two problems are known in this area: first, when the number of priorities is greater than 2, the analysis for priority processes is obviously very difficult to handle. Therefore, the earlier work on the subject was restricted mostly to two priority classes; and second, although analytically explicit results are available, they require sophisticated closed-form expressions of the mean queue length. One particular bulk size distribution - the GE distribution - is motivated by an ME (Maximum Entropy) formulation for the behavior of a G/G/1 queue. The choice of a GE distribution is motivated by the fact that measurements of actual inter-arrival traffic or service times may be generally limited and so only few parameters, such as mean and variance, can be computed reliably. Thus this paper we can obtain very simple and analytic closed-form expression for the mean queue length of a GE/G/1 priority queue and a significant increase in performance evaluation has been achieved. We present the four-class priority queues, performance analysis and simulation of the LER (Label Edge Router) system in the ATM-based MPLS (Multiprotocol Label Switching) network. We wish to obtain the boundary conditions of the mean queue lengths and the mean queueing delays for each priority class, since this metric is one of the most important in performance evaluation parameters for improving QoS and system performance of the LER system in ATM-based MPLS network. A significant numerical example is presented and discussed as well. In order to obtain optimizing the performance analysis for EF flow, AF 1 flow, AF 2 flow and BE flow, the optimum ratios of COV (Coefficient of Variation) can be found via many numerical experiments carried out by the authors for queueing network model with HOL (Head of Line) priority rules. However, the ratios of COV value constraints exist. Furthermore, we find that each service class gradually begins to deteriorate when SQVs (Squared Coefficient of Variations), C"a"i^215.524, C"s"i^212.96 and traffic intensity is greater than 0.95. We also find the values of maximum allowed burst size for EF flow and AF 1 flow and perform necessary policing actions on EF flow and AF 1 flow at the boundary node of the network. Finally, the four-class GE/G/1 priority queues and performance analysis of the LER system are shown accurate and robust after the comparison between theoretical evaluates and computer simulation results.