IEEE/ACM Transactions on Networking (TON)
IEEE/ACM Transactions on Networking (TON)
Large deviations and the generalized processor sharing scheduling for a multiple-queue system
Queueing Systems: Theory and Applications
Large deviations analysis of the generalized processor sharing policy
Queueing Systems: Theory and Applications
A Reduced-Load Equivalence for Generalised Processor Sharing Networks with Long-Tailed Input Flows
Queueing Systems: Theory and Applications
Testing the Gaussian approximation of aggregate traffic
Proceedings of the 2nd ACM SIGCOMM Workshop on Internet measurment
Reduced-Load Equivalence and Induced Burstiness in GPS Queues with Long-Tailed Traffic Flows
Queueing Systems: Theory and Applications
Generalized processor sharing with light-tailed and heavy-tailed input
IEEE/ACM Transactions on Networking (TON)
Optimal call admission control on a single link with a GPS scheduler
IEEE/ACM Transactions on Networking (TON)
Sample-path large deviations for generalized processor sharing queues with Gaussian inputs
Performance Evaluation - Long range dependence and heavy tail distributions
Traffic theory and the Internet
IEEE Communications Magazine
Characterizing Superposition Arrival Processes in Packet Multiplexers for Voice and Data
IEEE Journal on Selected Areas in Communications
Analysis of a two-class FCFS queueing system with interclass correlation
ASMTA'12 Proceedings of the 19th international conference on Analytical and Stochastic Modeling Techniques and Applications
Hi-index | 0.01 |
We consider a two-class Generalized Processor Sharing (GPS) queueing system, in which each class has its specific traffic characteristics and quality-of-service (QoS) requirements. Traffic of both classes is assumed to be Gaussian (a versatile family of models that covers both long-range dependent and short-range dependent traffic). In this paper we address the question how to select the GPS weight values. To do so, we first characterize the admissible region of the system for fixed weights. Then we obtain the realizable region by taking the union of the admissible regions over all possible weight values. The results indicate that, under a broad variety of traffic characteristics and QoS requirements, nearly the entire realizable region can be obtained by strict priority scheduling disciplines. In addition, we indicate how the buffer thresholds, QoS requirements and the traffic characteristics of the two classes determine which class should get high priority.