Skip to Main Content
In this paper we have developed a new service classification strategy in hybrid scheduling scheme to support differentiated quality of service (QoS) among the different set of clients. The scheme dynamically computes the data access probabilities and amalgamates the push and pull scheduling schemes to develop the hybrid scheduling framework. While a flat scheduling is used for push system, the major novelty of the work lies in differentiating the clients based on their priority-classes and incorporating the effect of priority in selecting an item from the pull-system. Modeling and analysis of the system is performed to get an average behavior of the QoS parameters like delay in our hybrid scheduling framework. Simulation results points out that the average waiting time for the highest priority clients can be kept very low, while simultaneously minimizing the number of requests dropped by assigning appropriate fraction of available bandwidth. It also demonstrates that by intelligent selection of the cut-off point, used to segregate push and pull systems, the overall cost associated with the system can be minimized.
Date of Conference: 14-17 June 2005