Enhanced Bandwidth Reconfiguration for On-Demand QoS Path Framework

This paper proposes an enhancement to a novel end-to-end QoS framework that has been recently introduced in the literature and referred to as the On Demand QoS Path (ODP). The ODP provides the end-to-end QoS guaranty of the Integrated Service (IntServ) while keeping the scalability nature of the Differentiated Service (DiffServ). The proposed enhancement, named E-ODP, efficiently reconfigures network bandwidth through the utilization of the self-similarity feature of network traffic. Specifically, bandwidth reconfiguration is handled using traffic prediction, fuzzy control, and adaptive learning. Simulation results are presented. They demonstrate a significant and persistent improvement in network utilization.