Contention Resolution Based On Node Trust Classifier To Guarantees Quality Of Service In Optical Burst Switched Network

Abstract

In optical burst switching (OBS) network, burst loss that specifically caused by burst contending to reserve the
network resource at the intermediate core switch is considered one of the key issues today. Burst segmentation is
one way that resolves contention which considered being an effective solution at current, since it can enhance the
quality of service (QoS) of OBS network’s resource utilization and throughput. Due to the lack of suitable
contention resolution technique in reserving the network resources to transmit data burst (DB); bursts loss is high
in OBS network. Contention happens when multiple bursts from different input ports are attempting to reserve
the same wavelength Division Multiplexing (WDM) channel on the same outgoing port at the same time. In this
paper, we propose and develop a new quality of service model that can be embedded into OBS core switch
architecture to enhance the node’s performance and the quality of burst transmission. The model allows the OBS
core switch to classify the ingress nodes into four trust classes and assign the DB into three priority levels based
on their behavior and the amount of reserved resources that are not being utilized. The contending node that
causes burst loss will be blocked by the developed model until the node improves its performance. The quality of
service model is implemented, tested and verified using a modified NCTUns network simulator. The analysis
conducted reveals that our proposed model is effective in enhancing node performance and the DB transmission
as well as in providing the network resources to the well behaved nodes.