Optimized burst assembly algorithm for multi-ranked traffic over optical burst switching network

Abstract

Optical Burst Switching (OBS) is seen as the most vital technology for the coming era of Internet backbone infrastructure. The OBS cardinal role, especially for data of high priority belonging to real time systems, makes optimizing the OBS network one of the significant fields of research. Thus, OBS network fulfillment of all requisites and compulsions for optimized performance for all classes of traffic in terms of packet loss and end to end delay is a vital issue of study. This research presents a new optimized multi-class burst assembly algorithm over OBS network to ensure an enhanced performance via appropriate network service that could reduce packet loss rate and delay, especially for high priority classes of data, by the mean of preventing contention along OBS resources. A simulation model using National Chiao Tung University network simulator (NCTUns) simulator has been used to evaluate the performance of the proposed schemes, where the scheme is runned with three types of traffic: Constant Bit Rate (CBR), Variable Bit Rate (VBR) and Available Bit Rate (ABR). The traffic classes served by the proposed scheme can be extended to N number of classes, Where N value is calculated upon the burst size. The scheme is named Multi-Class Adaptive Burst Assembly (MC-ABA), and it represents an optimized version of the Real-Time Adaptive Burst Assembly (RT-ABA), where the MC-ABA not only serves N number of classes of Internet traffic over OBS networks but also considers the intensity of high priority data within each assembled burst to avoid contention and reduces packet loss and delay of high priority data over OBS core nodes. Simulation results showed that MC-ABA scheme could reduce the end-to-end delay and packet loss, besides providing suitable service for all types of data traffic. Generally, the proposed scheme can improve OBS network to be an appropriate environment for high priority traffic.