An efficient real-time service discipline for the control plane to enhance the performance of optical burst switching networks

Optical burst switching (OBS) is widely believed to be the technology for the future core network in the Internet. Traditionally, the burst header packet (BHP) is served by the control processor on first-come-first-serve basis which increases the burst loss probability (BLP) at the core nodes closer to the destination due to insufficient offset time. Efficient real-time service disciplines in the control plane can reduce the BLP due to insufficient offset time by giving higher priority to the BHPs with smaller residual offset time. This in turn improves the throughput of OBS networks and reduces the BLP. This paper studies the role of real-time service disciplines used to service the BHPs in the queue in improving the performance of the OBS networks. We propose Least per-Hop Worst-Case Laxity First (LHWLF) algorithm which picks the BHP with least available delay budget at that node after accounting for the waiting time in the queue. We demonstrate the effectiveness of our LHWLF algorithm by comparing with other competing ones in the literature in terms of BLP and throughput. We also demonstrate that the proposed algorithm gives fair treatment to bursts with different lengths and the bursts travelling paths of different lengths.