Optik çoğuşma anahtarlamalı ağlarda akış denetimi ve hizmet ayırımı

ABSTRACTFLOW CONTROL AND SERVICEDIFFERENTIATION IN OPTICAL BURSTSWITCHING NETWORKSHakan BoyrazM.S. in Electrical and Electronics EngineeringSupervisor: Asst. Prof Dr. Nail AkarApril 2005Optical Burst Switching (OBS) is being considered as a candidate architecturefor the next generation optical Internet. The central idea behind OBS is the as-sembly of client packets into longer bursts at the edge of an OBS domain and thepromise of optical technologies to enable switch reconfiguration at the burst leveltherefore providing a near-term optical networking solution with finer switchinggranularity in the optical domain. In conventional OBS, bursts are injected tothe network immediately after their assembly irrespective of the loading on thelinks, which in turn leads to uncontrolled burst losses and deteriorating perfor-mance for end users. Another key concern related to OBS is the difficulty ofsupporting QoS (Quality of Service) in the optical domain whereas support ofdifferentiated services via per-class queueing is very common in current electroni-cally switched networks. In this thesis, we propose a new control plane protocol,called Differentiated ABR (D-ABR), for flow control (i.e., burst shaping) andservice differentiation in optical burst switching networks. Using D-ABR, weshow with the aid of simulations that the optical network can be designed towork at any desired burst blocking probability by the flow control service of theiiiproposed architecture. The proposed architecture requires certain modificationsto the existing control plane mechanisms as well as incorporation of advancedscheduling mechanisms at the ingress nodes; however we do not make any spe-cific assumptions on the data plane of the optical nodes. With this protocol, it ispossible to almost perfectly isolate high priority and low priority traffic through-out the optical network as in the strict priority-based service differentiation inelectronically switched networks. Moreover, the proposed architecture moves thecongestion away from the OBS domain to the edges of the network where it ispossible to employ advanced queueing and buffer management mechanisms. Wealso conjecture that such a controlled OBS architecture may reduce the numberof costly Wavelength Converters (WC) and Fiber Delay Lines (FDL) that areused for contention resolution inside an OBS domain.Keywords: Optical burst switching, rate control, service differentiation, conges-tion control.iv