The Research On QoS Resolution Of OBS Network

With the explosive growth of Internet business, the traditional network to a multi-service network in the direction of development, and in recent years integrated network development, the IP suited to become the technology of choice for all kinds of business. In optical networking, WDM technology provides enormous bandwidth, and has become the undisputed backbone network in most major transmission technology. IP and optical network integration can supplement each other's advantages. It is clear that IP over WDM is imperative.The IP over WDM is a key issue in the light of how to support Quality of Service (QoS). In this paper, we first make a comprehensive and systematic study about the basic principle, architecture and key technologies of OBS network. Next, we discuss several of providing QoS support Resolutions: additional offset-time, wavelength division, the priority level of competition strategy, and the proportional QoS, and introduce a method QoS Resolutions based on FLR-M. Then, we construct a three-level simulation platform on the basis of this theory, with OPNET network simulation tools from Network level, Node level, and Process level. The design of the Node model of the core node and edge node are introduced particularly, as well as the import module and the export module of the edge Node model, the processing model of the core Node model and three link model, such as IP_link, BDP_link and BCP_link. The simulation results show that the introduction of PLR-M algorithm to replace initiative discarded algorithm, maintaining QoS controllability at the same time, significantly improves the unexpected high rate of data loss, increases the efficiency of the QoS control.For a carrier network, QoS controllability is a basic requirement. On how to absorb the advantage of QoS Resolution based on offset-time, and further enrich the contents of QoS, how to maintain and strengthen the program of controllable capacity of the QoS Resolution, and how to determine the value of parameters M in PLR-M algorithm, will be the next step research work.