Research Of The Cross-layer Resource Scheduling Algorithm In IP+Optical Networks

With the increasing popularity of Mobile Internet and explosive growth of Cloud computing,the backbone transport network has suffered greater bandwidth stress and been required to provide more flexibility to accommodate diverse services.However,the traditional telecommunication networks have separate control and management mechanisms on IP and optical layers,which inevitably leads to low resource utilization and poor agility.Thus,the emerging Software Defined Network(SDN)technology is widely studied in that it can dynamically manipulate heterogeneous resources on different layers and improve network performance.In this thesis,we study the unified resource scheduling in IP+Optical network under the SDN architecture regarding the following three aspects:1)a unified information model to describe the cross-layer resources,2)various routing algorithms capable of efficiently scheduling cross-layer resources,3)performance evaluation of various algorithms.The major innovations of this thesis can be summarized as follow:1)A unified information model is built for IP + Optical cross-layer network.In this model,an auxiliary graph is built according to different switching granularities(including packet,timeslot,wavelength,wave band and fiber)supported by IP layer and optical layer in the network,and the properties of edges in the graph model reflect the relationships and constraints between different layers and different granularities.2)Multiple unified IP + Optical cross-layer routing algorithms are proposed based on the cross-layer resource information model.In the proposed algorithm,the weight of each edge is appropriately assigned and dynamically updated according to specific routing strategies,and then a simple weighted-SPF algorithm can be applied to generate the optimized cross-layer switching path.Therefore,we investigate four kinds of cross-layer routing strategies including the IP-preferred strategy,the optical-preferred strategy,the IP-preferred strategy with TE,and the optical-preferred strategy with TE.3)A simulation platform is built in order to evaluate the performance of various algorithms proposed in this paper comprehensively.In the simulation,three network topologies(i.e.NSFNET,Ring and Grid)and two traffic scenarios(i.e.large-granularity traffic and small-granularity traffic)are used,and the network performance in terms of blocking rate,average path length and the energy consumed by inter-layer transceivers are measured accordingly.