Researches On Resource Optimization Method Of Software Defined Elastic Optical Network

The increasing of Internet users and devices has put forward higher requirements for flexible bandiwidth allocation of the Space-Earth Integrated Information Network.With the advantages of controllable and resilient,the Software Defined Network technology will become a possible solution for the future Space-Earth Integrated Information Network architecture.To meet the diverse service requirements of satellite information network,it is necessary to expand its service elasticity and research the differentiation resource allocation customized.To cope with the growth of bandwidth-consuming requests of terrestrial backbone network,it is necessary to expand its rate elasticity to realize multi-rate carrying and optical path distribution on demand.In this paper,in view of the diverse differentiated service and large bandwidth transmission demands faced by the Space-Earth Integrated Information Network,aiming at rasing utilization efficiency of spectrum resources,the service elasticity of satellite optical network and the rate elasticity of terrestrial optical network are expanded,and research on the resource optimization methods of Software Defined Resilience Optical Network is carried out.The paper focuses on the hybrid service resource allocation scheme,the low dispersion degree elastic resource allocation algorithm,the elastic resource allocation based on reprovision,and the static spectrum allocation algorithm of space division multiplexing spectrum flexible optical network.The main research contents of this paper are as follows:(1)Hybrid service resource allocation scheme for software defined satellite optical networksTo meet the requirements of diverse services,a resource allocation scheme for hybrid service in software defined satellite optical networks is proposed.The scheme serves the hybrid service composed of Immediate Reservation(IR)and Advanced Reservation(AR)service,and allocates wavelength and timeslot for the service according to the global network state information.Based on the mathematical model of satellite optical network,hybrid service,and resource allocation,an Elastic Resource Allocation(ERA)algorithm for hybrid service in satellite optical network is proposed.Simulation results show that the resource allocation scheme with software defined network architecture achieves flexible and fine-grained dominate of network resource.Compared with the Nodiff-RA algorithm which does not distinguish IR and AR services,the ERA algorithm can reduce the blocking rate of hybrid services by about 27.2%and improve the wavelength utilization by about 2.91%when AR services account for 50%and the traffic intensity is 110 Erl.(2)Elastic resource allocation algorithm based on low dispersion degree for hybrid services in satellite optical networksTo improve the adverse impact of resource fragmentation on network performance in satellite optical networks,based on the proposed resource allocation scheme for satellite optical network,an elastic resource allocation algorithm considering the degree of fragmentation is studied.Combined with the dynamic characteristics of satellite optical network,resource fragments occurred in global resource matrice are divided into tail fragments and non-tail fragments.By proving the numerical relationship between the resource slots state jump and the number of resource fragments,the clutter degree of fragments is quantified.A description method of fragment degree called Dispersion Degree is defined,and an Elastic Resource Allocation algorithm based on Dispersion Degree(ERA-DD)is proposed.The restriction of the algorithm on the priority of candidate schemes is also analyzed.Simulation results show that the Dispersion Degree can describe the resource fragmentation accurately,and the ERA-DD algorithm limits the priority of various candidate schemes effectively.Compared with ERA and MTLSC algorithms,the ERA-DD algorithm respectively reduces the traffic blocking rate by 16.8%and 9.21%,and improve the wavelength utilization by 2.82%and-0.22%when AR traffic accounts for 50%and the traffic intensity is 110 Erl.(3)Elastic resource allocation algorithm based on Reprovision for hybrid services in satellite optical networksTo improve the adverse effect of advanced resource reservation on IR traffic in hybrid services,an Elastic Resource Allocation algorithm based on Reprovision strategy(ERA-RP)is proposed.By designing the reprovision strategy of AR requests,the scheme optimizes the resource allocation of AR traffic which conflicts with IR request.The simulation results show that the reprovision strategy makes improvements on IR traffic blocking rate and wavelength utilization.Compared with the ERA-DD algorithm,the ERA-RP algorithm can reduce the blocking rate of IR traffic by 62.2%and increase the wavelength utilization by 16.1%when AR service accounts for 50%,reprovision interval of AR service is 5 s and the traffic intensity is 110 Erl.(4)Static spectrum allocation algorithm for SDM spectrum flexible optical networksWith the growth of various bandwidth consumption applications,the terrestrial backbone network of the Integrated Information network is evolving towards spectrum flexible optical network and SDM optical network.To deploy the spectrum resources needed in terrestrial backbone networks,the joint allocation of Routing,Modulation levels,Spatial dimension and Spectrum resources(RMSSA)is studied.Based on the research of adaptive modulation strategy,the spatial switching mode and other factors,an ILP model of RMSSA problem is constructed,and the Local Optimized RMSSA(LO-RMSSA)heuristic algorithm is proposed to compute in large-scale networks.Simulation results show that the proposed ILP model can solve the static RMSSA problem effectively,and the LO-RMSSA algorithm can approach the solution of ILP model greatly.Compared with the SGA algorithm,the LO-RMSSA algorithm can reduce the network frequency slots by about 28.24%when a single bandwidth request is 60?80 Gbps.