Research On Resource Optimization Of Optical Interconnection

In recent years,with the rapid development of the Internet,as well as the rise of cloud computing,data centers and other new concepts,network traffic is growing at an unprecedented rate.Optical network technology has become the main network technology.Optical interconnects has been widely used in various types of networking environments,and have achieved good results.With further study to optical network technology,Elastic Optical Networks(EONs)have demonstrated agile spectrum management and smaller switching granularity.Compared with the traditional WDM/DWDM network,EONs showing an incomparable advantages.Because of the flexibility for spectrum resource utilization,EONs has become major focus of today's research direction.Meanwhile,the development of cloud computing has led to the rise of the data center.The rapid growth of internet users,the changing service models cause the traffic of inter-DC network is growing.The issue of the applying EONs in Inter-DC networks,and set up Inter-DC networks over EONs have attracted intensive interests from both academia and industry.This paper focus on the resource optimization of Inter-DC networks over EONs.The problem of resource optimization in EONs is Routing,Modulation,Spectrum Assignment problem(RMSA).Compared with the Routing Wavelength Assignment problem(RWA)in traditional WDM/DWDM network,the RMSA problem with smaller switching granularity have higher computing complexity.Currently,to solve such problems,there are two main methods:heuristic and linear programming.The heuristic algorithm is not very good for global optimization,and the linear programming algorithm can get global optimum results,but limited computing time.This paper mainly completed the following work:Survey the RMSA problem in EONs and the Inter-DC networks,learning the current research status and unresolved issues of domestic and international;Focus on the offline networking planning problem of Inter-DC networks over EONs,proposed an Integer Linear Programming(ILP)model to optimize resource utilization and load balancing to target.Select the GNU Liner Programming Kit(GLPK)as simulation tool,and analyze the algorithm of main solver drives.Completed the solving of mathematical model;Using the heuristic algorithm,comparing the result with ILP and draw conclusions.