Research On Resource Optimization Based On Data Center Optical Network

In recent years, with the rapid development of optical network technology, network bandwidth grew rapidly, and the speed of data transmission was creating new heights again and again. With the convenience and efficiency of the data transmission, a large number of new services have emerged, such as mobile video, online payment, cloud computing, big data,networking, mobile banking and so on. At the same time, data centers have also developed rapidly, as the actual load of these emerging businesses.However, the communication of data centers was also facing great challenges, because of the rapidly increase of traffic and the diversification of business.With the rapid increase in traffic and the diversification of business, data center communication was also facing great challenges. With excellent bearing capacity, the optical network can be priority for the communication of data centers. The elastic optical network, coming out with small switching granularity and excellent flexibility in resource management, can be suitable for the data center optical network,to support the growing demands and increasingly diverse business needs for data centers. Therefore, to study the planning method for the data center network based on elastic optical network was of great significance.This thesis focus on the RMSA planning problem in the data center network based on elastic optical network. Firstly, this paper introduced the present situation of the data centers and the optical network; then the relevant basic knowledge as the basis of the theory behind; next, emphasized on the knowledge of GLPK tools and the principle of Branch-and-cut Method;finally, designed the model for the characteristics of RMSA problems, and then simulated it and analyzed the results. The results showed that, with the design and modification for RMSA problems, the final model had a better load balancing performance, and the convergence speed improved. At the same time,the maximum occupied frequency gap and the average routing length are relatively large. It provided a solution to solve the RMSA problems in optical network of data center.