Research On Traffic Elasticized Scheduling And Differentiated Management Optimization Strategy For Cloud Data Center

With the rapid development of information technologies such as cloud computing,big data and artificial intelligence,more and more emerging businesses are converging on the cloud.Cloud data center is the information infrastructure of cloud business.With the growth of business and the number of users,the traffic scale of cloud data center has exploded,which brings new challenges to cloud data center traffic scheduling and management.Moreover,the rapid development of new formats such as Augmented Reality(AR)and Virtual Reality(VR)in the 5G era,the surge in cloud data center traffic will push to a new peak.On the one hand,most cloud data centers still use traditional data center network architectures and technologies.The traditional distributed IP network architecture is rigid,and traffic scheduling and management technologies have been difficult to adapt to the requirements of its network management,especially operator's cloud data centers that carry multiple services and are oriented towards multiple users.On the other hand,the diversification of services in cloud data center and the differentiation of user experience,traffic scheduling and management must be able to enable global,dynamic,adaptive,elastic,and differentiated optimization,to achieve efficient use of resources in cloud data centers and enhance network performance and user business experience.Software-defined network(SDN)as a new generation network architecture can promote network management and business support innovation,realize global traffic scheduling and fine-grained management optimization,which can provide a good system architecture and innovation for the traffic scheduling and management optimization in cloud data centers.The thesis focuses on the actual needs of network management and business support of the operator's cloud data center,and conducts in-depth exploration and research on key issues such as traffic elasticized load balancing scheduling,multi-service differential traffic management optimization,and traffic scheduling and management optimization algorithm design.To achieve the goal of flexible scheduling and differentiated management in operator cloud data center traffic.The main innovations and contributions in this dissertation are as follows:1.The elasticized traffic scheduling strategy in cloud data centers is proposedIn this dissertation,considering both the network structure and network traffic based on the actual network management in the operator's cloud data center,and a traffic analysis and prediction online scheduling mechanism is designed.Aiming at the multi-dimensional,multiconstrained and multi-modal problems of traffic scheduling in cloud data center,an elasticized traffic scheduling strategy for cloud data centers is innovatively proposed.In the two stages of analysis and prediction and online scheduling,the Fibonacci tree optimization algorithm(FTO)is used to optimize the global local alternate iteration and multi-modal and the optimal solution for traffic scheduling and multiple valuable sub-optimal solutions are obtained.Improving the elasticized scheduling capability of operator cloud data center traffic.The simulation platform verification shows that the proposed strategy can reasonably scheduling the traffic of the cloud data center and effectively improve the load balancing capability of the operator's cloud data centers.2.The multi-service differentiated traffic management optimization(MSD-FTO)strategy in cloud data centers is proposedIn order to cope with the traffic management for multi-service differentiated in cloud data centers,improving network performance and service experience,the multi-service differentiated(MSD)traffic management model was designed that can suit operational requirements in cloud data center.Fibonacci tree optimization(FTO)algorithm was improved according to the MSD model.MSD-FTO traffic management strategy was proposed in SDN cloud data center.Simulation results show that the strategy takes advantage of FTO global optimization ability and multi-modal adaptive performance.Through the global local alternating optimization of the algorithm,differentiation traffic management schemes are obtained as needed,the problem of multi-services differentiated traffic management is solved in cloud data center that improve network performance and service experience in cloud data center effectively.3.FTO algorithm for traffic scheduling and management is designed and improvedIn this dissertation,the traffic scheduling and management of the operator cloud data center as the research object,according to the actual network management and business support,the model that is designed to adapt its traffic characteristics and optimization goals.The FTO algorithm for the traffic scheduling and management is designed and improved innovatively,to solve the multi-dimensional,multi-constrained,multi-mode adaptive optimization problem of the traffic scheduling and management in cloud data center.The experimental results show that within the feasible region of traffic scheduling and management,the stable global optimization capabilities and flexible multi-mode adaptive characteristics of FTO can obtain multiple optimized schemes of the traffic scheduling and management optimization at one time.It satisfies the needs for cloud data center traffic elasticized scheduling and differentiated management of multi-service traffic,and can effectively solve cloud data center traffic scheduling and management optimization issues,and achieve the purpose of improving network performance and business service quality.To sum up,this thesis studies the traffic elastic scheduling and differentiated management optimization strategy in operator cloud data centers.The elasticized load balancing capability and the multi-service differentiated traffic management capability in operator cloud data center can be improved.It's helpful to the traffic scheduling and management optimization in operators' large-scale cloud data centers.And it's valuable for the study of cloud data center traffic scheduling and management strategies has theoretical and practical significance.