Design And Implementation Of A Virtualized Network Resource Optimization System Based On Multipath Transmission In Cloud Environment

In recent years,with the continuous development of cloud computing technology and the gradual popularity of cloud application ecology,it has been an important trend to store data and applications in cloud data centers with cloud computing technology as the core.To ensure the flexibility and reliability of cloud computing systems,virtualization technology has been introduced into the cloud environment to provide flexible resource management and high availability support for cloud computing platforms.The virtualized network carries the communication between different modules and virtual machines in the cloud computing system,and its performance affects the performance of the whole cloud computing environment.The optimization of the virtualized network is the key to improve the performance and reliability of the cloud computing system.Based on the multiple physical and virtual paths existing in virtualized network of the existing cloud environment,deploying multipath transmission in virtualized network can effectively improve the performance and availability of virtualized network,but this also brings challenges in path planning and path deployment,which are 1)difficult path planning due to performance differences of nodes in the cloud environment;2)incomplete virtual multiple paths and non-aware physical multiple paths leading to low network throughput performance.To address the above issues,this thesis investigates two aspects of deploying multipath transmission in cloud environment.First,in the aspect of path planning,for the problem of difficult path planning due to performance differences of nodes in cloud environment,this thesis proposes a path planning mechanism for node performance load balancing,models and solves the path planning problem to achieve the minimization of network latency after deploying multipath transmission in cloud environment under the premise of satisfying node performance load balancing.Second,in the aspect of path deployment,for the problem of incomplete virtual multiple paths and non-aware physical multiple paths leading to low network throughput performance,a path deployment mechanism based on complete virtual multiple paths and network traffic sensing is proposed to improve the network throughput performance of multipath transmission in cloud environment with complete virtual multipath and trafficsensing physical multipath as the solution.Finally,based on the theoretical study,this thesis designs and implements a virtualized network resource optimization system based on multipath transmission in cloud environment and deploys it in the actual environment of cloud data center to verify the theoretical results of this thesis.In summary,this thesis provides a feasible and effective virtualized network resource optimization scheme for cloud data centers based on multipath transmission in cloud environments,in terms of path planning and path deployment.With the increasing popularity of cloud application ecology,the research results of this thesis will be applied to cloud data centers to provide high-performance and highly reliable virtualized network support for cloud applications,which has important theoretical and application values.