Font Size: a A A

The Controller Placement Problem And Cloud Storage Assignment Problem In SDN

Posted on:2017-03-22Degree:MasterType:Thesis
Country:ChinaCandidate:S LiuFull Text:PDF
GTID:2308330485482222Subject:Software engineering
Abstract/Summary:PDF Full Text Request
With the growing demand for network services, network provide more and more types of services to users. The data traffic is increasing quickly. With computation becoming more and more complex, the traditional network architecture and service model has been unable to deal with these problems. Software defined network (SDN) and cloud computing provide an effective way to solve the above problems. More and more cloud computing services are built on the SDN network structure. However, the structure of SDN and SDN based cloud services are still have a lot of problems to be studied and resolved.First, SDN is the logic centralized network structure. The controller is called as the "brain" of SDN. The controller install the forwarding rule to switch through the standard protocol. The delay between the control plane and the data plane directly affect deployment of forwarding rule into switch. With the tremendous increase of traffic, the controller’s capacity is constrained. It is difficult to timely install the forwarding rule on the switch. The delay between the controller and controller directly affect the timely communication between controllers. The delay between controllers affects the network logic consistency. How to reasonably deploy SDN controller’s position, so as to ensure control plane and data plane delay, guarantee delay between controllers, guarantee the, has been became a hot issue in the current research.Second, cloud storage is a new concept. Cloud storage is the extension and development of cloud computing. In the network environment the speed and performance of cloud storage service largely depends on the location of storage device. In this paper, we proposed the SDN dynamic cloud storage problem based on user’s demand. The purpose is dynamically adjust the data in the cloud storage server according to the needs of the user to meet performance requirements for cloud storage.This paper first introduces the concept of SDN and cloud. The paper analyses the problems in this new network structure and service mode. Then, we solve three problems: How to deploy SDN controllers to reduce the delay of overall network; How to deploy SDN controller to get the minimum load controller placement; How to dynamic store data on cloud storage server according to the user demand change. According to the above three problems, we proposed the PSO-GDM algorithm, the NCPSO algorithm and the PSO-NFM algorithm. In this paper, we give the three algorithms detailed simulation results. The experiments show that the solutions can effectively enhance the performance of the network. And the solutions can optimize the network quality of service. For the large scale deployment of SDN and application of cloud storage based on SDN provides a good theoretical and practical support.
Keywords/Search Tags:Software Defined Network, Controller Placement, Cloud Storage, Particle Swarm Optimization
PDF Full Text Request
Related items