| With the continuous development of the Internet, people demand for network performance is getting higher and higher. Now the TCP/IP network is faced with complex network management, closed network equipment and other issues. Software defined network (SoftWare-defined Netwrork,SDN) is a new network innovation framework, can effectively solve the current problems which the current network is difficult to solve.As the SDN network scales up, a single controller may encounter a single point of failure and scalability problems. The distributed control system is an effective solution to solve the problem of scalability. Distributed control system contains multiple controllers,so the deployment of multiple controllers is also challenge to address.In the multi-controller deployment problem, the switch relies on the controller to perform actions, so the link stability between the switch and the controller is a very important factor. Therefore, the goal of this paper is to find a controller deployment plan to ensure the reliability of the communication between the switch and the controller, while minimizing the communication delay between the switch and the controller.At the same time,when we deploy multi-controller in the small and medium-sized network, in addition to propagation delay, the controller response time is also a very important factor affecting QoS. Therefore, this paper studies how to use reasonable network resources to meet QoS requirements in practical network deployment.In this paper, the above two aspects were entered into the in-depth study,the main contributions are as follows:(1) Based on the research of SDN distributed controller and multi-controller deployment algorithm, the influence of forwarding plane fault on network performance is studied, especially the reliability between switch and controller. And then we proposed the cycle k-medoids algorithm.(2) Simulation results show that the cycle k-medoids algorithm proposed in this paper can improve the reliability of the communication between the switch and the controller, while maintaining a relatively low delay.(3) The quality of service (QoS) problem in multi-controller deployment is studied. In this paper, the process of the flow in the SDN is modeled and the response time of the controller to the flow and its influencing factors are analyzed. Considering the load and propagation delay of the controller, a heuristic algorithm to guarantee QoS is proposed.(4) In this paper, we do simulation with the real topology. Results shows that improving the controller capacity and delay limit can reduce the number of controllers in a certain range. This is a practical guide to the deployment of the network. |
