Font Size: a A A

A Robust Controller Balanced Deployment Method In Software-Defined Networking

Posted on:2021-01-11Degree:MasterType:Thesis
Country:ChinaCandidate:X ZhangFull Text:PDF
GTID:2518306041961699Subject:Master of Engineering
Abstract/Summary:PDF Full Text Request
With the rapid development of Internet technology,the traditional network architecture can't meet the growing demand for reliability,scalability,flexibility and so on,so there is Software Defined Network(SDN)that is one of the new network architectures that can change the existing network dilemma.The core idea of SDN is to separate the control plane of the network from the data plane,so as to realize the flexible control of the network.Although SDN has many advantages that traditional networks can't match,SDN in the practical application of many problems,single controller deployment can't meet the needs of large-scale network,so the study of multi-controller deployment has become an urgent problem to be solved.Similar to traditional networks,SDN also suffer from links failure or nodes failure,in order to improve the network performance of SDN under links failure,it is of great significance to study the balanced deployment of robust controller in SDN.This thesis focuses on the issue of SDN's Robust Controller Placement(RCP)problem in the context of links failure.In view of the RCP problem,an Integer Linear Planning model(ILP)is established,and the goal of the ILP model is to minimize network latency and the cost of placing.In order to solve the RCP problem,we propose an effective heuristic algorithm,referred to as RCP heuristic algorithm.Experimental results show that the RCP heuristic algorithm can also improve the transmission effectiveness and controller load balancing of the SDN while satisfying the given control coverage.The main work of this thesis is as follows:(1)Put forward the relevant evaluation index.In the case of SDN suffering links failure case,the control coverage index is used to characterize the robustness of the network,the higher the index,the higher the robustness of the network;the network transmission efficiency index is used to characterize the controller transmission effectiveness,the higher the index,the higher the effectiveness of the controller's transmitted information;the controller's load Gini coefficient index is defined to characterize the load balance of the controller,the smaller the index,the more balanced the controller load.(2)Establish an RCP problem model.An integer linear programming model is established for the RCP problem.The goal of this model is to optimize the deployment of controller while considering network delay and load balancing,so that SDN can deploy the controller while meeting the given control coverage.The minimum number of controller(that is,the least cost),the minimum time delay and the highest efficiency for the controller to transmit information.(3)An effective RCP heuristic algorithm is proposed.In order to solve the RCP problem in the case of links failure,modified GN(Girvan-Newman)algorithm is used to simulate the links failure in SDN,and the minimum and optimal controller deployment nodes set is obtained through the Breadth-First Search algorithm and node effectiveness.The experimental results show that the RCP heuristic algorithm can meet the given control coverage while obtaining higher network transmission efficiency and lower controller load Gini coefficient,thereby improving the effectiveness of controller transmit information and load balancing of controller in SDN.In summary,in order to effectively improve the network performance of SDN under the links failure case,we propose the RCP problem and RCP heuristic algorithm to improve network robustness,control effectiveness,and achieve controller load balancing.Experimental results show that the RCP heuristic algorithm can effectively improve the network performance of SDN and reduce the complexity of solving RCP problem.
Keywords/Search Tags:Software defined network, Controller placement, Robustness, Load balancing, Network delay
PDF Full Text Request
Related items