Font Size: a A A

Research On Multi-constrained Routing Algorithm For Software Defined Network Based On Nonlinear Annealing

Posted on:2016-10-25Degree:MasterType:Thesis
Country:ChinaCandidate:Z K SongFull Text:PDF
GTID:2348330488473347Subject:Computer system architecture
Abstract/Summary:PDF Full Text Request
With the rapid development of the network, the increasing size of the expansion of existing networks, network management, monitoring becomes more and more difficult, the traditional network architecture has been unable to meet the current needs of the network management. In recent years, research on next-generation network architecture in full swing, made a lot of new network architecture, to solve the problem of network management, SDN(Software Defined Network) as a representative of the next-generation network model, has become the future focus of network Architecture Research.Different from the existing distributed network management, SDN separates the forwarding logic and control logic of traditional network equipment and instead uses a central controller to control the network. The centralized management approach of SDN solves the problem of functions complex, management inconvenient, help to strengthen the network management, expand the network performance, creates a programmable network, and indicates the direction of the next-generation network architecture research.The existing networks mainly provide "best-effort" service, but with the massive popularity of the network, network applications become more diverse, and the network needs to provide different standard quality of service for different network applications. The concept of QoS(Quality of Service) become more and more popular, studies on the quality of service control are also deepening.QoS is designed to provide different standard of services for applications according to the different characteristics of applications. In the related study on QoS, QoS routing is one of the focus, whose main achievement is to find a data packet transmission path for the network applications to meet the constraint parameters. As a development trend of the next-generation network architecture, the research of QoS routing problem on SDN means more. Based on the architectural features of SDN, this paper proposed a heuristic QoS routing algorithm based on nonlinear coupling function, reduce feasible solution space and the thought of annealing, and compared the performance of this algorithm.First, this paper introduces the development process and architectural features of SDN, points out the advantages of SDN. Next, highlight the relevant knowledge of QoS routing, detailed analysis of the advantages and disadvantages of some existing typical QoS routing algorithm.Then, based on the characteristics of SDN and the disadvantages of high time complexity, easy to local optima, poor portability, bad scalability for the existing QoS routing algorithms, this paper proposes a new heuristic QoS routing algorithm on SDN. Firstly, algorithm couplings multi-parameters as one parameter according to the linear coupling functions, does twice labeled according to the parameter use Dijkstra like algorithm, and simplifies the network size based on the labeled value to reduce feasible solution space; Then this algorithm use the nonlinear coupling function to coupling multi-parameters, do Dijkstra like algorithm, while introducing the idea of annealing, the probability of the choice of site, to avoid falling into local optima; At the same time by the parameter of iterations defined this paper can weights between the time complexity of the algorithm and algorithm performance.Finally, by using the network simulation software to build a SDN model, this paper analysis and test the performance of this algorithm based on this model. Compared to some other typical QoS routing algorithms, the algorithm we proposed showes a higher success rate and reasonable running time, and has good scalability for large-scale network.
Keywords/Search Tags:Software Defined Network, Quality of Service, simulated annealing, coupling function, QoS routing
PDF Full Text Request
Related items