Research Of Multi-constraint OSPF Routing Method Based On Genetic Algorithm

With the rapid development in the broadband IP technology, there are more and more multimedia application in the Internet, such as video meeting. On the one hand, these new services require the support of multicast communication to reduce the traffic in the network. On the other hand, there must be a QoS (Quality of Service) guaranteed mechanism for these realtime services, they are different from traditional routing in existence, and solving them in classical SPF algorithms is very hard.This thesis focuses on the application of genetic algorithms in the routing search of OSPF network and emphasizes on how to improve performance in congestion in order to obtain the optimal global result. The genetic algorithms combined with the parameters of OSPF network, using the constraints to confirm the direction of the routing search of OSPF network, solve the difficulties in the choice problem for the routing of OSPF network. The purpose is to applying genetic algorithms for the routing search of OSPF network and to open up a new path for the routing of OSPF network.The important role of OSPF network in nowadays and hereafter information society is presented. The characters of the OSPF network and the routing, and the importance and the necessity are introduced. Then popular searching methods are described and the basic conceptions of genetic algorithms are elaborated, such as problem formulation, coding, crossover, mutation and so on. Under the leading of fitness value, it makes an efficient searching in complex spaces until acquiring the best result. The new methods of routing search of the OSPF network are suggested, which are based on genetic algorithms and take into account the requirement of QoS parameters in the routing search and the satisfaction of the real time. Combining with the genetic algorithms, a search strategy to diminish delay parameter is put forward. The results from emulation tests demonstrate the advancement and practicality of the new method in congestion.This thesis models and predicts network performance to improved routing strategy using network simulation tool OPNET, it designed a node system based on OSPF. Then this node system is simulated in improved routing strategy on Modeler platform.At last, the future development of the OSPF routing algorithm is pointed out. The future of it and its development are also discussed.