Research On Optimal Route Selection Algorithms For Load Balance Based On SDN

With the continuous rise and development of cloud computing,big data and e-commerce,and the gradual expansion of network scale,the traditional network has been unable to meet the growing needs and services of users.In this large network era,in order to redefine the Internet,in 2008,Professor Nick McKeown of Stanford University put forward the concept of OpenFlow.Then,according to the programmable feature of OpenFlow,SDN network came into being at the historic moment.In SDN network,the separation of control plane and data plane is its greatest feature.Because of the flexible programmable ability of controller and the ability of centralized control,SDN network has become a hot topic for many researchers.However,the research of load balance in SDN network has been in the exploratory stage.Today,with the development of SDN network,this new network architecture is in the process of increasing development.Load balancing is necessary.Through the study of SDN network system,the understanding of load balancing and the research of corresponding algorithms,many current SDN network load balancing implementations depend on traditional load balancing algorithms.The static and dynamic methods of traditional load balancing algorithms can solve the network problems to a certain extent,but in the case of high network complexity or large network flow.It will cause server burden to varying degrees,it is difficult to complete the tasks required by the network,and it will seriously waste network resources.In order to further study the load balancing problem in SDN network,ant colony algorithm is selected as the basic algorithm.Firstly,the ant colony algorithm is improved.The improved ant colony algorithm optimizes the pheromone updating strategy,and also introduces the search frequent region,so that the ant colony will get the best set of paths when searching the path.Secondly,in order to get the better solution,the ant colony algorithm is modified.The genetic algorithm is introduced into the ant colony algorithm,and a hybrid ant colony algorithm is proposed.This algorithm regards the best path set as the initial population of the genetic algorithm,and then obtains a better solution to replace the original solution through crossover and mutation operations.Finally,the global optimal path is obtained.The hybrid ant colony algorithm has stronger search ability and can effectively avoid calculation.Finally,a load balancing module is introduced in SDN network,in which a hybrid ant colony algorithm is used to achieve load balancing in SDN network,and at the same time,it can improve the speed and accuracy of the network.In the experimental simulation stage,the mainstream Floodlight controller and Mininet simulation platform are selected to carry out the experiment.After determining the reliability performance of the network platform,the proposed algorithm is simulated and tested.Theresults are mainly analyzed from several evaluation indicators,such as load balance,stream reception rate,bandwidth occupancy rate,delay and packet loss rate.Compared with the traditional ant colony algorithm,the hybrid ant colony algorithm proposed in this paper can greatly improve the search ability and global convergence of the algorithm,greatly reduce the burden on the server,improve the load balancing ability of the network,at the same time,it can also improve the resource utilization of the network,reduce the consumption of the network,and improve the quality of service of the network,to a certain extent.It also enhances the stability and rapidity of network traffic forwarding.