Research On Role Identification Method For SDN World Communication Network

Based on the rapid development of SDN's(Software Defined Network)world communication network,and gradually expand its application and practice in the civil field,it is necessary to face the problem of surge in network users and increase the amount of data.However,there is a time domain for SDN-based world communication network,Spatial domain and spectral domain.At present,the engineering realization of the network user role recognition algorithm is far less than that of the land mobile communication network,the Internet,the Internet of things,the television network and other fields.Therefore,it is necessary to study and optimize the existing user identification of the world telecommunication network.The role of the user of the world communication network user first to a large number of users are not obvious characteristics of the cluster,and then by analyzing the user relationship to identify the user role.Aiming at the problem that the current clustering method mainly through sampling clustering,the sample size is not enough to reflect a large number of all user characteristics,resulting in a larger error.In the CUER algorithm,the interval distance is introduced,and the Map / Reduce mechanism is proposed Based on the interval clustering method,all users are divided into multiple subclasses,and then the subclasses are clustered(from the similar behavior point of view),and the results of the initial clustering are clustered together.It can avoid sampling clustering lead to a large error in the results,but also improve the real-time clustering.Aiming at the problem that the current user identification algorithm is a large number of users,it can not tell all the key users,and the recognition accuracy is low.The user identification method based on the relationship between heaven and earth communication network is proposed.First,the HBACA algorithm(mixed ant algorithm)is used to find or identify different elements of the relationship between the strong ability to extract the network events in the user relationship,mainly by searching for the behavior of the ant mixed with the behavior of reconnaissance ants,that is,the search ants in the search process,will find the investigation of ants have been marked pheromone and then identify the network user's relationship;then,in the network user relationship to calculate the center of each user node center degree,compact center and node active according to the important or critical degree of network communication users in the world,the users are divided into key users,important users and general users,and the criteria for judging the roles of users of the world communication network are put forward.Based on the above research,the simulation system of the world communication network is designed and the simulation experiment is carried out.The experimental results show the effectiveness and feasibility of the improved algorithm.The research results of this thesis can further enhance the accuracy and efficiency of user identification and recognition in the world network communication network,and lay the foundation for the further research on the following network control and data mining.