The Research And Application About Telecommunication Network Relationship Based On Cloud Computing

With more communication happening on the mobile network, telecom operators have accumulated much user information and behavior data which have large value. How to analyze these data is becoming a very interesting research topic for assisting operation and providing better service to users. In this article, both individual and group perspectives are used to study.From the individual perspective for assisting operation, new customers who have no communication with others through this network will be predicted to different levels based on their base information. Experience is not necessary for label customers because of introduce of facile, and customers are classed in relative way, so the customer values is relative. From the individual perspective for providing services for customers, customers who have communication with others through this network will be recommended new friends based on their base information and behavior data. In this article, cloud computing is used to calculate two-degree familiarity and friends features are used to get preferences of customers.In addition to analysis it from individual perspective, community structure and community evolution are also studied as group perspective. From the group perspective, some customers are studied as a group. The group which has more communication inside it is called community.For the network structure of group perspective, a community structure description algorithm based on statistical normalization is introduced and the result can be stored in a storage structure, so it can be used many times but calculated just once. In the empirical research part, outer calls distribution, external calls distribution, and total calls distribution are analyzed to given a different perspective for telecom community study. From the group point of view, in addition to study it in such a static way, the dynamic nature on the timeline is also studied called community evolution. For the community evolution based on cloud computing, parallel algorithm is implementation based on the data structure of evolution to tracing the evolution path.