Data Processing And User Behavior Analysis In Mobile Communication Network

At present, the data analysis technology for mobile communication network data is mostly based on the voice data collected by operators rather than IP data. With the development of the IP-based process of mobile communication network, it is necessary to analysis the IP data in mobile communication network. Although there are many valuable researches on IP data, the corresponding research content is focus on Internet. With the continuous progress of the IP-based of mobile communication network, it is necessary to research the whole scheme of IP data processing and user behavior analysis in the mobile communication IP network.In this paper, the research on IP data processing and user behavior analysis in mobile communication network starts from the IP data of mobile communication network. The research and implementation of big data processing technology is taken as the basis and means. The paper studies the User Behavior Analysis in Mobile Communication Networks from two dimensional of complex network and data mining.The main contributions and innovation points are listed below.The paper proposes and implements the model of distributed preprocessing. The paper codes the data import module based on Hadoop programming which can be used to parse the original PCAP data and send the split IP protocol packet to Hadoop's Mapper function for distributed processing. And programs the data selection and processing module based on Map-Reduce which module uses the distributed idea to extract, clean, convert and contract the fields in the incoming IP protocol packet. The paper provides a unified data structure and format for the establishment of network and the analysis of data.The paper analysis the IP network model based on the complex network. The paper establishes a network model with IP as node and IP connection as edge for IP data of mobile communication network. Adopts the research method of complex network analysis to study the IP network model and the topological properties of the network. Extracts and marks the key nodes of the network. The paper also optimizes the principle of label propagation algorithm, and use the algorithm to group the network model dataset.The paper analysis user behaviors of the IP network model based on data mining. According to the characteristics of mobile IP network model,the paper decomposes the result dimension of clustering algorithm. By weakening the internal relations of the same behavior with different ports,the paper realizes association analysis of mobile communication IP data based on different behaviors of different port.