User Mobility Behavoran Alysis Based On3G Network

Since the3G network begin in China at2009,it develop so rapidly that China has become the world’s largest mobile communications consumer country. With the popular of the concept called SoLoMo which means social, locate and mobility. One of the biggest business needs of network enterprise today is achieving interactive information between businesses and users to interact and conduct in-depth interactions based on mobile Internet network. Most of the Start-up companies have started to focus on this field right now. The information of mobile users play a key role in range of application on mobile Internet network such as personalized advertising recommendation, personal content search and traffic forecasting. The research of the mobile behavior based on3G network traffic in China is just on the way.This paper models massive3G traffic in China and tries to model it for further analysis. Firstly the application layer protocol has been parsed using the deep packet inspection and separated the signaling and data in order to associate data with the device type. After analysis the mobile behavior based on the modeled data. This paper parsed base station data which contains more than60000users. Clustering the related base stations and create the path of every mobility users for related patterns mining and use the parallelism frequent sequence mining algorithm FUFP-tree for predicting the next sign location of the user corresponding the application layer information.The following are the main research in this paper:Firstly modeling and parsing massive3G network data based on a distributed platform. Analysis user behaviors from the point of view of the data plane and signaling surfaceThen analysis different terminal behavior of mobile users after matching terminal with the data.At last clustering the base and create the path of every user and get the frequent items using a parallel FP-Growth mining. To improve the prediction accuracy combined with application information.