移动互联网是包含用户、网站、移动基站的超大规模异构复杂网络，用户是移动互联网的主体，移动互联网中信息的传播、网络优化、网络规划、多网协同均与用户行为密切相关，因此有必要研究移动互联网用户行为特性及用户行为动力学模型，发现移动互联网演化的内在动因和规律，以实现用户行为预测、优化移动通信网性能、多网协同分流、提高通信资源利用率。本项目基于多个省域的移动核心网实测数据，采用统计与数据挖掘的方法研究移动互联网用户移动、访问行为时空特性。同时根据用户行为特性，将用户行为抽象为粒子运动采用统计力学的方法建立用户行为动力学模型并求解，实现对用户移动和访问行为的预测。本项目的研究成果将从用户行为角度为移动互联网的优化、管理、维护提供新的基础模型，同时为探索异构复杂网络演化机制提供新方法，摸索从实测数据到模型，再到应用的研究途径，并且将为舆情监控、社会感知计算、下一代移动通信网络规划优化提供理论基础。

The mobile Internet is an ultra-large-scale heterogeneous complex networks which contain users, web sites, mobile base stations. The user is the main body of the mobile Internet, user behavior is closely related to network planning, network optimization, the evolution of network structure. It is necessary characteristics of user behavior and to establish the dynamic model, describing the intrinsic motivation of user behavior and the evolution of mobile Internet. The dynamic model can be used to predict user behavior, optimize the network performance of mobile communication, improve the utilization of communication resources, and found that public opinion and information dissemination mechanism. Based on measured data from a large number of mobile core network, this project adopts statistical mechanics and data mining to analysis users motion and access behavior characteristics from the motion of particles, establish user behavior dynamics model. This project also uses user behavior dynamics model to predict the user behavior, the result is used in improve the resource utilization of mobile communication and provide the theory basis of mobile internet management and optimization. Research results of this project will explore the evolution mechanism of heterogeneous complex networks and provide a new method to explore the application of complex networks from the measured data to the model, and then way, and will provide some theoretical basis for monitor public opinion, the next-generation mobile communications network optimiztion.