Modelling and Analysis of Regional Service Behavior Properties of Mobile Internet Applications
In view of the facts that traditional traffic models are less effective in describing the new mobile internet applications and that the mobile user service behavior among regions with different levels of economic development are always different, a new modeling methodology of user service behavior specifically for mobile internet applications is proposed in this paper. Based on the crowdsourcing measurement data of the service behavior collected through massive end users in live mobile networks of China, the properties of three typical mobile internet services in three typical geographical regions are analyzed and compared with each other, by utilizing the proposed modeling methodology. And the relationships between some key properties of service behavior are analyzed with the maximal information coefficient method and the Pearson correlation coefficient method. The key parameters, differences, and internal relations among these services in typical geological regions of China are then identified. It provides a solid foundation for wireless product design, network planning, and optimization of mobile Internet services-oriented networks specifically in China, and is also a valuable reference for other countries of the similar situation.
Traffic model, mobile Internet application, service behavior, crowdsourcing.