Characterization of 3G Data-Plane Traffic and Application towards Centralized Control and Management for Software Defined Networking

With the wide deployment of 3G/4G cellular data networks there is a tremendous growth of mobile Internet access worldwide. We conduct a detailed measurement study about user behaviors, application usage and location patterns of users. We present a methodology that correlates different network attributes and user information together. For user behaviors we classify all users into four groups by time division and investigate the behaviors of different types of users. For example, we find that midnight users consume tremendous network bandwidth. We categorize applications into several groups and provide insights of each type of application usage about up/down flow user numbers and user request times. We also analyze the location information of users that reveal the relationship between network usages with user mobility. Our study successfully provides network operational insights with the details of network usage or the user remarkable behaviors that can help network planners and operators to plan and adaptively adjust network capacity assignment and flow or application management. Most importantly it will help service providers to reduce their operating cost and offer more flexible and variety of services to meet the needs of today´s mobile clients without the need of increasing capacity in congested areas. Finally, our analysis will enable much more efficient routing and resource allocation management applications in the recent endeavor for software defined networking that will put such responsibility into the logically centralized controllers.