Mobility management: Deployment and adaptability aspects through mobile data traffic analysis

Abstract

The expected boost in mobile data traffic and the evolution towards the next generation of networks are making cellular operators reconsider whether current approaches for handling mobility could be im proved, according to the characteristics of the mobile traffic that actually flows through real networks. In this work, we make use of extensive analysis of real network traces to infer the main characteristics of mobile data traffic for a particular operator. Our analysis focuses on the features related to mobility, i.e., location information, number of handovers, or duration of the data traffic exchange. New techniques to gather the mobility characteristics of the user based on data and control packets correlation are designed and applied to compare the gains of deploying different mobility management approaches. We show that adapting the mobility management mechanism to the degree of mobility and the network characteristics brings some benefits to the network operator over the current approach, especially in scenarios of low mobility, where a dstributed mobility management solution proves to be more efficient.

Distribution and Allocation of Network Resources Based on Predictive Analyses of Time-Series Telecommunications Data

“…In processing time-series data, one of the most commonly used methods includes data classification. There are many examples of usage of data classification processes, some of which are applied, for instance, in adapting the mobility management mechanisms [37], prediction of applications' data consumption [38], and user activity [39].…”