Unveiling Radio Resource Utilization Dynamics of Mobile Traffic through Unsupervised Learning

Understanding mobile traffic dynamics is a keyissue to properly manage the radio resources in next gener-ation mobile networks and meet the stringent requirementsof emerging heterogeneous services, such as enhanced mobilebroadband, autonomous driving, and extended reality (just toname a few). However, radio resource utilization patterns of realmobile applications are mostly unknown. This paper aims atfilling this gap by tailoring an unsupervised learning methodology(i.e. K-means), able to identify similar radio resource utilizationpatterns of mobile traffic from an operating mobile network.Our analysis is based on datasets referring to residential andcampus areas and containing wireless link level information(e.g., scheduling, channel conditions, transmission settings, andduration) with a very precise level of granularity (e.g., 1 ms).Obtained results reveal the properties of groups of sessions withsimilar characteristics, expressed in terms of bandwidth demandsand application level requirements