Unveiling Radio Resource Utilization Dynamics of Mobile Traffic through Unsupervised Learning

Abstract

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

In-depth study of RNTI management in mobile networks: Allocation strategies and implications on data trace analysis

“…When T L instead of T R is used, non-existent users are created, that have a smaller load value than the T R extracted ones. These errors can have substantial impact [0-1] min,TR [2][3][4] min,TR [5][6][7][8] min,TR [9][10][11][12][13][14][15][16] min,TR [17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32] min,TR min,TR [0-1] min,TL [2][3][4] min,TL [5][6][7][8] min,TL [9][10][11][12][13][14][15][16] min,TL [17][18][19][20]…”

“…At the same time, the incorrect identification of user sessions has clear consequences on traffic analysis: counting multiple users as one, or vice versa, results in incorrect estimates of traffic distribution, user activity, and resource allocation, which affects modeling and simulations and potentially biases the conclusions. Traffic traces containing RNTI information have been proven useful to identify radio resource utilization of mobile traffic patterns [19,20], to fingerprint applications [21,22] and possibly reveal the user identity [23], to proactively identify user behavior for resource optimization [24] and to perform classification of Downlink Control Information (DCI) messages [25]. While previous studies on mobile data traffic were only able to differentiate traffic profiles according to time usage patterns [26,27], the use of BS traffic traces allows more sophisticated studies, for example to identify application classes [19] and standalone applications [21,22].…”

Metodologia Automatizada para Descrever a Cobertura de Sinal Móvel em Rodovias: Estudo de Caso Considerando Incerteza de Dados

“…Experimentos foram realizados através da leitura da quantidade de sinal móvel em um misto de áreas urbanas e rurais da Dinamarca. Rago et al [3] buscaram identificar padrões de utilizac ¸ão de trafego móvel através aprendizado não supervisionado. A análise apresentada é baseada em conjuntos de dados com informac ¸ões do nível de sinal disponível.…”

“…Para cada registro armazenado no banco de dados, os dados de estradas ausentes são inseridos usando dados de latitude e longitude. Dados rodoviários são obtidos de acordo com as informac ¸ões e classificac ¸ão disponíveis via APIs de Overpass 2 e OpenStre-etMap 3 , e incluem:…”