Characterization of Traffic for Live Video Streaming Services on DASH in 4G Networks Based on Syntactic Analyzers

Abstract Context: Mobile data traffic generated by video services increases daily. To address this situation, telecommunication service providers must know the behavior of video traffic and thus adjust network resources to meet and maintain the quality levels required by users. Traffic characterization studies in 4G networks for Live Video Streaming (LVS) services are scarce, and those available are obtained from simulation scenarios in which the real operating conditions of these types of networks are not considered. Method: This work focuses on finding a model that characterizes traffic from the probability density functions of LVS services under the adaptive streaming DASH technique in LTE networks. The traces analyzed to carry out the modeling study were acquired in real emulation scenarios considering the operating conditions frequently presented in the actual provision of the service, for which five test scenarios were defined. Results: Based on the parameterization of a number of probability density functions found, a description of different traffic models of the service under study is presented, as well as for each of the pre-established test scenarios in a 4G-LTE network. Conclusions: From the results, it is concluded that the traffic model depends on the conditions of each scenario, and that there is no single model that describes the general behavior of LVS services under the adaptive streaming DASH technique in an emulated LTE network.