Neural Network Model of QoE for Estimation Video Streaming over 5G network

Asma Lounis (1) , Farid Alilat (1) , Nazim Agoulmine (2)

1 Image Processing and Radiation Laboratory
2 IBISC - Informatique, Biologie Intégrative et Systèmes Complexes

With the rapid increasing demand of commercial video streaming, the satisfaction of the end user is becoming more and more important to measure and assure.The quality of experience (QoE) is defined as the measure of the overall level of customer satisfaction with the usage of a service provided by a vendor. Many works have addressed this issue in many different scenarios in cellular networks however most of these works have addressed video streaming over LTE networks (Long Term Evolution network). Up to day, there are few contributions of work that address the QoE over 5G network since there still still some challenges in this later to address. In this paper, we present the specific aspects we consider important in the evolution from 4G to 5G in term of traffic management and a solution to estimate this QoE in this new context. We adopted an approach based on Neural Network (NN) to estimate the QoE parameters. NN have been successfully used in many domains where it was difficult to derive an exact analytical model of the system so is the case of the 5G network.

Domaines: Réseaux et télécommunications [cs.NI]

Asma Lounis, Farid Alilat, Nazim Agoulmine. Neural Network Model of QoE for Estimation Video Streaming over 5G network. 6th International Workshop on ADVANCEs in ICT Infrastructures and Services (ADVANCE 2018), Jan 2018, Santiago, Chile. pp.21--27. ⟨hal-01777376⟩