A Data-Driven Architecture for Personalized QoE Management in 5G Wireless Networks
Ying Wang,Peilong Li,Lei Jiao,Zhou Su,Nan Cheng,Xuemin (Sherman) Shen,Ping Zhang
DOI: https://doi.org/10.1109/mwc.2016.1500184wc
IF: 12.777
2017-01-01
IEEE Wireless Communications
Abstract:With the emergence of a variety of new wireless network types, business types, and QoS in a more autonomic, diverse, and interactive manner, it is envisioned that a new era of personalized services has arrived, which emphasizes users' requirements and service experiences. As a result, users' QoE will become one of the key features in 5G/future networks. In this article, we first review the state of the art of QoE research from several perspectives, including definition, influencing factors, assessment methods, QoE models, and control methods. Then a data-driven architecture for enhancing personalized QoE is proposed for 5G networks. Under this architecture, we specifically propose a two-step QoE modeling approach to capture the strength of the relationship between users and services. Thereafter, the preferences of a user is introduced to model the user's subjectivity toward a specific service. With the comprehension of users' preferences, radio resources can be distributed more precisely. Simulation results show that overall QoE can be enhanced by 20 percent, while 96 percent of users have an improved QoE, which validates the efficiency of the proposed architecture.