Dynamic and adaptive QoE management for OTT application sessions in LTE

Radio resources in Long Term Evolution (LTE) systems are distributed among User Equipments (UE) without considering the application and content being transmitted. When the radio capacity is fully utilized, the bearer based service strategy aiming to maximize the spectral efficiency may not support good customer experience for popular Over The Top (OTT) applications where each session has an individual and highly dynamic bandwidth demand. In this paper, we argue that maximizing the system efficiency cannot be decoupled from customer experience, therefore maximized spectral efficiency does not necessarily improve the service quality perception of end users. Furthermore, we show that the existing LTE QoS architecture does not bring sufficient benefit on this matter and conclude that customer experience is a better system level target. Such operation requires a different service paradigm where resources are allocated to individual application sessions adaptively and dynamically by understanding their specific requirements. We propose a Quality of Experience (QoE) Manager that can deliver superior QoE for each session by loosening the radio efficiency targets to maximize the customer experience. Still, the radio scheduler retains sufficient freedom to optimize on channel diversity. The solution is demonstrated via simulations including the successful management of encrypted YouTube downloads.