Video-QoE aware resource management at network core

We address the problem of video-aware multiuser resource management in modern wireless networks such as 3GPP Long Term Evolution (LTE) in the context of video delivery systems like HTTP Adaptive Streaming (HAS). HAS is a client driven video rate adaptation and delivery framework that is becoming popular due to its inherent advantages over existing video delivery solutions. Quality of Experience (QoE) is the prime performance criterion for adaptive video streaming and wireless resource management is a critical part in providing a target QoE for video delivery over wireless systems. However modifying resource management functionality to be video-aware at the network edge is rather difficult in practical networks. In this paper, we propose an alternative architecture to enhance video QoE in QoS (Quality of Service)-aware networks wherein the intelligence for video aware resource management resides at the network-core rather than at the network edge. In this new architecture, a "Video Aware Controller" (VAC) is placed at the network core. The VAC periodically receives HAS-related feedback from adaptive streaming clients/servers which it converts to QoS parameters for each user. Further, we propose an algorithm to dynamically compute the Maximum Bit Rate (MBR) for each streaming user based on media buffer feedback. Our simulation results on an LTE system level simulator demonstrate significant reduction in re-buffering percentage and enhanced QoE-outage capacity compared to existing schemes.