Modeling QoE for streamed H.265/HEVC content under adverse network conditions

Abstract

Networked visual applications such video streaming have grown exponentially in recent years, yet are known to be sensitive to network impairments. Increasingly the focus of both content and service providers is turning from Quality of Service (QoS) awareness to customer experience awareness. The challenge of managing users' Quality of Experience (QoE) is particularly acute when the content is delivered over loss-prone network links. The newly standardized High Efficiency Video Coding (H.265/HEVC) standard offers bandwidth savings of up to 50% over the current Advanced Video Coding (H.264/AVC) standard without loss in video quality. This bandwidth saving makes H.265/HEVC an attractive compression method for the delivery of video content over wireless networks, which are often resource-constrained. The impact of network impairments such as packet loss on highly compressed HEVC encoded content is an area that has not yet been sufficiently investigated. This work reports on a user-centric study of the effects of network impairments on HEVC video streams in IP networks. HEVC encoded video streams were delivered over a loss-prone network and the resulting impaired videos were subjectively evaluated for perceptual quality. A comprehensive statistical analysis of the results of these subjective tests and concurrent objective testing using traditional video quality metrics has identified the primary factors that affect the user's perception of quality and therefore their QoE. The relative importance of each QoE factor for HEVC streaming in lossy network conditions is identified and user tolerance levels for packet loss in HEVC streaming established. This timely study addresses the important issue of QoE in HEVC streaming and provides valuable insights that can be built upon in the design of user-centric streaming applications and services based on the H.265/HEVC standard.