Delivery of High Definition Video Content over Bandwidth Constrained Links in Heterogeneous Wireless Networks

In order to deal with the fluctuations in available bandwidth that mobile users experience, modern video streaming employs adaptive bitrate (ABR) algorithms. MPEGDASH enabled clients adapt the bitrate of requested video segments to match as closely as possible the available bandwidth of the wireless link in use. If link conditions are poor or the link is congested, the ABR algorithm will select the lowest available bitrate supported by the video stream. This approach works well for standard video content since there is no restriction on using low bitrates.However, for users wishing to view High Definition (HD) content, there is a lower bound to the bitrate below which they cannot go if the content is to be considered HD. ABR algorithms, forced to maintain HD bitrates over wireless links with constrained bandwidth, experience significant numbers of stalling events. High numbers of stalling events, particularly towards the end of a video, have a serious negative impact on end-user QoE.This paper proposes Adaptive Interface Selection (AIS) to address this issue. AIS leverages the multi-homed nature of modern mobile devices and the segmented nature of video content prepared for DASH-enabled clients to overcome the bandwidth constraints that hobble ABR algorithms. AIS uses multiple wireless interfaces in parallel to download content to maintain the playout buffer level to prevent stalling events. By minimizing the number of stalling events and streaming at HD compatible bitrates end-user QoE is maintained. Downloading as little content as possible over mobile networks helps protect the end-user's mobile data-cap.