Abstract:
Transmission of high quality video over wireless networks is increasingly important as users consume more video on mobile devices. Improvement in display technologies wil...Show MoreMetadata
Abstract:
Transmission of high quality video over wireless networks is increasingly important as users consume more video on mobile devices. Improvement in display technologies will drive demand for higher quality experiences such as multiview video, at the expense of greater bandwidth requirements. The viewing experience can be disrupted while streaming video due to packet loss caused by network congestion and interference. Packet loss in a 3D-HEVC encoded stream can lead to long periods of rebuffering, causing discomfort. To counter this, we have designed a rate adaptation method that uses the packet buffer size as an indicator of network load. During congestion, it proactively drops packets belonging to layers with less impact. The data from the dropped layers is approximated through view synthesis at the user's device using an error concealment strategy to maintain a higher visual Quality of Experience (QoE). The transmission method is evaluated through simulation using the Vienna LTE-A simulator. The QoE is measured using the received video's PSNR and the frequency/duration of rebuffering. The experimental results demonstrate the performance gains for this approach over the benchmark method.
Date of Conference: 03-05 May 2017
Date Added to IEEE Xplore: 03 August 2017
ISBN Information: