Video Quality Adaptation to Improve The Quality of Experience in DASH Environments

Department of Communication Engineering, Kwangwoon University, Korea Summary Recently, DASH (Dynamic Adaptive Streaming over HTTP) is gaining attention because it is possible to use an existing web server, and not be restricted by the firewall or NAT (Network Address Translator). However, the existing video quality adaptation methods for DASH do not consider the frequent change of video quality, which degrades the QoE (Quality of Experience) of DASH service. In this paper, we propose an SDASH (Smooth Dynamic Adaptive Streaming over DASH) scheme to improve the QoE of DASH service by avoiding frequent changes in the video quality. The proposed S-DASH scheme consists of a video quality adaptation algorithm, and a segment duration determination algorithm. The video quality adaptation algorithm controls the video quality based on MSD (Media Segment Duration), SFT (Segment Fetch Time), and the buffer state of the client to reduce the change of video quality and to prevent buffer underflow of the client. The segment duration determination algorithm determines the segment duration based on variation of chunk throughput to reduce the change of video quality and to improve the video quality. As a result, the proposed scheme can improve the QoE of DASH service. Through the simulation results, we prove that the proposed S-DASH scheme improves the QoE of DASH service by utilizing the smooth video quality adaptation algorithm and the segment duration determination algorithm.