Bitrate selection for MPEG-DASH considering QoE by queueing theory

In recent years, video streaming data have increased. Therefore, MPEG-DASH (Moving Picture Experts Group-Dynamic Adaptive Streaming over HTTP) [1], which is one of the protocols that adaptively changes the video bitrate according to the user's network throughput in video streaming services, was developed. In the research on DASH, it is noted that the loss of the segment, which is a video playing unit, affects the video quality of the multicast environment [2]. It has also been noted that the occurrence of buffer underruns, which means that video playback stops, leads to user dissatisfaction [3]. In this paper, we propose an evaluation function for QoE (quality of experience) considering segment loss and buffer underrun probability. Moreover, we derive an optimal video bitrate to maximize the QoE value. The characteristics are analyzed by the simulation using the M/D/1/K model of queueing theory.