HTBT: A Hybrid DASH Adaptation Algorithm Using Takagi-Sugeno-Kang Fuzzy Model.

Video streaming takes the largest share of internet traffic today, and MPEG dynamic adaptive streaming over HTTP (DASH) has become dominant among other video streaming standards and protocols. According to the DASH standard, multimedia content is encoded in different quality levels with different bitrates located on the server, and users can request multimedia content of any available bitrate. The user side determines the desired bitrate in the unit called adaptation bitrate (ABR) logic. Many ABR algorithms have been proposed to improve the quality of experience (QoE). The main criteria for determining QoE are average bitrate, number of switches between resolutions, and number of buffer underflows. This paper presents a hybrid DASH adaptation algorithm that uses the following input values: current buffer occupancy level, network throughput value calculated on the last downloaded DASH segment, and Takagi-Sugeno-Kang model output that represents expected throughput in the next segment download iteration. We compared the proposed algorithm with several other algorithms and the results show that it outperforms others in average bitrate and number of bitrate switches. Furthermore, our algorithm prevented all buffer underflows. [ABSTRACT FROM AUTHOR]