Efficient Short-Form Video Streaming: An Integration of Dynamic Bitrate Adaptation and Predictive Segment Preloading

Nowadays, short-form videos have become popular, particularly among mobile users. However, unlike traditional videos, users can flexibly engage in brief viewing sessions and effortlessly skip content that fails to captivate their interest. This user behavior causes significant data wastage during short-form video streaming. From another perspective, fluctuations in network conditions during streaming pose additional challenges in providing a good preloading strategy to meet users' quality of experience (QoE). In this paper, we develop a new solution to improve the quality of user experience (QoE). Our scheme adapts to fluctuating network conditions, varying available bitrates, and user behavior patterns to optimize preloading strategies. Furthermore, our solution automatically adjusts the video downloading speed to ensure the appropriate amount of data downloaded for each video. Experiments show that our method improves QoE by ranging from 10% to 25% compared to reference methods.