Time-Scalable Low-Latency Video Bitrate Adaptation
Heng Jiang, Yu Liu, Yumei Wang · 2021
High-quality real-time video requires large bandwidth and low latency. These requirements bring huge challenges to network performance. In this paper, a dynamically balanced framework for the bandwidth and delay of the video is proposed. The frame rate of the video segment being played will adaptively change with the download time. For example, When the throughput is slightly lower than the selectable bitrate, a longer download time is required, then the playback frame rate will slow down, and vice versa. This method not only reduces the buffer occupancy and enhances the real-time performance, but equivalent to increases the optional bitrate of the video. We have implemented our libVLC player using the SDK provided by VLC. The simulation results show that our method improves significantly the downloaded bitrate at low, medium, and high throughput.