On stabilizing buffer dynamics for adaptive video streaming with a small buffering delay

Abstract:

Adaptive video streaming is today's predominating video delivery model in the Internet. However, existing bitrate adaptation achieves satisfactory performance only under buffering of several tens of seconds. The major challenge is the stabilization of client buffer dynamics, because a small buffer lacks resilience to the buffer fluctuation imposed by the mismatch between the network throughput and the average bitrate of the selected video chunk. This study tackles this fundamental challenge by developing a novel adaptation algorithm that minimizes buffer deviation from the desired level despite a small client buffer and jointly reduces the rate of bitrate switching. Our evaluation methodology combines a bitrate-adaptive streaming simulation with a trace-driven simulation of 3G mobile network in the context of video streaming. Simulation results show that our algorithm effectively supports low-latency adaptive video streaming with a buffering delay as small as a chunk-duration.