Towards simple and smooth rate adaption for VBR video in DASH
Yanping Zhou, Yizhou Duan, Jun Sun, Zongming Guo
Abstract
Yanping Zhou, Yizhou Duan, Jun Sun, Zongming Guo
Abstract
Rate adaption in Dynamic Adaptive Streaming over HTTP (DASH) is widely applied to adapt the transmission rate to varying network capacity. For rate adaption on variable bitrate (VBR) encoded video, it is still a challenge to properly identify and address the dynamics of bandwidth and segment bitrate. In this paper, the trend of client buffer level variation (TBLV) is analyzed to be a more effective metric for detecting the dynamics of bandwidth and segment bitrate compared to previous metrics. Then, a partial-linear trend prediction model is developed to accurately estimate TBLV. Finally, based on the prediction model, a novel simple rate adaption algorithm is designed to achieve efficient and smooth video quality level adjustment. Experimental results show that while maintaining similar average video quality, the proposed algorithm achieves up to 47.3% improvement in rate adaption smoothness compared to the existing work.
OpenAlex reports 19 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Rate adaption in Dynamic Adaptive Streaming over HTTP (DASH) is widely applied to adapt the transmission rate to varying network capacity. For rate adaption on variable bitrate (VBR) encoded video, it is still a challenge to properly identify and address the dynamics of bandwidth and segment bitrate. In this paper, the trend of client buffer level variation (TBLV) is analyzed to be a more effective metric for detecting the dynamics of bandwidth and segment bitrate compared to previous metrics. Then, a partial-linear trend prediction model is developed to accurately estimate TBLV. Finally, based on the prediction model, a novel simple rate adaption algorithm is designed to achieve efficient and smooth video quality level adjustment. Experimental results show that while maintaining similar average video quality, the proposed algorithm achieves up to 47.3% improvement in rate adaption smoothness compared to the existing work.
Key concepts: Variable bitrate, Dash, Dynamic Adaptive Streaming over HTTP, Computer science, Constant bitrate, Bandwidth (computing), Real-time computing, Smoothness