基于最佳分段点估计的流媒体非均匀分段方法

Non-Uniform Segmentation Method for Streaming Media Based on Best Segmentation Point Estimation

针对命中率随存储的流媒体片段流行度变化的特征,提出了一种新的基于最佳分段点估计的流媒体非均匀分段方法,根据不同存储大小下的流媒体外部分界流行度对其内部最佳分段点进行估计,进而把每个视频分成高流行度段和低流行度段两个片段。实验结果表明,与均分分段相比,该方法能减少流媒体的片段数,提高缓存命中率。

According to the relationship between the byte bit rates and the popularities of the stored streaming media segments, a new non-uniform segmentation method for the streaming media based on best segmentation point estimation was proposed. In the proposed method, the best internal segmentation points of the streaming media were estimated according to the external boundary popularity of the streaming media at the current storage size of the proxy. Then each video was divided into high popularity and low popularity parts using the estimated best internal segmentation point. Experimental results show that the proposed method can effectively reduce the number of segments and improve byte hit rates comparing with the uniform segmentation method.