一种基于视频分类的码率自适应算法

An Adaptive Bitrate Algorithm Based on Video Classification

流媒体的码率自适应算法依据网络状态动态调节视频块的码率,提升用户体验质量,但忽略了视频类型的差异对用户体验质量的影响,导致算法性能下降。提出区分视频类型特征的码率选择算法C-ABR。设计相应的用户体验质量效用函数,使用强化学习算法训练模型A3C,提升用户体验质量。实验结果说明,相对于典型的码率自适应算法Pensieve和MPC,C-ABR算法用户体验质量分别提升22.7%和50.4%。

The Adaptive Bitrate(ABR)algorithms for streaming media can dynamically adjust the bitrate of video blocks according to network status,and thus improve the user Quality of Experience(QoE).However,the existing algorithms usually ignore the impact of video types on the QoE,resulting in performance degradation.This paper proposes a bitrate selection algorithm,C-ABR,which adapts to different types of videos.Corresponding utilization functions of user experience quality are designed,and the A3 C model is trained by using the reinforcement learning algorithm to improve the QoE.The experimental results show that compared with the current typical bitrate self-adaption algorithms,including Pensieve and MPC,the C-ABR method improves the QoE by 22.7%and 50.4%,respectively.