边缘计算中面向多码率视频的缓存、转码和传输联合优化策略

Joint optimization strategy for caching,transcoding and transmission of multi-bitrate video in edge computing

结合边缘缓存技术与流媒体传输技术能有效提升视频服务质量,为降低视频内容提供商的边缘资源租赁成本,提出一种视频缓存、转码和传输联合优化策略。首先,综合考虑视频的缓存、转码、边缘传输和云端传输的成本,以最小总租赁成本为目标建立整数规划模型,并证明其NP-complete性质;其次,根据历史请求数估计视频流行度变化,并对流行视频进行缓存;最后,基于视频的缓存状态,为用户的请求选择成本最低的响应方式。仿真实验表明,所提策略与现有策略相比,可提升请求命中率并有效降低内容提供商的资源租赁成本。

Combining edge caching and adaptive bit rate streaming transmission can effectively improve the quality of video services.In order to reduce the cost of leasing edge resources for video content providers,this paper proposed a joint optimization strategy for video caching,transcoding and transmission.Firstly,considering the cost of video caching,transcoding,edge transmission and cloud transmission,the integer programming model aiming at the smallest total rental cost was formalized,which was proved to be NP-complete.Secondly,the strategy estimated the change in video popularity based on the number of historical requests,and cached popular videos.Finally,based on the video caching status,the strategy chose the lowest cost approach in response to user requests.Simulation results show that compared with the existing strategies,the proposed strategy improves the request hit rate and effectively reduces the leasing cost of resources.