Joint Optimization Strategy Of Communication And Caching Resources In Real-time Immersive Videos Transmission System Based On Edge-cloud Collaboration

With the maturity of Virtual Reality(VR)technology,immersive videos have become more popular.In order to ensure the viewing quality of users,the transmission of immersive videos needs to consume a lot of bandwidth resources to provide high-definition and lowdelay content.However,the bandwidth resources are limited,which is difficult to guarantee the time constraint.At the same time,most VR terminal video sources come from the cloud,and the acquisition of video content from the cloud further increases the transmission cost.In view of the limited bandwidth resources and the transmission cost of video from the cloud,this paper designs a utility-driven joint bitrate allocation and caching scheme to improve the system utility of the transmission of the immersive video based on the Mobile Edge Computing(MEC)technology.The main research contents and contributions of this paper are as follows:1)A new conception,unfreshness indicator,is proposed to present the outdated degree of tiles that can impair video quality.Depending on the tile’s unfreshness,the Quality of Immersive Videos(Qo I)model is constructed in this paper,which presents the video quality perceived by users.The quality of an immersive video is influenced by not only the bitrate version and the unfreshness of tiles,but also the content feature,temporal and spatial quality distortion;2)The online algorithm is used to solve the bitrate allocation and caching problems of the real-time immersive video,and the computational complexity is reduced by iterative solution in stages.Firstly,the greedy algorithm is used to get the approximate optimal solution in the bitrate allocation stage.Then,the 0-1 branch and bound algorithm is used to obtain the optimal solution in the caching stage.Finally,the purpose of improving the quality of immersive videos and reducing the transmission cost is achieved to get the approximate optimal utility of the whole system;3)In the experimental simulation stage,different video contents are used to simulate and compare the proposed algorithm by adjusting the bandwidth resources and the caching capacity of the edge node.The experimental results show that the system utility obtained by the proposed algorithm is higher than that of other benchmark algorithms,and the proposed algorithm is suitable for videos with little dynamic and obvious content change.