Research On Optimal Method Of Short Video Cache Placement In Edge Computing Network

Edge computing is an emerging computing architecture that can provide better performance improvements for applications with low latency and high bandwidth requirements.As a part of cloud computing sinking to the user side,edge computing can effectively play the role of data aggregation.Edge computing collaborates with the cloud and terminals to analyze and control massive data.5G technology solves the problem of network access,but not the bandwidth and delay of backbone and metropolitan area networks.The connection of countless 5G smart terminals and Internet of Things terminals will result in the snatching of bandwidth resources.Therefore,the communication between the cloud,the edge,and the end nodes still faces the blocked network problem,and edge computing is a good solution.Internet-based online multimedia video(long video and short video)gradually replaces traditional TV,and the new format of video(4K,QHD,360~o,etc.)increases the bandwidth consumption of each content request.These factors have contributed to the explosive growth of video traffic,which is expected to exceed 80%of the total Internet traffic.Compared with the long video,the short video is a delay-sensitive application.The popularity of massive short video files and the characteristics of short duration have an uncertain impact on the user's experience quality.Therefore,in order to ensure the user's experience quality,caching technology in edge computing is a very promising solution to achieve the purpose of reducing latency and increasing the cache hit rate.In this paper,based on the data analysis of the emerging short video content,through the introduction of deep reinforcement learning algorithm,a cache model based on threshold-based partial cache short video is designed to achieve efficient caching of multimedia content on edge nodes.The main work of this paper is as follows:1)This article introduces the concept of edge computing technology,enabling technology and application scenarios,comprehensively introduces in-depth technology,and analyzes the characteristics of 5G technology communication and its impact on social and multimedia services.2)Researched the characteristics of short video content,namely Tik Tok short video,analyzed the metadata of Tik Tok short video,and carried out the attribute characteristics(length,bit rate,and size)of Tik Tok short video file Mathematical analysis.Subsequently,the Pearson correlation was used to analyze the popularity index of the Douyin short video.A sample study that uses the popularity distribution to increase the cache hit rate provides further guidance for the design of efficient edge cache systems in 5G networks.3)According to the characteristics of the file size and popularity of short video content,combined with the characteristics of the 5G edge computing network service,this paper proposes a strategy for placing cache content based on the segmentation threshold of the partial block cache.A good blocking threshold to maximize cache hit rate.4)The simulation experiment results show that the placement strategy of the partial cache based on the blocking threshold can obviously increase the number of short video caches and improve the cache hit rate while ensuring the user experience quality.Based on deep reinforcement learning to solve the optimal block threshold,the strategy of partial block caching can quickly obtain the maximum cache hit rate of about 90%and reach a stable state.