Research On Optimization Strategy Of Streaming Media Cache Based On CDN

According to relevant statistics,short video,live / on-demand and other streaming media traffic has increased dramatically in recent years,which has exceeded 70% of the total network traffic.Every day,a large number of users use various handheld mobile devices to record and upload small network videos.These small videos will be cached in the CDN source server.When network users request to access the video,they will,CDN source server will use streaming media technology to transmit video cache to CDN cache server and realize real-time playback of small video.In order to improve the cache efficiency of CDN server and enhance the user's audio and video viewing experience,it is necessary to study the streaming media cache optimization strategy.This topic collects video data from a domestic short video platform by writing a crawler program,sets a preference value calculation formula,establishes a video value model according to the formula,and then proposes a cache strategy algorithm based on the video value model,which aims to improve the hit rate of users' requests for hot events in a certain period of time,reduce the back source of edge node server,and save bandwidth overhead Reduce the pressure of backbone network.The main work of this paper includes the following three parts:(1)The video content requested by the user is collected and feature extracted.The calculation formula of video preference value is formulated and the video value model is built.The preference value data of each video is calculated according to the calculation formula of video preference value.The videos are sorted according to the preference value from high to low.The top ranked videos first enter the server cache space,while the bottom ranked videos will be eliminated on a regular basis.(2)A cache algorithm based on video value model is proposed.The video value model takes the time-varying characteristics of user access requests into account.The impact of user access request behavior on cache hit rate is fully considered,and the cache strategy of local area cache server cooperative operation is adopted.(3)A simulation experiment is designed.Part of the experiment is to simulate the user request behavior,and the other part is to simulate the streaming media cache algorithm.The cache hit rate is taken as the measure of the performance of the cache algorithm based on the video value model.Experimental results show that the proposed caching algorithm based on video value model can improve the hit rate of users' requests to a certain extent and reduce the load of the source server.