Proxy Cache Replacement Algorithm Based On Popularity Prediction Gain Model Of Streaming Media

Network technology developed rapidly today, the portion of streaming media ininformation transfer of network is becoming larger and larger. And because of the growth ofnetwork application, and Internet totally assigns such as WAN, the congested phenomenon ofthe network grows in intensity too.In order to improve streaming media transmission performance of data in WAN, proxycache server is added into content distribute system in streaming media. Set proxy cache atthe edge of network, and useful streaming media data is saved on the acting server, the proxycache server will be asked to provide service to shorten the delay of starting for the users.Because the capacity of the proxy cache server is limited, the most useful streaming mediadata must be saved on the proxy cache server to guarantee the server can give full play to therole of it. So proxy cache replacement algorithm determines performance of the acting serverdirectly.This paper makes improve based on traditional proxy cache replacement algorithm.Media quality, delayed start and cost of transfer factors are added in popularity predict proxycache replacement algorithm of streaming media. Experiment show that the improvedalgorithm can improve the proxy system performance flexibly in different environments;also it can reflect the users’ services and technique requirements of agent server.This paper researches impact on user’s end the factor of media quality on the conditionthat starts delay and network to consume certainly.The validity of this algorithm has already operated proving through experiments.