The Network Optimization Based On Resources Regionalization And Flow Localization

With the rapid development of the internet business, the demand of network broadband is also accelerating. Many factors have affected the user experience, among them, The Internet Service Provider plays a important role, In order to reduce the high-risk single-point, almost all the inland deploy resources of Iternet Service Provider are distributed uniformly in different areas; meanwhile, to meet user's needs, it is necessary to use a long trunk to dispatch the different resources between each district.As is known to all, the large Internet Service Provider transmission backbone range of the interval will be in the form of circular or multipoint, but for the long trunk, the cost will be huge, no matter in investment, maintaining or expansion, The key problem for Internet Service Provider is how to slow down the expansion of the long-distance link while facing the fast growing source deployment.Applying the basic theory of data communication, This paper mainly analyses and concludes these two kinds of application which take up the largest internet traffic proportion: P2P and network streaming media, tests one of the highly used of platform : Xunlei and Tudou Net ,studies its work principle and mechanism. What's more, analyses and reseaches the Server Assistant , Peer Assistant and other concepts from P2SP ,And then analyses the collection of the technical infomation for derivation reduce ,Proposing flow localization and the resources regionalization . Using hierarchical scheduling mechanism and combining with cloud computing as a blueprint of resources reasonable distribution to realize the flow localized in the best effect , which all for reducing the operating costs of Internet Service Provider's . In order to test and verify this project , The paper has designed the experimental and the real situation tests to prove the proposal. Test Results show that the flow localization and resources regionalization are effective and viable.