| To efficiently deliver streaming media, researchers have developed technical solutions that fall into three categories, each of which has its merits and limitations. Infrastructure-based CDNs with dedicated network bandwidths and hardware supports can provide high-quality streaming services, but at a high cost. Server-based proxies are cost-effective but not scalable due to the limited proxy capacity in storage and bandwidth, and its centralized control also brings a single point of failure.Client-based P2P networks are scalable, but do not guarantee high-quality streaming service due to the transient nature of peers.Live streaming based on P2P has been developing rapidly. A number of popular software such as PPSTREAM, PPLIVE have emerged. Through these softwares users can enjoy real-time live P2P streaming media.But P2P live streaming has a natural disadvantage: users can enjoy the media like watching TV, but they are not able to execute the vcr operation to enjoy a complete streaming media ,and also they are not able to seek what they want.Therefore this paper focuses on how to supply vedio-on-demand service through P2P network. Because video-on-demand system is a asynchronous system,each user's start time is not the same, and users will implement operation sucn as Pause. Seek etc, so it is a challenging issue to provide large scale VoD services through p2p network. We have two ideas,one idea is how to share the resource of memory,the other is to divide a video file into segments and store these segments in the entire P2P network,so the p2p network can serve as a powerful server for the client as a substitude of the streaming server. Based on the above ideas,we give our double-cache stratege.We designed a popularity-based cache replacement strategy to replace cache in the disk. The main issue in this paper is: (1)To realize a P2P streaming for vedio-on-demand service. (2)Propose a MD cache strategy and a ring structure.(3) To study how to cache the media file in the entire P2P network by segment, and design a popularity-based cache replacement policy to reach a better QoS for users. |
