Outliers Based Caching of Data Segment with Synchronization over Video-on Demand using P2P Computing

Abstract

Nowadays live streaming plays an important role in various field of real time processing like education, research etc. The videos are maintained at the server then the clients may access the videos from the server may leads the performance problem. So the videos are downloaded by the clients and send it to the requesting clients. The mesh network is suitable for the live streaming because there is no master/slave relationship among the clients and in P2P (Peer-to-Peer) live streaming, each peer holds the video segment for satisfying the needs of the requesting peer. If the cache is full the data segments are replaced by using various replacement algorithms. These algorithms are mainly used in the data segment at the head part of the cache. The tail part of the data segments are never used to satisfy the peers even though it is a relevant data segments. The proposed work mainly focuses on the data segment of the tail part for live streaming is called outliers. The tail part of the cache is synchronized with the other neighboring peers with the help of the segment table. This segment table has to maintain each peer to overcome the unavailability of the data segment at the peers. There is various tag formats are proposed for representing the tail part of the cache. In future the performance will be improved in maximum level.