Abstract

From the past few years, peer-to-peer (P2P) applications have becoming well-liked magnificently. BitTorrent (BT) has one of the most effective mechanisms for P2P content distribution among all P2P applications. In P2P applications, each node plays a role both as a client and a server simultaneously, comparing to the traditional client-server systems, where every node has only one responsibility to act as either a client or a server. Therefore, in P2P applications, the upload bandwidth of each peer can be counted as a significant resource of the network. Although BT was created for file-sharing purposes at the beginning, which is a time-insensitive distribution, after a while, it has attracted the attentions to be use for video/audio streaming purposes too, which are time-sensitive. The importance of this capability is that peers now, can be able to watch, or listen to, their favorite live streaming content concurrently. Motivated by this fact, a stochastic model for a BT-based P2P live streaming system is proposed and numerically solved and based on what we gain, it is also shown how the performance of the system can be affected by different parameters of the system (such as the number of neighbors, delay time, size of the buffer, etc.). Moreover, we also try to apply some minor changes needed in the BitTorrent's mechanisms, in order to support the video/audio streaming more efficiently.