Abstract
Peer-to-peer (P2P) live streaming systems gain a lot of attention recently. Among the various architectures, the gossip-based approach holds great potential and has been used in real systems. In a gossip-based P2P streaming system, peer nodes are grouped together to exchange pieces of the live stream with each other. Maintaining the group, called a swarm, and pulling the streaming data from other peers incur nontrivial overhead. This thesis proposes to put the nodes in a swarm into a cooperative relationship. The idea is to gather nodes that are physically close into clusters, and let nodes in the same cluster cooperatively pull streaming data from outside of the cluster and actively push collected data to other nodes in the cluster. By cooperating with cluster members, the proposed CSLive system can reduce the number of partners with which a node needs to maintain, whereby cutting down the control overhead in gossip-based live streaming systems. Furthermore, with the cooperative operations, a node can expect to receive most streaming data from nodes in the same cluster, which have good communication quality, and only need to pull the responsible data from nodes outside of the cluster. The performance of the whole system may be improved. The proposed system is evaluated with simulation.