A Method for P2P Streaming System Join Latency Reduction Using Preferred Peer List

Abstract

In peer-to-peer streaming systems, a lower join latency is desirable. We propose a method to reduce the initial join latency by providing a list of preferred peers that have a high probability of being available to other peers. By collecting partial knowledge about the peer-to-peer overlay, the source peer maintains information about preferred peers. This information is supplied to new peers entering the streaming system. We develop lightweight cache algorithms and a P2P protocol. Analytical models are presented for the cache design and the probability distribution of join latency. The models and simulation results show that the PPL method can significantly reduce the initial join time compared to a prior art probe-based join scheme.