Playout adaptation for peer-to-peer streaming systems under churn

We address the problem of playout adaptation in peer-to-peer streaming systems. We propose two algorithms for playout adaptation: one coordinated and one distributed. The algorithms dynamically adapt the playback delay of the peers so that the playout miss ratio is maintained within a predefined interval. We validate the algorithms and evaluate their performance through simulations under various churn models. We show that playout adaptation is essential in peer-to-peer systems when the system size changes. At the same time, our results show that distributed adaptation performs well only if the peers in the overlay have similar playback delays. Thus, some form of coordination among the peers is necessary for distributed playout adaptation in peer-to-peer streaming systems.