Content and overlay-aware scheduling for peer-to-peer streaming in fluctuating networks

Abstract

Due to the scalability of peer-to-peer structure, it is widely applied in multimedia streaming applications to provide services for large number of clients concurrently. A lot of research work has been done on improving the data distribution efficiency by adapting the overlay structure to the dynamically changing networks. Instead of changing overlay structure, we propose a data packet scheduling algorithm to improve transmission efficiency of the streaming system. Our scheduling algorithm distributedly prioritizes the data packets and sends the most important data to the whole system first. It avoids wasting network resources while frequently switching peer connections. The algorithm serves to provide an optimal ordering solution to minimize overall latency based on overlay structure information. With the help of content-aware weighting scheme, the scheduling algorithm also improves the streaming quality at the peers. Our algorithm shows good performance even under a dynamic and challenging network environment with negligible algorithm overhead.