Issue No. 09 - September (2008 vol. 20)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/TKDE.2008.18
Live peer-to-peer (P2P) streaming has become a promising approach for broadcasting non-interactive media content from a server to a large number of interested clients. However, it still faces many challenges such as high churn rate of peer clients, uplink bandwidth constraints of participating peers, and heterogeneity of client throuput capacities. This paper presents a new P2P network called LSONet, a collaborative peer-to-peer streaming framework for scalable layer-encoded bit streams. The contributions are the combination of the advantages of both layered conding and mesh-based packet exchange. With layered coding, it overcomes overlay bandwidth limitatioins and heterogeneity of client capacities. With mesh based overlay streaming, it can better handle peer churns, as compared to tree-based solutions. For achieving these targets, this paper employs a gossip-based data-driven scheme for partnership formation, and proposes two algorithms, optimized transmission policy (OTP) and graceful degradation scheme (GDS), for multi-layers allocation. The proposed system is completely self-organizing, and in a fully distributed fashion. Extensive simulations show that LSONet achieves higher quality of service by peer-assisted streaming and layered video coding. Also, through comparison, results show that the system outperforms some previous schemes in resource utilization and is more robust and resilient for nodes departure, which demonstrate that it is well-suited for quality adaptive live streaming applications.
Modeling structured, textual and multimedia data, Distributed networks, Multicast, Reliability, availability, and serviceability
H. Guo and K. Lo, "Cooperative Media Data Streaming with Scalable Video Coding," in IEEE Transactions on Knowledge & Data Engineering, vol. 20, no. , pp. 1273-1281, 2008.