P2P Video-on-Demand: Steady State and Scalability

Abstract

Abstract-The fundamental P2P principle that downloading peers help other peers can be applied in the context of videoon-demand. This represents a demanding application combining aspects of other well-known P2P applications, i.e., live streaming and traditional file sharing. We seek to provide insight on fundamental questions about the performance and scalability of the system. A deterministic fluid model is derived that explicitly takes into account the video transfer and playback phases. The analytical results are complemented with extensive simulations from the corresponding stochastic model, as well as traces from a more realistic BitTorrent simulator.

How impatience affects the performance and scalability of P2P video-on-demand systems

“…The corresponding download and upload rates are: c = d/m and µ = u/m. As justified in [1], we assume that d > w.…”

“…Previous P2P video-on-demand models [6,1], however, omit the fact that the leechers may become impatient during the transfer phase and leave the system before the video file is completely downloaded. In this paper, we study how impatience affects the performance and scalability of BitTorrentlike P2P video-on-demand systems.…”

“…Therefore, it is clear that η < 1 in this case. One of our main conclusions in [1] was that the system operates properly (guaranteeing sufficient playback quality) whenever the efficiency parameter η exceeds the following threshold:…”

“…In [1], we presented a slightly different model. An essential difference between our model and that of Parvez et al is in the handling of the playback phase and the modeling of selfishness.…”

“…The model in [6] ignores the playback phase and simply assumes that any seed, whether it has played back the media file or not, departs with a constant rate. Instead of that, we include the playback phase explicitly in [1].…”