No abstract

No abstract

Common peer-to-peer (P2P) file sharing clients usually download at an unstable rate and hardly exploit the available bandwidth offered by low rate sources. The characteristic fluctuational throughput of the source peers might be caused by user behavior factors such as running other bandwidth consuming tasks, throttling of download speed by P2P software or even termination of the source.In this paper we propose Collabory -a solution for stabilizing and accelerating the download speed rate in existing P2P networks. We introduce a new role: "Feeders" -peers that collaboratively aggregate the downloads from multiple sources into a single, stable stream served to the downloading peer. We show that the solution utilizes source nodes with an extremely low and unstable throughput without reducing the download rate of the downloading peer.Measurements in a test suite expressed a major increase in download rate and stability. Upgraded & stabilized throughput is demonstrated on eMule.

In today's extensive worldwide Internet traffic, some 60% of network congestion is caused by Peer to Peer sessions. Consequently ISPs are facing many challenges like: paying for the added traffic requirement, poor customer satisfaction due to degraded broadband experience, purchasing costly backbone links and upstream bandwidth and having difficulty to effectively control P2P traffic with conventional devices.Existing solutions such as caching and indexing of P2P content are controversial as their legality is uncertain due to copyright violation, and therefore hardly being installed by ISPs. In addition these solutions are not capable to handle existing encrypted protocols that are on the rise in popular P2P networks.Other solutions that employ traffic shaping and blocking degrade the downloading throughput and cause end users to switch ISPs for a better service.LiteLoad discerns patterns of user communications in Peer to Peer file sharing networks without identifying the content being requested or transferred and uses least-cost routing rules to push peer-to-peer transfers into confined network segments. This approach maintains the performance of file transfer as opposed to traffic shaping solutions and precludes internet provider involvement in caching, cataloguing or indexing of the shared content. Simulation results expresses the potential of the solution and a proof of concept of the key technology is demonstrated on popular protocols, including encrypted ones.

In theory, peer-to-peer (P2P) based streaming designs and simulations provide a promising alternative to serverbased streaming systems both in cost and scalability. In practice however, implementations of P2P based IPTV and VOD failed to provide a satisfying QoS as the characteristic fluctuational throughput of a peer's uplink leads to frequent annoying hiccups, substantial delays and latency for those who download from it. A significant factor for the unstable throughput of peers' uplink is the behavior of other processes running on the source peer that consume bandwidth resources.In this paper we propose Maxtream -a machine learning based solution that actively predicts load in the uplink of streaming peers and coordinates source peers exchanges between peers that suffer from buffer underrun and peers that enjoy satisfactory buffer size for coping with future problems.Simulation and experiments have shown that the solution successfully predicts upcoming load in popular protocols and can improve the QoS in existing P2P streaming networks.