Randomized Decentralized Broadcasting Algorithms

— We consider the problem of broadcasting a live stream of data in an unstructured network. Broadcasting has been studied extensively for networks with capacity constraints at the edges. We give the first completely distributed algorithm that optimally solves the broadcast problem in edge-capacitated networks, providing a new proof of Edmonds’ theorem. Motivated by the access capacity limitations in todays Internet, we study the problem of efficient decentralized broadcasting in node capacitated networks. We present the first completetly decentralized algorithm for solving the node-capacitated broadcast problem, and show some of its optimality properties analytically and through simulation. We study the delay that users must wait in order to playback the stream with a small number of skipped packets, and discuss the suitability of our algorithms for live video streaming.

• Type: Conference
• Year: 2007
• Where: INFOCOM
• Authors: Laurent Massoulié, Andrew Twigg, Christos Gkantsidis, Pablo Rodriguez
• Keywords: Broadcast Problem, Broadcasting, Communications, Efficient Decentralized Broadcasting, INFOCOM 2007