Randomized rumor spreading is an efficient protocol to distribute information in networks. Recently, a quasirandom version has been proposed and proven to work equally well on many graphs and better for sparse random graphs. In this work we show three main results for the quasirandom rumor spreading model. We exhibit a natural expansion property for networks which suffices to make quasirandom rumor spreading inform all nodes of the network in logarithmic time with high probability. This expansion property is satisfied, among others, by many expander graphs, random regular graphs, and Erdos-Rényi random graphs. For all network topologies, we show that if one of the push or pull model works well, so does the other. We also show that quasirandom rumor spreading is robust against transmission failures. If each message sent out gets lost with probability f, then the runtime increases only by a factor of . © 2009 Springer Berlin Heidelberg.
CITATION STYLE
Doerr, B., Friedrich, T., & Sauerwald, T. (2009). Quasirandom rumor spreading: Expanders, Push vs. Pull, and robustness. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 5555 LNCS, pp. 366–377). https://doi.org/10.1007/978-3-642-02927-1_31
Mendeley helps you to discover research relevant for your work.