Lightweight Coloring and Desynchronization for Networks
Arik Motskin, Tim Roughgarden, Primož Škraba, Leonidas Guibas · 2009
We study the distributed desynchronization problem for graphs with arbitrary topology. Motivated by the severe computational limitations of sensor networks, we present a randomized algorithm for network desynchronization that uses an extremely lightweight model of computation, while being robust to link volatility and node failure. These techniques also provide novel, ultra-lightweight randomized algorithms for quickly computing distributed vertex colorings using an asymptotically optimal number of colors.