TítuloHow to Implement Doubly-Stochastic Matrices for Consensus-Based Distributed Algorithms
Tipo de publicaciónConference Paper
Year of Publication2014
AutoresValcárcel-Macua, S, Moreno-León, C, Romero, JS, Pereira, SSilva, Zazo, J, Pagès-Zamora, A, López-Valcarce, R, Zazo, S
Conference NameIEEE Sensor Array and Multichannel Signal Processing Workshop (SAM)
Conference LocationA Coruña, Spain
Palabras clavecompass, dynacs, wsn
Resumen Doubly-stochastic matrices are usually required byconsensus-based distributed algorithms. We propose a simpleand efficient protocol and present some guidelines for implementingdoubly-stochastic combination matrices even in noisy,asynchronous and changing topology scenarios. The proposedideas are validated with the deployment of a wireless sensornetwork, in which nodes run a distributed algorithm for robustestimation in the presence of nodes with faulty sensors.