Distributed agreement in the presence of noise


In this paper, we propose a distributed dynamic consensus scheme which exhibits noise resilient property. In our setup, each node is modeled as a simple integrator. We use frequency-domain analysis to establish that the output of each node tracks the (weighted) average of the inputs of all nodes with zero steady-state error and show that a certain kind of graph Laplacain (called feasible Laplacian) is necessary in our framework. The eigenvalue location of feasible Laplacians turns out to be an LMI region, which allows us to derive an LMI test condition. When the noise is present, we show that the output tracking error has a bounded covariance.

3 Figures and Tables

Cite this paper

@article{Wang2009DistributedAI, title={Distributed agreement in the presence of noise}, author={Jing Wang and Nicola Elia}, journal={2009 47th Annual Allerton Conference on Communication, Control, and Computing (Allerton)}, year={2009}, pages={1575-1581} }