Distributed Estimation Under Network Model Uncertainty

Saurabh Sihag; Ali Tajer

ICASSP 2018accepted0 citations

Distributed Estimation Under Network Model Uncertainty

Abstract

This paper considers the problem of distributed state estimation in an interconnected network, in which there is uncertainty in the true model. Such uncertainties are due to the possibility of disruptions or changes in the nominal model. The focus is on the setting in which the true network model belongs to a set of possible models. Forming an optimal estimate has high computational complexity in large networks and, therefore, this paper treats this problem in a distributed framework. The key observation is that the estimation quality critically depends on successful isolation of the true model. On the other hand, the true model cannot be determined perfectly due to noisy measurements. Based on these observations, this paper formulates a composite hypotheses testing problem and provides optimal decision rules that account for estimation quality and detection performance. The theory developed in this paper is evaluated via a case study.

BibTeX

@inproceedings{icassp2018_distributedestim,
  title = {Distributed Estimation Under Network Model Uncertainty},
  author = {Saurabh Sihag and Ali Tajer},
  booktitle = {ICASSP 2018},
  year = {2018}
}