Distributed estimation by consensus filtering
speaker: Sandro Zampieri (Università di Padova)abstract: Assume we have a large number of sensors which can communicate each other through a communication network. The task of each sensor is to estimate a quantity which is modelled by a filtered white noise, as in the Kalman filtering classical set up. The goal is to obtain a variation of the Kalman filtering in which the sensors can take advantage both of their measures and of the information they receive from the other sensors. In other words the sensors have to collaborate by sharing their information and by reaching a consensus of their estimates. We analyse how the performance of this distributed estimation filter is influenced by the network topology and how random communication protocols can improve the estimation convergence. More precisely, in the context of randomized consensus, we will review some algorithms presented in the literature and we will propose new ones. We will present some performance results which will allow to make some comparison. Finally, we will establish some probabilistic concentration results which will give a stronger significance to results already proposed.
timetable:
Fri 6 Apr, 11:30 - 12:00, Aula Dini
<< Go back