Distributed state and field estimation using a particle filter

This paper addresses distributed tracking of the probability distribution of state parameters of a space-time-variant process described by a linear partial differential equation using a particle filter (PF). We focus on localizing an acoustic source in a given region. The underlying wave equation leads to a high-dimensional estimation problem. We propose a technique for reducing the computational effort of the PF which allows to distribute the signal processing algorithm over the nodes of a wireless sensor network (WSN) without a fusion center.