Stochastic Filtering with Networked Sensing
Résumé
This paper presents a solution to the problem of target tracking within a sensor network. This is based on modeling the target dynamics as a Markov process, and is reformulated as nonlinear filtering. When the target motion is a diffusion process, the optimal filtering involves a resolution of the backward Kolmogorov equation. Since an explicit solution for this partial differential equation does not exist in general, we recover the filtering by an alternative Monte-Carlo approach.
Origine : Fichiers produits par l'(les) auteur(s)
Loading...