Abstract
A decomposition is given for the implementation of the Kalman filter as a collection of parallel processors. This decomposition is based on the representation of the system as a direct sum of observability subspaces.

This publication has 1 reference indexed in Scilit: