ABSTRACT
This chapter discusses the general recursive estimation and, in particular, the Kalman filter. It outlines a Bayesian approach to probabilistic information fusion. All data fusion problems involve an estimation process. An estimator is a decision rule which takes as an argument a sequence of observations and computes a value for the parameter or state of interest. The Information filter is essentially a Kaiman filter expressed in terms of measures of information about the parameters (states)of interest rather than direct state estimates and their associated covariances. The notion of Fisher information is useful in estimation and control. It is consistent with information in the sense of the Cramer-Rao lower bound. The two key information-analytic variables are the information matrix and information state vector. The chapter describes the Kalman and Information filters are compared by simulating the constant velocity system. The Kaiman and Information filters are demonstrably equivalent.
