GPS positioning, filtering, and integration

This paper explores the possibility of improving navigation solutions and easing integration requirements by using non-linear filters based on direct solutions-to the GPS equations. An alternative approach to navigation with GPS and integration of GPS with other sensor systems is discussed. This two-stage method is based on the use of statistical point estimation. After discussing the underlying concept of two-stage estimation, problems encountered in the statistics of point estimation of position and bias from GPS solutions are surveyed. This is followed by a presentation of a closed form maximum likelihood estimator for position and bias based on the assumption of Gaussian errors. To the authors' knowledge, this MLE algorithm has not previously been published. Applications to the integration of the GPS with the INS are presented.<>