Terminal area navigation using relative GPS bias states with a terrain scene

Abstract

The effects of large GPS pseudorange bias errors and large absolute target location errors can be removed almost entirely from a guided weapon incorporating GPS updates in its Kalman filter by utilizing GPS bias states following a high quality terrain scene. Specifically, target relative bias states are added to the Kalman filter, and these states provide an accurate estimate of the total GPS bias after the terrain scene update is accepted by the Kalman filter. The performance of two different bias state configurations, one having three states and the other seven, is analyzed by means of a covariance simulation. The scenarios investigated include short and long terrain scene to target separations, both with and without the presence of high-powered GPS jammers and a softwareimplemented satellite constellation freeze. It is found that because of its virtual elimination of the pseudorange and absolute target location errors, and its affordability when compared to a terminal guidance seeker, the relative GPS bias state implementation is a viable precision strike candidate.