Planning for terrain-aided navigation

The purpose of this paper is to investigate the planning of a mobile trajectory in order to use its own motion for improving its position estimation. This optimization procedure relies upon two basic ingredients: 1) a reference measurement map of the area of interest is available before departure, and 2) real time measurement sensors placed aboard the mobile. The main objective is to plan a trajectory which minimizes the localization error along the path or/and at the arrival area. The general framework of the Markov decision process coupled with an auxiliary local cost function are the basic ingredients of a sub-optimal algorithm. The quality of the optimization scheme is evaluated by deriving the posterior Cramer-Rao bounds of the non linear discrete-time system.