GPS multipath detection and rectification using 3D maps

Global Navigation Satellite Systems (GNSSs) suffer from the problems of invisible satellites and multipath effect in urban canyons. Many approaches have been employed to eliminate multipath signals in order to reduce positioning errors. Among these, those that consider the geometry of surrounding buildings to improve the evaluation of possible multipath signals have gained most interest. However, such approaches, although successful, require many satellites for positioning after eliminating multipath signals. This study proposes an approach in which the multipath signals themselves are used for positioning error correction. The proposed algorithm evaluates the pseudoranges of the possible multipath signals by referring to the building geometry. The assumed position is estimated by using the pseudoranges and is evaluated by the likelihood of the possible positioning error. The proposed method was verified through field experiments in urban canyons in Tokyo.