An iterative LiDAR DEM-aided algorithm for GNSS positioning in obstructed/rapidly undulating environments

This paper describes a new algorithm to aid stand-alone GNSS positioning in areas of bad signal reception using a Digital Elevation Model (DEM). Traditional Height-Aiding (HA) algorithms assume either a preset (fixed) value for the receiver elevation or rely on the elevation value that corresponds to the nearest available position fix. This may lead in erroneous receiver elevation estimates that, under circumstances, are inefficient to aid effectively GNSS positioning. In this study, the receiver elevation is updated at every iteration step of the navigation solution through dynamic interpolation of the elevation model. The algorithm, because of its ability to extract and fully exploit the elevation information derived from a digital model, it can prove particularly useful in forested areas with steep-sloped terrain. Extended test runs were undertaken to validate the correctness of the mathematical model and the feasibility of the algorithm and associated software. Particularly, analysis of a dataset acquired in a forested, rapidly undulating environment reveals significant average improvement in all performance metrics of positioning, namely the GNSS position availability (50%), accuracy (56%) and external reliability (86%) compared to the Standard Point Positioning (SPP) solution. Moreover, it was found that the method can cope successfully in marginal operating conditions with situations of bad satellite geometry and satellite signals affected by interference due to tree canopy.