Bridging Integrated GPS/INS Systems with Geospatial Models for Car Navigation Applications

M. Attia, A. Moussa, N. El-Sheimy

Abstract:	Most vehicle navigation systems rely mainly on the Global Positioning System (GPS) receivers as the primary source of information to provide the position of the vehicle. GPS, however, can provide positioning information only under ideal conditions which require an open environment (i.e. open space areas). In other words, the system doesn’t work very well in urban, canopy areas due to GPS signal blockage. For the moment, any sophisticated urban application which demands essentially continuous position determination, cannot depend on GPS as a stand-alone system. More recently, the concept of combining complimentary navigation systems, such as Inertial Navigation System (INS), has been introduced even for commercial applications. The integration of the inertial sensors with GPS provides a system that has superior performance in comparison with either a GPS receiver or INS. However, INS systems typically exhibit very high positional drift when working in stand-alone mode. In this paper, we present a new technique to overcome this drift based on the integration of Geospatial Information System (GIS) model and map matching as an update to the navigation system solution. Vehicle navigation, unlike other applications such as aviation, marine or pedestrian navigation, is restricted to specific paths which is the street network. In order to reach a more reliable map matching results in vehicle navigation the actual maps geometry and topology was used to make a logical threshold for the GPS/INS solution. The proposed approach uses IMUs data to provide a more reliable navigation solution using GPS/INS integration, and also to aid the map matching stage by taking reliable decisions about the maneuvers. The proposed technique was tested in Calgary downtown. Four datasets were used, the first dataset for developing the algorithms and the other datasets for verifying the new technique. The data was collected in a land vehicle equipped with low-cost MEMS IMUs, a high sensitivity GPS receiver. A Geospatial data model was created for the study area, where the roads are modeled as segments connecting nodes, which represent the intersections. The GPS/INS solution is used to initiate the position of the vehicle in order to begin tracking its path. Whenever there is degradation in GPS coverage, the system will use the raw gyroscope measurements of the heading gyro to update the map matching with the actual driving path based on the developed spatial database. The developed system was proved to give map matching results identical to the actual path, which improves the positioning accuracy.
Published in:	Proceedings of the 23rd International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2010) September 21 - 24, 2010 Oregon Convention Center, Portland, Oregon Portland, OR
Pages:	1697 - 1703
Cite this article:	Attia, M., Moussa, A., El-Sheimy, N., "Bridging Integrated GPS/INS Systems with Geospatial Models for Car Navigation Applications," Proceedings of the 23rd International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2010), Portland, OR, September 2010, pp. 1697-1703.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In