The needs of vehicle navigation system are increasing because they provide the solution for many problems which today's society is faced with, such as traffic accidents, heavy congestion, and a shortage of parking lots. Although technologies related to the current navigation system cover numerous areas, including positioning, map indication, display, communication and etc., the accurate determination of position location is still of primary importance.
GPS (Global Positioning System) receiver and low cost dead reckoning sensors have been used for vehicle position location. The integration of these can improve the system performance to cope with accumulating errors and multipath problem in urban area. However, the implementation of integrated system of two sensors is not easy, for their real data have noises that are not apt to be modeled.
This paper discusses how to combine a GPS receiver and dead reckoning sensors. The characteristics of sensors and their data are investigated, and the system for field test is manufactured. The models for the system and the sensors are suggested and extended Kalman filter is applied to fuse them. The test results of a selected typical route are presented.
