Abstract
Vehicle self-localization is an important and challenging issue in current driver assistance and autonomous driving research activities. An accurate and reliable localization technique can make the vehicle be guided along the desired trajectory, furthermore, contribute to vehicle cooperation. Global Positioning System (GPS) has been proven itself reliability for vehicle self-localization in the open sky. However, it suffers from the effect of multipath and Non-Line-Of-Sight (NLOS) propagation in urban canyon. In order to reduce both multipath and NLOS effects in GPS positioning, this paper utilizes 3D building map and ray-tracing algorithm to simulate the reflecting path of satellite signal transmission. Moreover, to increase the accuracy and the robustness with respect to the localization result, the information from on-board sensors, including speedometer and gyro sensor, are integrated with the GPS positioning result in a Kalman filter framework. A series of experiments in different scenarios are conducted in an urban canyon. The experimental results demonstrate the accuracy of our proposed method and its feasibility for autonomous driving.
| Original language | English |
|---|---|
| Pages (from-to) | 565-572 |
| Number of pages | 8 |
| Journal | Proceedings of the Institute of Navigation Pacific Positioning, Navigation and Timing Meeting, Pacific PNT |
| Volume | 2015-April |
| Publication status | Published - 2015 |
| Externally published | Yes |
| Event | Institute of Navigation Pacific Positioning, Navigation and Timing Meeting, PACIFIC PNT 2015 - Honolulu, United States Duration: 20 Apr 2015 → 23 Apr 2015 |
ASJC Scopus subject areas
- Transportation
- Ocean Engineering
- Aerospace Engineering
- Electrical and Electronic Engineering
- Computer Science Applications