[Back]


Talks and Poster Presentations (with Proceedings-Entry):

M. Ashury, C. Eliasch, T. Blazek, C. Mecklenbräuker:
"Accuracy Requirements for Cooperative Radar with Sensor Fusion";
Talk: European Conference on Antennas and Propagation (EuCAP 2020), Kopenhagen, Dänemark; 03-15-2020 - 03-20-2020; in: "Proceedings of the 14th European Conference on Antennas and Propagation (EuCAP 2020)", (2020), 1 - 5.



English abstract:
Reliability and robustness are the essential requirements for automotive radar systems. However an automotive radar system suffers from environmental conditions and interference. Applying a proper radar data fusion algorithm can significantly increase the detection probability and robustness. This article investigates the accuracy requirements based on the geometric parameters for cooperative radar sensors with central sensor fusion. The results show that a radar sensor fusion for cooperative radar can increase the detection probability and therefore the system robustness by using at least two radar sensors. Furthermore, results show that three to four sensors at a minimum distance of ten meters are sufficient for high quality estimation. At the same time the system requirements regarding the accuracy for every single radar sensor can be relaxed to a range accuracy as low as one meter standard deviation.

German abstract:
Reliability and robustness are the essential requirements for automotive radar systems. However an automotive radar system suffers from environmental conditions and interference. Applying a proper radar data fusion algorithm can significantly increase the detection probability and robustness. This article investigates the accuracy requirements based on the geometric parameters for cooperative radar sensors with central sensor fusion. The results show that a radar sensor fusion for cooperative radar can increase the detection probability and therefore the system robustness by using at least two radar sensors. Furthermore, results show that three to four sensors at a minimum distance of ten meters are sufficient for high quality estimation. At the same time the system requirements regarding the accuracy for every single radar sensor can be relaxed to a range accuracy as low as one meter standard deviation.

Created from the Publication Database of the Vienna University of Technology.