Spatiotemporal Local-Remote Senor Fusion (ST-LRSF) for Cooperative Vehicle Positioning

Area of Science:

• Intelligent Transport Systems (ITS)
• Robotics and Autonomous Systems
• Sensor Fusion and Estimation Theory

Background:

• Accurate vehicle positioning is critical for ITS protocols, algorithms, and applications.
• Current methods often rely on single-source data, limiting precision.
• Cooperative perception using vehicle-to-everything (V2X) communication offers potential for improvement.

Purpose of the Study:

• To develop a novel framework for cooperative vehicle positioning.
• To enhance absolute vehicle positioning accuracy by fusing local and remote sensing data.
• To analyze the theoretical and practical performance of the proposed fusion method.

Main Methods:

• Introduced the spatiotemporal local-remote sensor fusion (ST-LRSF) framework.
• Utilized on-board exteroceptive sensors for local estimates and V2X for remote estimates.
• Employed a greedy algorithm for minimal weighted matching (MWM) of vehicle states.
• Integrated an extended Kalman filter (EKF) with refined ST-LRSF positions.

Main Results:

• The ST-LRSF framework effectively fuses local and remote sensing data for improved vehicle positioning.
• Theoretical analysis shows position uncertainty is inversely proportional to the square root of matching size.
• Numerical simulations confirm high positioning accuracy across various cooperative scenarios.
• The EKF model further reduced positioning uncertainty using ST-LRSF refined measurements.

Conclusions:

• The proposed ST-LRSF framework significantly enhances absolute vehicle positioning accuracy in cooperative ITS.
• The fusion approach effectively addresses limitations of single-source positioning systems.
• The methodology provides a robust solution for precise vehicle localization in dynamic environments.