3D LiDAR Aided GNSS NLOS Detection and Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current GNSS-RTK positioning systems face significant accuracy degradation in urban environments due to Non-Line-of-Sight (NLOS) receptions caused by signal reflections and blockages from buildings and dynamic objects, limiting their effectiveness in autonomous driving vehicles.
Innovation Solution
A 3D LiDAR aided GNSS NLOS mitigation method that integrates LiDAR and IMU factors using a local factor graph optimization to estimate relative motion, generates a 3D point cloud map for environment description, detects and corrects NLOS receptions, and employs a sliding window map to enhance satellite geometry and accuracy, using a weighting scheme to de-weight NLOS measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GNSS-RTK positioning is used in urban environments, then global-referenced positioning is provided, but positioning accuracy is significantly degraded due to NLOS reception and signal reflection
Solution Approach 1:
The patent uses the 3D LiDAR point cloud map, which was originally designed for environmental perception and obstacle detection, to identify NLOS satellites by checking whether satellite lines-of-sight are blocked by buildings. This converts the LiDAR data into a beneficial tool for GNSS signal quality assessment, transforming a potential harm (signal blockage) into a useful detection mechanism for improving positioning accuracy
Solution Approach 2:
The patent introduces an intermediary mechanism between LiDAR and GNSS by using the 3D point cloud map as a mediator. The LiDAR point cloud map serves as an intermediary data structure that translates environmental geometry into satellite visibility information, enabling the GNSS receiver to identify and exclude NLOS satellites without direct interaction between the LiDAR and GNSS systems
2Measurement precision
If 3D LiDAR is used to detect NLOS satellites, then positioning accuracy is improved, but system complexity increases due to integration of multiple sensors and processing algorithms
Solution Approach 1:
The patent makes the 3D LiDAR point cloud map serve multiple functions: it is used for both environmental perception (original function) and NLOS satellite detection (new function). By making the LiDAR data multi-functional, the system avoids adding separate detection hardware, thereby reducing overall system complexity while maintaining improved positioning accuracy
Solution Approach 2:
The patent merges the LiDAR processing pipeline with the GNSS positioning pipeline by integrating NLOS detection into the existing point cloud map generation process. Instead of creating separate systems for LiDAR mapping and GNSS correction, the patent combines them by using the same point cloud data structure for both environmental representation and satellite visibility analysis, reducing system complexity through consolidation
3Measurement precision
If all NLOS satellites are excluded from positioning calculations, then measurement accuracy is improved, but the number of available satellites decreases
Solution Approach 1:
The patent applies local quality by selectively excluding only those satellites whose lines-of-sight are blocked by buildings, rather than excluding all low-elevation satellites. Each satellite is individually assessed based on its specific geometric relationship with the vehicle and surrounding buildings, allowing the system to maintain the maximum possible number of usable satellites while removing only the locally problematic NLOS signals
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The method achieves high precision positioning with 10-centimeter accuracy in urban canyons by effectively mitigating NLOS effects, improving satellite geometry, and reducing positioning errors, thereby supporting the navigation requirements of autonomous driving vehicles.
Implementation Method 1
a 3D LiDAR sensor configured to generate a description of a local environment in which the autonomous driving vehicle is located
Implementation Method 2
receiving an IMU factor, from an inertial measurement unit, corresponding to a relative motion between two epochs
Data Source
AI summary
A method for supporting positioning of a vehicle using a satellite positioning system is disclosed. The method includes generating, in real-time, a sliding window map (SWM) based on 3D point clouds from a 3D LiDAR sensor and an attitude and heading reference system (AHRS), wherein the SWM provides an environment description for detecting and correcting a non-line-of-sight (NLOS) reception; accumulating the 3D point clouds from previous frames into the SWM for enhancing a field of view (FOV) of the 3D LiDAR sensor; receiving global navigation satellite system (GNSS) measurements from satellites, by a GNSS receiver; detecting NLOS reception from the GNSS measurements using the SWM; correcting the NLOS reception by NLOS remodeling when a reflection point is not found in the SWM; and estimating a GNSS positioning by a least-squares algorithm. It is the objective to provide a method that mitigates NLOS caused by both static buildings and dynamic objects.