Skip to Main Content
Next generation roadway maps and vehicle navigation systems have the objective of reliably achieving where-in-lane positioning accuracy. Various methods are under consideration both to attain the requisite roadway map accuracy via post-processing and real-time vehicle positioning accuracy and reliability. Fundamental to these methods is the problem of accurately and reliably estimating a sensor platform trajectory in a post-processing environment. For mapping, the platform trajectory provides the pose for feature sensors (e.g., camera, LIDAR, RADAR). For navigation, the platform trajectory is the ground-truth reference. This article describes a smoothing framework for estimating sensor platform trajectories using an Inertial Measurement Unit (IMU) and a dualfrequency GPS pseudo-range and carrier-phase receiver. A Bayesian estimation framework is presented and transformed to a series of nonlinear least squares problems. The result of this optimization process is the platform trajectory estimate at the IMU measurement rate (200 Hz) with position accuracy at the centimeter level. One of the contributions of this research is the method developed to solve for the carrier-phase integer ambiguities. Real-world experimental results are presented to validate the proposed smoothing framework.