Real-Time and Post-Mission Heading Alignment for Drone Navigation Based on Single Antenna GNSS and MEMs-IMU Sensors

This paper aims to develop a heading alignment procedure for drone navigation employing a single hover GNSS antenna combined with low-grade MEMs-IMU sensors. The design was motivated by the need for a drone-mounted differential interferometric SAR (DinSAR) application; however, the methodology proposed in this work can be applied to any Unmanned Aerial Vehicle (UAV) application that requires high-precision navigation data for short-flight missions utilizing cost-effective MEMS sensors. The method proposed here involves a Bayesian parameter estimation based on a simultaneous cumulative Mahalonobis metric applied to the innovation process of Kalman-like filters, which are identical except for the initial heading guess. The procedure is then generalized and called parametric alignment. The motivation for the multidimensional extension in the scenario is also presented. The method is highly applicable for cases where gyro-compassing is not available, usually for low-cost UAV applications. It employs the most straightforward optimization techniques that can be implemented using a real-time parallelism scheme. Numeric simulations and experimental evaluations using a real UAV demonstrate that the proposed method can provide initial heading alignment when the heading is not directly observable during takeoff.