Abstract
The advancement of low-earth-orbit (LEO) communication constellations has revitalized interest in Doppler-based positioning. However, conventional Doppler positioning algorithms struggle with dynamic receivers under unknown initial states due to the inherent nonlinearity of the observation model. To address this challenge, we propose an improved least-squares-based algorithm that decouples the estimation of position and velocity, enabling robust positioning from a zero initial state. Simulation results demonstrate that the proposed method achieves meter-level positioning accuracy and decimeter-per-second velocity accuracy under various dynamic scenarios, including high-speed motion. This approach establishes a viable framework for real-time navigation in GNSS-challenged environments using LEO signals.