Exploring Observation-Domain with Nonlinear Time-Series Warping for Aided GNSS Navigation.

This paper investigates the application of sequence matching (SM) techniques to enhance aided GNSS navigation in lunar mission scenarios. Focusing on the challenges of aligning GNSS derived Position, Velocity, and Time (PVT) solutions with pre-designed Aiding Trajectories (ATs) under varying geometric conditions, the study identifies Dynamic Time Warping (DTW) as the most effective SM strategy. The novelty lies in extending nonlinear time-series warping methods towards the exploration of the observation domain, introduced as a promising alternative of our previous work based on a state domain approach. Merging GNSS observations as primary data sequences avoids the burden of GNSS only state estimation, particularly limited in cislunar environments, while enabling data fusion at an earlier stage of processing. A key advantage of the observation domain is its ability to support multidimensional sequence matching even in low visibility conditions typical of high-altitude navigation. Throughout this work, our aim is to refine weighting strategies and expand the use of observation domain SM for deep space navigation and signal processing, leveraging the flexibility and robustness it offers to address geometric challenges in lunar GNSS applications.