A new dual-filter tightly coupled model for pseudolite augmented GNSS real-time PPP

Abstract

The long convergence time is a key factor that hinders the wide application of Global Navigation Satellite System (GNSS) Precise Point Positioning (PPP). Due to the close distance of the pseudolite (PL) from the receiver, the moving receiver can produce rapid spatial geometric variations, which is beneficial to the fast convergence of the parameters. Therefore, this advantage of the pseudolite system (PLS) can be utilized to augment GNSS PPP for fast convergence. However, PLS may have limited coverage, which usually means a short augmentation time. Therefore, it is vital to make the GNSS PPP parameters (especially the GNSS ambiguity parameters) converge quickly. Thus, the potential of PLS should be exploited as much as possible, as well as the gain from PLS should be effectively utilized. In order to fully and reliably utilize the gain of ambiguity-fixed PLS solutions, we have proposed a dual-filter tightly coupled (DF-TC) model. The contributions of the PLS to GNSS have been theoretically analyzed under DF-TC, tightly coupled (TC), and semi-tightly coupled (STC) models. The experiments have demonstrated that all these coupled models can achieve good augmentation performance when the augmentation time is long enough. However, in the case of short-time augmentation, better augmentation performance can be achieved by utilizing the ambiguity-fixed PLS solutions. Among them, the DF-TC model shows the best augmentation performance, which can significantly reduce the augmentation time required by GNSS PPP, which means that the PLS coverage can be smaller, and the use of PLS can be more flexible and convenient. In addition, a short-time PLS augmentation can also lead to a significant improvement in the GNSS ambiguity fixing rate.