Performance comparison of a KF-based and a KF+VDFLL vector tracking-loop in case of GNSS partial outage and low-dynamic conditions

New GNSS tracking architectures are becoming of scientific interest since the last decade. The traditional DLL/PLL tracking-loops have been progressively substituted by KF/EKF scalar tracking and VDLL/VDFLL vector tracking methods. In this paper we will analyze in depth the performance of a KF scalar tracking and VDFLL vector in case of signal outage. Differently from papers on the same topic, we will show how the performance of a KF, when used within the tracking stage during a signal's blockage, are strongly dependent on a proper setting of the noise covariance matrices Q and P of the system, otherwise cases of frequency false-lock can occur when the outage condition lasts for a long time. Conversely, a VDFLL vector tracking algorithm is able to provide more precise frequency and code-phase's estimations then the KF scalar in case of signal's unavailability, but its performance relies on the navigation solution. Therefore, the effects of the PVT accuracy on the estimation of the main tracking parameters are also presented in this paper.