Results of a real time implementation of vector delay lock loop in a dynamic field test

Regular GNSS receivers track the signal of each satellite in view independently, which commonly referred to as scalar tracking. However, with this method of signal tracking the visibility of the satellite and signal degradations have a high impact on the tracking and and reduces the robustness of the receiver. To overcome these common weaknesses, other approaches such as vector tracking (VT) exist. This approach benefits from the usage of the antenna movement and the dependence of the satellite distances to each other. Hence it is possible to provide a robust and continuous navigation solution as it is more robust against shadowing effects. One method of the VT is the so called vector delay lock loop (VDLL). The scalar code tracking loops are replaced with a single interlaced estimation algorithm. Present developments have shown good results in post processing for this. Our previous implementation in real-time already showed good results with simulated GPS L1 C/A signals. This paper presents the results of the ongoing work and research of a VDLL implementation with an extended Kalman filter (EKF) on a GNSS receiver in a field test. The test setup, the conditions and results are presented. A record and replay unit of a GNSS hardware receiver is used to evaluate the algorithm. This allows a repeatable and comparable real-time testing of the VDLL algorithm with pre-recorded real GNSS signals. In such scenarios, the VDLL must take into account more uncertainties, such as atmospheric and satellite clock errors.