Multipath Rejection Using Multi-Correlator Based GNSS Receiver in Vector Tracking Loops

Abstract

In global navigation satellite systems (GNSSs), multipath propagation is to this day a major source of error. Especially in urban environments, satellite signals are reflected by buildings and other objects, leading to additional delayed signal replicas at the receiver. The frequent blockage of line-of-sight (LOS) signals further increases complexity here. A conventional GNSS receiver relying on scalar tracking (ST) and a delay locked loop (DLL) for the code tracking can only provide inaccurate position estimates under such conditions. Newly arising safety-relevant applications like self-driving cars, unmanned aerial vehicles, but also autonomous maritime vessels in harbor proximity raise demand for high accuracy and reliable positioning. Conventional GNSS receivers with errors of several meters are here insufficient. Vector tracking (VT) is a powerful solution for a reliable positioning in such challenging environments. However, its capability to suppress the effect of multipath propagation is limited and further reduced the more satellites are affected by multipath. Therefore, a VT loop with an integrated multipath mitigation is proposed. This is accomplished with an Extended Kalman Filter (EKF) replacing the DLL discriminator for the code tracking. Its measurement model incorporates the radio propagation channel between satellite and user, inherently accounting for multipath. The joint estimator determines therewith not only the code delay, but also an estimate for the channel impulse response (CIR). A drawback of this algorithm – but also of a conventional DLL discriminator – is, that an unobstructed LOS path is required. Combining this approach with VT overcomes this limitation and reduces the effect of multipath in challenging environments.