Multipath Rejection Using Multi-Signal Multicorrelator Based GNSS Receiver With an Extended Kalman Filter

Abstract

In Global Navigation Satellite Systems (GNSSs), multipath propagation is still a major source of error, in particular in urban environment. Obstacles in the nearer receiver environment can reflect satellite signals leading to additional delayed signal replicas at the receiver. This causes conventional GNSS receivers to provide biased range estimates leading to Position, Velocity and Time (PVT) errors. Especially newly arising safety-relevant applications, such as autonomous cars or unmanned aerial vehicles, raise demand for high accuracy and fail-safe positioning. GNSS positioning with deviations of several meters, as it may well occur in multipath environments, would be insufficient here. A novel multi-signal Extended Kalman Filter (EKF) based multipath mitigation solution is proposed. It relies on a multicorrelator structure replacing the conventional Delay Locked Loop (DLL) code tracking. The multipath rejection is implemented through the underlying signal model of the EKF. It is incorporating the radio propagation channel between satellite and receiver, inherently accounting for reflected signal replicas. In addition to the increased resilience against multipath propagation, an estimate of the Channel Impulse Response (CIR) of the radio channel is obtained. Furthermore, this approach is also able to track in one GNSS frequency multiple signals per satellite simultaneously, exploiting the advantage of increased signal power and robustness. The functionality of the multi-signal EKF has been demonstrated with simulations and compared to a conventional receiver with a hardware GNSS constellation simulator. The results were subsequently verified with actual measurements. The proposed approach turned out to be very effective against multipath. Pseudorange errors were able to be reduced drastically leading to more accurate positioning.