Exploiting WSSUS Multipath for GNSS Ranging

Kurzfassung

We propose a low-dimensional parameter estimation scheme for satellite-based ranging in the presence of an attenuated line-of-sight (LOS) signal and a time-varying number of multipath signals. This is of interest for landmobile users of global navigation satellite systems (GNSS) in diverse scattering environments with limited satellite visibility (e.g. a car moving through an urban canyon). Conventional ranging techniques, which rely solely on estimation of LOS parameters, ignore or suppress multipath and thus sacrifice available information; on the other hand, exhaustive approaches involving detection and estimation of each time-variant echo are too cumbersome in terms of parameter dimension. We suggest an unconditional maximum likelihood estimator (U-MLE) for joint estimation of LOS synchronization parameters and a small set of multipath distribution parameters. Our key assumptions are Gaussian wide-sense stationary uncorrelated scattering (WSSUS) and a known structure of the power delay profile (PDP). The proposed estimator uses as input data only a few complex-valued correlator outputs, which are provided by conventional navigation receivers, and its complexity is independent of the actual number of present echoes. Simulations demonstrate that the U-MLE improves accuracy and robustness of LOS synchronization for various model-matched and -mismatched multipath settings, including standardized and approved urban scenarios.