Dual-Polarization MSBL-Based Beamforming for GNSS Multipath Mitigation

Multipath environment may enormously hamper the tracking performance. To enhance multipath mitigation, we propose a novel method based on multi-dimensional processing and multiple sparse Bayesian learning (MSBL). Our method utilizes sparsity in the spatial domain and signal characteristics in the polarization domain. Then we derive a joint off-grid angle and polarization MSBL-based beamformer. The LOS time delay is estimated from the proposed beamformer output with a maximum likelihood estimator. Simulation results demonstrate that our approach achieves close-to-optimal performance compared to the Cramer-Rao lower bound and outperforms an equivalent single-polarization beamformer with multipath signals.