GPS Multipath Mitigation Based on Improved Empirical Mode Decomposition

Multipath is a challenging error in short-baseline GPS applications. A new multipath mitigation method is proposed based on empirical mode decomposition (EMD). Firstly, the propagation of noise during the EMD of noisy signal is studied by singular spectrum analysis (SSA) and then an improved EMD method is developed based on the compression of noise. Secondly, considering the position sensitivity of EMD an iterative method is adopted to average the error caused by position change of noise. Finally, a de-noising method based on the improved EMD and iterative algorithm is proposed. Both simulated signals and real GPS observations are adopted to verify the new method, results show that our method is outperform of wavelet-based method and has more robustness than EMD-CIIT.