Wavelet Decomposition for Denoising GPS/INS Outputs in Vehicular Navigation System

This paper aims to bring together Global Positioning System (GPS) and Inertial Navigation System (INS) integration approach for data fusing using wavelet analysis. The wavelet is employed to compare the sensors outputs at different resolution levels and for several types of errors in INS and GPS as well as to smooth and predict the INS errors. Simulation results show that, the types of filters which are used in multiresolution algorithm have large effects in reducing the INS/GPS error in position and especially in velocity components. Furthermore, soft thresholding technique is introduced in multi-resolution algorithm, where six rules are proposed to select the threshold value. Real GPS and INS measurements were used in the Matlab simulation to illustrate the effectiveness of the proposed method. Key-Words: Vehicular navigation, Inertial navigation system, GPS, Wavelet Analysis, Denoising.