EMD filter method and its application in GPS multipath

Empirical Mode Decomposition (EMD) is a relatively new signal processing technique. It starts from the data itself to decompose the signals in spatial domain, so it can discriminate the signals from the noise. A noise filter (EMD noise filter hereinafter) based on the characteristic of IMF (intrinsic mode function) is developed. Then, simulated data and real data (i.e., GPS data) are used to test the method separately. The following are the conclusions drawn from these tests: 1 EMD noise filter can better mitigate the random errors. Compared with wavelet method, it is more straightforward, and not subject to the effects from the selection of certain wavelet function. 2 EMD noise filter is more efficient in eliminating instantaneous strong noise than the wavelet method. It therefore can better extract and thus model the repeatability of the multipath errors, which is crucial in mitigating the errors.