A neural network receiver for EM-MWD baseband communication systems

Baseband digital communication in “electromagnetic measurement while drilling” systems (EM-MWD) is often corrupted by surface noise. The conventional correlation receiver works well under the assumption of additive white Gaussian noise (AWGN); however in practice, the noise is actually non-stationary and usually contains spectral peaks in lower frequency range. In this research, a new approach based on artificial neural network is investigated. The neural network receiver has adaptive learning ability and outperforms the correlation receiver under various noise conditions, especially in the situation of non-white noise as well as the real world noise taken from actual drilling sites.