Blind channel estimation combined with matched field processing in underwater acoustic channel

Blind channel estimation methods normally use the statistical properties of received signals to restore the original waveform without a priori information. However, these methods need a high computational complexity especially in multipath environments, and their performances are not good enough. A blind channel estimation method utilizing acoustic propagation physics via normal mode is proposed in this paper. Some environmental information and matched field processing (MFP) is used for source localization by matching array data to solutions of an ocean waveguide. This method could in theory acquire more accurate channel information and without requiring any information about transmitted signal. Numerical simulations are presented to demonstrate the accuracy and robustness of this method in multipath environments.