A new blind adaptive beamformer for underwater communications

In this paper, a new blind adaptive beamformer is proposed for underwater communications. It exploits the cyclostationary property of modulated communication signals, which arises from carrier frequencies or baud rates. Spectral lines are generated by processing the underlying QPSK modulated signal via down-conversion, conjugation and multiplication. The beamformer coefficients are determined by minimizing the mean square error between the processed signal with spectral lines and a complex exponential extracted by phase-locked loop. Compared to traditional adaptive beamformers, the proposed beamformer is robust against the pointing error of the desired signal and the positioning error of the hydrophone array. To overcome the intersymbol interference, a decision feedback equalizer is combined with the proposed beamformer. Simulation results show that the proposed space-time processor is very effective for various channel models.