Abstract:

An estimation algorithm of multiple-input multiple-output orthogonal frequency division multi-plexing (MIMO-OFDM) underwater acoustic channels, based on compressed sensing (CS） theory, is proposed, which makes full use of sparse feature of the underwater acoustic channels. A complete dictionary is designed to meet the framework of CS theory under the effect of Doppler frequency shift, which is based on MIMO-OFDM underwater acoustic communication system model, and then a series of non-orthogonal basis in a complete dictionary is used to describe the signal under reconstruction. The effectiveness of the proposed algorithm is demonstrated by comparing the channel estimation performances of three algorithms: basis pursuit de-noising, Dantzig selector and orthogonal matching pursuit. The experimental results show that the sparse channel estimation algorithm based on CS is superior to the traditional least square algorithm of channel estimation in terms of accuracy, and can still estimate channel parameters accurately even in the case of least square matrix inverse singular. In relation to the effect of Doppler frequency shift, the direct CS estimation method overcomes the estimation method with compensated CS.