Expectation Maximization-Based Semi-Blind Channel Estimation for Underwater Acoustic OFDM Communications

In this paper, we propose a semi-blind channel estimation method based on the expectation maximization (EM) algorithm for underwater acoustic orthogonal frequency division multiplexing (OFDM) communications. This method, on the basis of comb-type pilots, can further utilize the power of data symbols to participate in channel estimation, thereby enhancing estimation accuracy at the cost of increasing computational complexity. The proposed method performs channel estimation for each OFDM symbol separately, facilitating the tracking of channel variation. Moreover, it allows non-uniform distribution of pilot symbols and can significantly alleviate the aliasing effect caused by non-uniform discrete Fourier transform (DFT). Neither the sparsity nor other prior knowledge of the channel is necessary for the receiver. Simulation results in the acoustic environment of the South China Sea demonstrate the advantages of our method.