Underwater acoustic Orthogonal Frequency Division Multiplexing(OFDM)systems suffer from frequency-selective fading and inter-carrier interference(ICI)caused by multi-path delay spread and Doppler effect respectively in underwater acoustic channels. Conventional receiver algorithms require to carry out complex two-dimensional(delay/Doppler)channel estimation，followed by complex channel equalization and symbol detection. Targeting at the characteristics of slow change of Doppler matrix elements and limited ICI interference range in the channels with limited Doppler spread，this paper draws on the polynomial fitting idea and proposes an innovative receiver algorithm. The algorithm approximates the Doppler matrix elements with polynomial functions， and iteratively performs polynomial fitting and data symbolic frequency-domain equalization. Compared with traditional methods，the algorithm does not require Doppler estimation of channels. Instead，it only requires one-dimensional(time delay)estimation of channels to achieve good performance. The feasibility of the algorithm is verified by the simulation results of underwater acoustic channels generated by Bellhop.
