Inter-block iterative sparse channel estimation method for long-range OFDM underwater acoustic communication in deep sea

In long-range communication of deep sea, underwater acoustic channel is characterized by long delay spread and serious frequency selective fading. During estimation of Orthogonal Frequency Division Multiplexing (OFDM) system in this channel, traditional block-independent Compressed Sensing (CS) algorithm need a larger number of pilot symbols to ensure acceptable estimation performance, which would lead to low spectrum efficiency. An inter-block iterative channel estimation method based on sparse time-varying modeling was proposed to deal with this problem. The temporal correlation features of deep-sea channel are exploited to establish iterative channel estimation model between adjacent data blocks, and based on this model, a number of elements are cut from the candidate dictionary matrix of traditional sparse channel estimation algorithm and the optimization equation is modified. In this process, the previous estimated channel could offer useful prior information for the estimation of current block, which would significantly decrease the number of pilot symbols. To validate the effectiveness of this method, an underwater acoustic communication experiment was carried out at different receiving depths and distances of deep sea. The experimental results show that this proposed method could get better estimation performance than traditional sparse estimation method with only half the pilot inserted.