Parameter Estimation of Delay-Doppler Underwater Acoustic Multi-Path Channel Based on Iterative Fractional Fourier Transform

The characterization of wideband underwater acoustic (UWA) multi-path signals is investigated in this paper. The UWA multi-path channel can be considered as a sparse model and it causes changes of the transmitted signal on Doppler factor, time delay, and amplitude, thus the signal component from each path can be differentiated from the set of parameters. This paper proposes an FrFT-based method to estimate the channel parameters, which can decompose and retrieve the multi-path signals from the UWA channels. The proposed algorithm proceeds in an iterative manner and returns a set of estimated parameters at each turn. The estimated time delay is obtained directly without any extra correlation computation. Then, the estimated component is extracted by eliminating its FrFT result from the fractional domain. Compared to the existing methods, the most improvement is that the method can be applied to the wideband LFM signals with a wider frequency range in low signal-to-noise environment. Because the indirect subtractions avoid the bad chain effect caused by the low resolution or estimation error. The effectiveness of the method is confirmed by several simulations with low and high frequency. The performance is compared with a matching pursuit-based algorithm and an FrFT-based algorithm. It is shown that the proposed method outperforms the existing algorithms in low-frequency multi-component separation as well as estimation accuracy.