A Newtonized Approach for Estimation of Doubly Spread Acoustic Channels

Underwater acoustic communication systems face considerable challenges due to the joint distortion of delay and Doppler in time-varying channels. Traditional grid-based channel estimation methods suffer from a trade-off between resolution and computational complexity. In this paper, we propose a Newtonized off-grid parameter estimation approach with second-order optimization techniques to address this challenge. By modeling the Doppler estimation part as a generalized likelihood maximization, we derive the closed-form for the first- and second-order derivatives of the cost function. This allows for Newton iterations to refine coarse grid estimates to sub-grid precision with minimal computational overhead. Numerical simulations verify that the proposed method performs better in root-mean-square-error(RMSE) compared to traditional grid-based methods.