Deep Learning-Based Low-Complexity Channel Impulse Response Estimation for Resource Management of Underwater Communication Network

This paper proposes a deep learning-based low-complexity channel impulse response (CIR) estimation method using a gated recurrent unit-convolutional neural network (GRU-CNN) based CIR estimator that estimates a CIR without feedback from a receiver. The proposed method estimates CIR using only SSP and seafloor information with low complexity using GRU and a one-dimensional convolutional layer. The simulation results show that the proposed method has the lowest mean squared error (MSE) of CIR estimation compared to other deep learning models and lower complexity than the BELLHOP estimator.