Channel Estimation Method for OTFS Systems Based on ResUNet Network

To address the challenges of inter-subcarrier interference (ICI) and inter-symbol interference (ISI) in communication systems, this paper proposes a deep learning-based channel estimation network, namely ResUNet. ResUNet is a novel neural network architecture that integrates U-Net with ResNet (Residual Network). By leveraging multiple convolutional layers and residual connections, the network effectively extracts deep-level features from the channel, ultimately reconstructing the channel matrix and obtaining full channel state information (CSI). Simulation results demonstrate that, in high-mobility scenarios, the proposed channel estimation method significantly outperforms conventional approaches, achieving high-precision channel estimation.