A Learned Denoising-Based Sparse Adaptive Channel Estimation for OTFS Underwater Acoustic Communications
Lianyou Jing, Qingsong Wang, Chengbing He, Xuewei Zhang
Abstract
This letter proposes a learned denoising-based sparse adaptive channel estimation method in the delay-Doppler domain for time-varying underwater acoustic (UWA) channels in an orthogonal time-frequency space (OTFS) system. We first propose a symbol-wise adaptive channel estimation method for the OTFS system. By leveraging the sparsity characteristic of the channels, we employ the improved proportionate normalized least mean squares (IPNLMS) algorithm. Based on the characteristic that the channel in the delay-Doppler domain is invariant, the multiple estimates obtained from the adaptive filter could be regarded as multiple noisy images derived from the same clean image. A neural network called FastDVDNet, commonly used in video denoising, is utilized to exploit the correlation among the multiple images. The simulation results demonstrate that the proposed denoising strategies significantly enhance the estimation performance, thereby achieving superior channel estimation results.