DSRNet: Hybrid Deep Learning-Based Channel Estimation for RIS-Aided Wireless Communication

EasyChair Preprint 15751

Abstract

Reconfigurable intelligent surfaces (RIS) are widely perceived as a transformative technology for 5G and beyond, enabling dynamic programming of wireless propagation channels. However, acquiring accurate channel state information (CSI) remains a major challenge in RIS-assisted wireless communication systems. Most existing research assumes the availability of full CSI, which is often impractical due to the passive nature of the RIS elements and the high-dimensional nature of the channels. To fill this gap, we introduce a two-stage framework, called the denoising super-resolution network (DSRnet), to estimate full CSI from partial representations. Then the estimated full CSI is utilized to maximize the weighted sum-rate (WSR) via phase shift prediction. DSRnet employs a hierarchical architecture consisting of a super-resolution sub-network for initial estimation, followed by a denoising sub-network enhanced with spatial attention modules for refined processing. The proposed model achieves impressive channel estimation performance with an NMSE of −13.14 dB while maintaining computational efficiency, utilizing only 80, 916 parameters. When the estimated full CSI is used for phase shift prediction, it shows an approximately 15% improvement in WSR compared to the methods with partial CSI. These findings verify DSR- net as a practical and efficient solution for large-scale RIS deployments, effectively balancing high performance with reduced CSI overhead.

Keyphrases: Channel State Information, Denoising, Reconfigurable Intelligent Surfaces, deep learning, super-resolution

@booklet{EasyChair:15751,
  author    = {Syed Samiul Alam and Haolin Tang and Guang Yang and Changqing Luo and Wei Wang and Yanxiao Zhao},
  title     = {DSRNet: Hybrid Deep Learning-Based Channel Estimation for RIS-Aided Wireless Communication},
  howpublished = {EasyChair Preprint 15751},
  year      = {EasyChair, 2025}}