基于贝叶斯分析的IRS辅助通信系统信道估计技术的研究

智能反射平面(IRS)辅助技术可实现配置无线传播环境的目的，从而极大提升了无线频谱利用率和能量效率。信道估计是IRS辅助通信系统工作的基础，对通信系统的性能起着至关重要的作用。然而，现有信道估计方法较少能充分挖掘IRS辅助通信系统信道的稀疏特性及非理想模型特征，在应对复杂的无线通信场景时，不可避免地存在信道估计性能的损失。本项目以完善和丰富复杂场景下IRS辅助通信系统的信道估计技术为目标，其主要包括：建立多种广义稀疏概率模型，在贝叶斯分析框架下实现多稀疏特性融合的IRS辅助通信系统信道估计技术，大幅提高贝叶斯分析方法的精确度；建立多种与实际IRS辅助通信系统相关的误差模型，实现复杂场景下稳健的信道估计技术，有效改善贝叶斯分析方法的的鲁棒性；提出若干低计算复杂度的信道估计方法，并从反射系数信息加载角度提出若干联合信息恢复和信道估计方法，显著降低IRS辅助通信系统信道估计方法的应用复杂度。

The intelligent reflecting surface (IRS) aided technique can configure the wireless communication environment to greatly improve the wireless spectrum utilization and energy efficiency. Channel estimation is a fundamental problem for the IRS-aided communication systems, which largely determines the performance of communication systems. However, the existing channel estimation methods are unable to fully exploit the channel sparsity property of the IRS-aided communication systems, as well as the non-deal model characteristics. Therefore, the channel estimation performance for the IRS-aided communication systems may degrade substantially in complex wireless communication scenarios. This project aims to improve and enrich the channel estimation technology for the IRS-aided communication system in complex scenes. Firstly, we try to provide a variety of generalized sparse probability models for the IRS-aided communication systems, and then establish a generalized Bayesian analysis framework for channel estimation under the multi-sparse feature fusion, so as to enhance the accuracy performance. In order to improve the robustness of Bayesian analysis method, a variety of error models related to the actual IRS-aided communication systems will be given to implement robust channel estimation technology in complex scenes. Finally, some low computational complexity channel estimation methods are proposed, as well as some low application complexity channel estimation methods using the information loading with the reflection coefficients, so as to significantly reduces the computational complexity and application complexity of the channel estimation methods for the IRS-aided communication systems.