| 摘要: |
| 在智能反射面(Intelligent Reflecting Surface,IRS)辅助毫米波多用户系统中，针对传统信道估计方案因不完全稀疏性导致性能下降的问题，提出了一种基于多用户公共稀疏结构的3阶段联合估计方案。该方案将基于完全稀疏结构的公共参数作为先验信息，通过对用户联合估计实现性能提升。第1阶段利用用户对级联信道特定结构的共享性，估计公共参数信息。第2阶段提出一种改进的正交匹配追踪算法(Orthogonal Matching Pursuit,OMP)估计用户独有的参数信息作为联合估计的初始值，还引入对支撑集原子的二次筛选，确保用于估计的支撑集最优。第3阶段基于先验信息和初始值对用户级联信道进行联合估计。仿真结果表明，所提方案较现有结构化稀疏方案可以显著减少冗余导频消耗，并且在相同导频开销下，估计精度最大提高3~4 dB。 |
| 关键词: 智能反射面(IRS) 级联信道估计 正交匹配追踪算法 公共稀疏结构 |
| DOI：10.20079/j.issn.1001-893x.231107001 |
|
| 基金项目:重庆市自然科学基金面上项目(cstc2021jcyj-msxmX0454) |
|
| Cascade Channel Estimation for IRS-assistedMillimeter-wave Multi-user Systems |
| LI Guiyong,YU Xiaona,GAO Xinyu |
| (School of Communication and Information Engineering,Chongqing University of Posts and Telecommunications,Chongqing 400065,China) |
| Abstract: |
| In intelligent reflecting surface(IRS)-assisted millimeter-wave multi-user systems,in order to solve the problem of performance degradation caused by incomplete sparsity of the traditional channel estimation scheme,a three-stage joint estimation scheme based on multi-user common sparse structure is proposed.In this scheme,the common parameters based on the completely sparse structure are used as the prior information,and the performance is improved through user joint estimation.Specifically,in the first stage the user’s commonality is used to estimate the common parameter information of the specific structure of the cascade channel.In the second stage,an improved orthogonal matching pursuit(OMP) algorithm is proposed to estimate the user’s unique parameter information as the initial value of joint estimation,the secondary screening of the support set atoms is introduced to ensure that the support set used for estimation is optimal.In the third stage,the joint estimation of the user cascading channel is carried out according to the prior information and the initial value.The simulation results show that the proposed scheme can significantly reduce the redundant pilot consumption compared with the existing structured sparse scheme,and the estimation accuracy can be increased by 3~4 dB at the maximum under the same pilot overhead. |
| Key words: intelligent reflecting surface(IRS) cascade channel estimation OMP algorithm common sparse structure |
