针对智能超表面(Reconfigurable Intelligent Surface,RIS)辅助通信系统中无源反射阵列所导致的信道估计导频开销大的问题,提出了一种低导频开销的压缩感知(Compressive Sensing,CS)信道估计算法.根据毫米波信道的特征,分析级联信道所存在的三重结构稀疏性,即所有用户角级联信道的公共行稀疏性、不同非零行之间非零元素偏移相同和所有用户不同非零行之间列偏移相同.根据毫米波信道的稀疏结构,对现有的正交匹配追踪(Orthogonal Matching Pursuit,OMP)算法进行改进,通过设计合理的判决准则减少噪音所带来的误差扩大.仿真结果表明,在实现相同的信道估计精度时,所提算法在低信噪比(Signal to Noise Ratio,SNR)条件下比现有方案所需的导频开销要低.
Reconfigurable Intelligent Surface Channel Estimation Based on Compressive Sensing
To solve the problem of high pilot overhead caused by passive reflecting arrays in Reconfigurable Intelligent Surface(RIS)-assisted communication systems,a low pilot overhead channel estimation algorithm based on Compressive Sensing(CS)is proposed.According to the characteristics of the millimeter-wave channel,the triple structure sparsity of the cascaded channel is analyzed,such as the common row sparsity of all user angular cascaded channels,the same non-zero element offset between different non-zero rows and the same column offset between non-zero rows of all users.Based on the sparse structure of the millimeter-wave channel,the existing Orthogonal Matching Pursuit(OMP)algorithm is improved,and reasonable judgment criteria are designed to reduce the error expansion caused by noise.Simulation results show that the channel estimation pilot overhead of the proposed algorithm is lower than that of the existing schemes under the same channel estimation accuracy in the low Signal to Noise Ratio(SNR)conditions.
万方数据
维普
