基于深度学习的IRS辅助无线通信系统信道估计OA

Intelligent Reflecting Surface(IRS)can perform certain phase and amplitude transformations on the incoming signals,thus achieving accurate signal transmission and improving signal coverage and transmission efficiency.However,this advantage can only be achieved when accurate Channel State Information(CSI)is known.While due to the passive nature of IRS elements,it is difficult to obtain accurate CSI.Therefore,the Compressive Sensing(CS)algorithm combined with Deep Learning(DL)is used to solve this problem.Firstly,a joint-link structure is used to optimize the traditional compressive sensing algorithms to achieve better Normalized Mean Square Error(NMSE)performance at lower pilot overhead and Signal to Noise Ratio(SNR).Secondly,the channel estimation problem is regarded as a denoising problem,the results obtained by the optimized compressive sensing algorithm are treated as CSI containing noise,and a multiple deep denoising block network is employed to further denoise the CSI to obtain more accurate CSI.Finally,the experimental results demonstrate that the proposed algorithm achieves higher accuracy than the algorithms for comparison at the same SNR.