基于TS-OMP的IRS辅助毫米波大规模MIMO信道估计方案

打开文本图片集

中图分类号： TN929.5 文献标志码：A DOI： 10.12305/j.issn.1001-506X.2026.03.32

Abstract：To address thehigh pilot overhead in channel estimation for inteligent reflecting surface（IRS）assisted large-scale multiple-input multiple-output （MIMO）milimeter-wave systems， a channel estimation scheme based on the triple-structured orthogonalmatching pursuit （TS-OMP）is proposed.By exploiting the angular spread characteristicsof the channel，the triple-structured sparsityof the beamspace cascaded chanel matrix isrevealed.Leveraging this structured sparsity，the problem of channel estimation is reformulated asa problem of triple-structured sparse signal recovery.The proposed TS-OMP algorithmaddresses this by first estimating the sparse row indices and intra-cluster column offsets for a subsetofusers.Theseestimated indicesandofsetsarethenutilizedaspriorinformation toestimatethechannel state information for allusers.Simulationresults demonstrate thatthe performanceand pilot overheadof thisscheme are significantly better than existing approaches.

Keywords： channel estimation； intelligent reflecting surface； compressed sensing；millimeter wave；angularspread

0引言

智能反射面（intelligentreflectingsurface，IRS）辅助的通信系统作为一种密集部署且低成本的无线连接方式，受到了广泛的关注[1-2。（剩余13352字）