基于RIS子结构的压缩感知多径抑制直接定位算法OA

To address the limited positioning accuracy of reconfigurable intelligent surface(RIS)-aided system in mul-tipath environments,a high-precision localization algorithm based on RIS substructures was proposed.A single RIS was divided into multiple substructures,and time-sequential orthogonal coding was applied to each substructure,enabling the reflected signal components from different substructures to be accurately separated at the receiver through orthogonal de-coding.In this way,a single RIS was made equivalent to multiple virtual anchor points,and the spatial observation di-mension of the system was significantly improved under a single RIS and single antenna configuration.On this basis,a compressed-sensing-based direct localization framework was constructed using an angle-domain signal model.An obser-vation dictionary was established according to the geometric mapping relationship between candidate positions and the substructures,and the target was directly estimated in the position domain,thereby avoiding the cascaded errors intro-duced by the estimation of intermediate parameters,such as the angle of arrival(AoA),in conventional two-step meth-ods.By jointly exploiting the consistency constraint of the target position and the sparsity of multipath components,the target path was effectively distinguished from multipath interference.Simulation results show that the proposed algo-rithm exhibits good robustness in complex multipath environments,achieves sub-meter-level positioning accuracy,and provides a new solution for low-cost and high-precision positioning aided by RIS.