EFFICIENT CHANNEL ESTIMATION IN RIS-ASSISTED SYSTEMS USING DS-COSAMP ALGORITHM

Reconfigurable Intelligent Surfaces(RIS)can change the wireless environment by adjusting the reflection coefficients of various components. However, due to the substantial number of passive components in RIS that lack signal processing capabilities, considerable pilot overhead is necessary for channel estimate in RIS-assisted wireless networks. In this second half of the invited work, we aim to reduce this complexity issue by using the dual coded stickiness of the angular domain cascaded channels used by communication consumers. We specifically observe that these channels exhibit a shared sparsity pattern in their rows (totally mutual pathways) and a partly shared pattern in their columns (user-specific paths). Building on this rationale, we offer the Double Structured Compressive Sampling Matching Pursuit (DS-CoSaMP) method, consistently achieves lower Normalized Mean Squared Error (NMSE) than DS-OMP, especially as the number of users or RIS elements increases. It is more effective than DS-OMP in jointly tracking and updating these structures through its two-stage support update and signal estimation process. The simulation results demonstrate that the overheadreduction proposed by our technique for the pilot is significant compared to conventional methods.