Receiver Design for Asynchronous Unsourced Random Access in Cell-Free Systems

Cell-free techniques and unsourced random access (URA) are effective in improving the capacity to accommodate massive user equipments (UEs). However, the variable transmission distances between UEs and access points in cell-free systems lead to inevitable asynchronous transmission, which would decrease URA performance. To address this, this paper introduces an asynchronous URA receiver for cell-free systems. The approach begins with expanding the URA codebook to account for phase shifts due to various delays, transforming the codeword detection problem in URA under unknown delays into a sparse signal recovery problem in compressed sensing. Further, we leverage a coordinate descent algorithm complemented by a convolutional neural network-based soft decoder to detect transmitted codewords efficiently. Simulation experiments validate the efficiency in handling asynchronous transmission challenges of the proposed asynchronous URA receiver in cell-free systems.