Random Matrix Spreading for Unsourced Random Access

Unsourced random access (URA) has broad prospects as a technology that meets the massive access requirements of massive machine type communication scenarios. This paper first establishes a URA system based on random spreading. To overcome the limitation of channel use imposed by the length of spreading codes, a method utilizing random matrices instead of spreading sequences for symbol spreading is designed. Bundling and spreading the adjacent symbols of the data sent by each user, the effective spreading length of each symbol becomes several times of the original spreading code length, while preserving the total number of channel uses. Simulation results demonstrate that using matrix-based spreading can effectively improve the active user detection performance, and enhance the data detection performance to a certain extent.