Grant-Free Massive Connectivity in Massive MIMO Systems: Collocated Versus Cell-Free

We consider the problem of user activity detection for grant-free massive connectivity, where a base station, equipped with a large number of antennas, serves a large number of machine type devices. Two massive multiple-input multiple-output (mMIMO) systems are studied, namely, a collocated mMIMO with physically collocated antennas and a cell-free mMIMO with geographically separated antennas. By utilizing the sporadic transmission nature and the asymptotical orthogonality of channel vectors associated with different users, we propose a correlation-based sparse signal recovery scheme to detect active users. Simulation results are provided to evaluate the performance of the proposed method. Our results show that the cell-free mMIMO achieves much better user activity detection performance than the collocated mMIMO, and thus is more suitable for grant-free massive connectivity. An empirical analysis is performed to explain the superiority of the cell-free system over the collocated system.