Trade-offs in Cell-free Massive MIMO Networks: Precoding, Power Allocation and Scheduling

Abstract

Cell-free Massive MIMO (CF-M-MIMO) networks have recently emerged as a promising architectural solutions to satisfy the requirements of future networks (i.e., beyond 5G, 6G). The CF-M-MIMO paradigm advocates for the irregular deployment of a large number of access points (AP) throughout the network coverage area, all connected to a central processing unit (CPU), with the aim of bringing the radio access front-end closer to the users. Indeed, the cell-free topology can be interpreted as a fully distributed implementation of the Massive MIMO (M-MIMO) technology that is currently permeating the rollout of 5G. Interestingly, the large body of theoretical results derived for M-MIMO over the last decade can be recast in the CF-M-MIMO framework yet the distributed nature of the system needs to be carefully factored in. This paper aims at highlighting the different trade-offs affecting various performance metrics in CF-M-MIMO networks. In particular, the influence and consequences of using different precoding strategies, various power allocation techniques and pilot allocation strategies will be discussed and assessed. Moreover, a novel technique, termed user-subset selection, is shown to significantly improve the performance of maxmin power allocation at the cost of penalty in the form of an outage probability.