Coordinated Massive MIMO with Power Control: An Efficient Solution for MU & SU-MIMO Scheduling

In this paper we propose an optimal strategy to do rank adaptation, power control and user selection (pairing) for both Single-User (SU) & Multi-User (MU) MIMO in Massive MIMO (MM) systems. In this approach, we redefine the user selection and rank adaptation problem as a power control problem. We then use the Network Utility Maximization (NUM) framework to optimize the power vector. The benefit is that user selection and rank adaptation problems are typically centralized, combinatorial and thus NP hard, while power control problems are typically distributed and only linear in complexity. The Coordinated Massive MIMO with Power Control (CMM-PC) solution proposed in this paper only operates on a cell level, it extends naturally to a network level. As such, we currently do not require inter cell information exchange. Simulations show a continuously increasing performance gain as the number of antennas are increased over the range from 8 to 64.