A Joint Channel and Queue Aware Scheduling Method for Multi-User Massive MIMO Systems

Massive multiple input and multiple output (massive MIMO) has attracted wide attention since it holds the promise of providing excellent spectral efficiency. One of the basic issues of single-cell multi-user massive MIMO system is how to design efficient scheduling methods under the quality-of-service (QoS) requirements. In this paper, we present a joint channel and queue aware scheduling method based on Lyapunov optimization theory. Specifically, the channel gain can be estimated by default according to the characteristic of channel hardening so that we can jointly optimize the user selection and power allocation. What remains to be done is only to measure the direction between the base station and the active users, which is in contrast to the conventional process. To this end, we formulate a cross-layer control problem, which is a mixed integer nonlinear programming problem. To obtain the optimal solutions to this problem efficiently, we present a two-layer iteration algorithm. The inner layer is a dynamic programming algorithm based on water-filling in cellars and the outer layer is a low-complexity one-dimensional search algorithm.