Adaptive Grouped Physical Layer Multicast and Beamforming for Massive MIMO

In this paper, we consider an adaptive grouped physical layer multicasting for large-scale multi-antenna wireless systems in which a set of users are divided into several groups and each user group are assigned with a unique beamforming vector for multicast transmissions on the orthogonal resources. Based on the adaptive grouped multicast framework, we consider the joint design of user grouping and multicast beamforming adapted to the user channels. Two design objectives are studied, i.e., the average-rate maximization and the max-min fairness. We propose an iterative user grouping and beamforming design method for both optimization objectives. For iterative user grouping, we present a method for selecting better initial grouping centers. Moreover, to overcome the issue of converging to a local optimum for the iterative approach, we propose a novel enhancement scheme via user grouping perturbation, which performs very close to the exhaustive grouping search. Simulation results demonstrate the efficacy of the proposed designs.