Precoding And Scheduling In Massive MIMO Systems

Massive MIMO bears remarkable potential to provide huge spectral-efficiency and energy-efficiency gains, and has become a hot topic in both industry and academia as one of the major candidate technologies for 5G. At the time of writing, the topic of user scheduling in massive MIMO systems is widely open. Aiming to exploit the full potential of massive MIMO, we focus on low-complexity user scheduling schemes in this paper. First, concerning the characteristics of massive MIMO, the sum rate performance of massive MIMO systems is anal-ysed in various scenarios by leveraging RMT tools. Then, with the help of ana-lytical results, user scheduling in massive MIMO systems is investigated from a new perspective. As a promotion of 5G, several user scheduling schemes for massive MIMO systems are proposed in this paper, with which we can signifi-cantly enhance the sum rate performance with very low complexity.The main innovations and contributions of the paper are as follows.1) We propose several low-complexity user scheduling schemes for mul-tiuser massive MIMO downlink systems. First, approximations of the sum rates of massive MIMO systems with ZF precoding are derived, which only rely on the system parameters (i.e., the number of base station antennas, the transmit power) and statistical channel information (i.e., the statistical distribution of channel matrix eigenvalues), independent of instantaneous CSI. Based on the approximate expressions, the optimal number of active users for maximizing the approximate sum rate is studied. Then, upon exploiting the optimal num-ber, a pair of low-complexity user scheduling schemes to improve the system sum rates are introduced. Both of the proposed schemes achieve significant sum rate improvement over the RUS scheme, while enjoying the same extremely-low computational complexity as that of the conventional RUS. In addition, a low-complexity transmission mode selection scheme is also proposed based on the analytical results, with which the system sum rate performance is improved through the joint design of precoding approach selection and user scheduling.2) For multi-cell massive MIMO systems, we also propose a low-complexity user scheduling scheme. Different from traditional interference management methods relying on complicate matrix computation, we introduce AAS into multi-cell massive MIMO systems to reduce interference through vertical spa-tial domain processing. First, all the users are divided into groups according to the vertical beam width of the base station antennas. Each scheduling slot is also divided into the same number of sub-slots. Then, benefiting from the RMT tools, sum rate approximations are derived, and the optimal number of active users and the optimal tilting angle for each group are found. At each scheduling sub-slot, one UE group is considered. The optimal number of users for simul-taneous data transmissions is selected, and the base station antenna tilting is adjusted to the optimal angle. Moreover, we develope a user selection scheme consisting of transmit beamforming codebook and threshold-based feedback.3) We discuss the relay power allocation and user scheduling in massive MIMO AF relay systems, in which the relay is equipped with large-scale an-tenna arrays and performs linear signal processing. Tight and tractable sum rate approximations for ZF and MMSE relay systems are first derived by em-ploying RMT, especially FPT tools. Upon exploiting the approximations, a re-lay power allocation method based on the LSF CSI is developed, avoiding the complicated matrix calculations in the conventional power allocations. Then, low-complexity user pair scheduling schemes are proposed, in which the system sum rate performance is enhanced through adjusting the number of active user pairs. Furthermore, we also discuss the collaborative design of relay antenna tilting with user scheduling in massive MIMO AAS relay systems.