Capacity maximizing in massive MIMO with linear precoding for SSF and LSF channel with perfect CSI

Abstract

The capacity of a massive MIMO cellular network depends on user and antenna selection algorithms, and also on the acquisition of perfect Channel State Information (CSI). Low computational cost algorithms for user and antenna selection significantly may enhance the system capacity, as it would consume a smaller bandwidth out of the total bandwidth for downlink transmission. The objective of this paper is to maximize the system sum-rate capacity with efficient user and antenna selection algorithms and linear precoding. We consider in this paper, a slowly fading Rayleigh channel with perfect acquisition of CSI to explore the system sum-rate capacity of a massive MIMO network. For user selection, we apply three algorithms, namely Semi-orthogonal user selection (SUS), Descending Order of SNR-based User Scheduling (DOSUS), and Random User Selection (RUS) algorithm. In all the user selection algorithms, the selection of Base Station (BS) antenna is based on the maximum Signal-to-Noise Ratio (SNR) to the selected users. Hence users are characterized by having both Small Scale Fading (SSF) due to slowly fading Rayleigh channel and Large-Scale Fading (LSF) due to distances from the base station. Further, we use linear precoding techniques, such as Zero Forcing (ZF), Minimum Mean Square Error (MMSE), and Maximum Ratio Transmission (MRT) to reduce interferences, thereby improving average system sum-rate capacity. Results using SUS, DOSUS, and RUS user selection algorithms with ZF, MMSE, and MRT precoding techniques are compared. We also analyzed and compared the computational complexity of all the three user selection algorithms. The computational complexities of the three algorithms that we achieved in this paper are O(1) for RUS and DOSUS, and O(M2N) for SUS which are less than the other conventional user selection methods.