Spectral and Energy Efficiency of Massive MIMO for Hybrid Architectures Based on Phase Shifters

Abstract

This paper investigates the downlink achievable spectral efficiency (SE) and energy efficiency (EE) of massive MIMO for hybrid architectures based on phase shifters, where the base station (BS) has perfect channel state information and baseband processing is done zero-forcing precoding. We derive an approximated upper bound on the achievable SE for hybrid architecture with ideal phase shifters. Based on the derived analytical expression, we find that the total achievable SE increases with the number of BS antennas and users and the signal-to-noise ratio (SNR). In order to acquire the required limited feedback, we propose an algorithm to generate the corresponding quantized matrix and study hybrid architectures with quantized phase shifters. Compared to full-digital architectures, results show that hybrid architectures enjoy a much higher achievable EE in massive MIMO systems, whereas its achievable SE is inferior to full-digital architectures. In addition, results also showcase that the achievable SE of hybrid architectures can be improved by increasing the bits of phase shifters, and there exists an optimal SNR and antenna number to maximize the achievable EE of the system.