Spectral Efficiency Analysis for Massive MIMO System Under QoS Constraint: an Effective Capacity Perspective

TLDR: This paper analyzes the achievable effective SE of a massive MIMO system under the probability constraint of delay violation by adopting the effective capacity (EC) theory of wireless channels and demonstrates how the optimal number of scheduled users changes for different QoS constraint levels.

Abstract: In massive multiple-input multiple-output (MIMO) systems, dozens of mobile users can simultaneously receive signals from one base station. To obtain maximum spectral efficiency (SE), researchers have investigated the optimal number of scheduled users for one time slot. However, in practical cases, we must consider the quality of service (QoS) constraint. The probability of delay violation is an important QoS index that depends on transmission stability over a long period rather than the instantaneous transmission rate of one time slot. In this paper, we analyze the achievable effective SE of a massive MIMO system under the probability constraint of delay violation. By adopting the effective capacity (EC) theory of wireless channels, we associate the delay violation probability with the transmission rate fluctuations caused by the massive MIMO scheduling strategy used. The relationship between the effective SE, the QoS constraint, and the number of scheduled users is formulated as a continuous function. Our simulation results demonstrate how the optimal number of scheduled users changes for different QoS constraint levels. According to the changing trend, the massive MIMO system can be programmed to choose between different simple scheduling strategies for different QoS constraint levels.