On the MISO Broadcast Channel With Finite Constellations and Imperfect CSIT: Rate-Splitting, SDMA, NOMA, and Space-Time Block Coding

Abstract

The multiple-input single-output broadcast channel (MISO-BC) models a scenario in which a transmitter equipped with multiple antennas communicates with multiple single-antenna receivers over the same time, frequency, and code resource block. Strategies for managing inter-user interference in this setting are typically studied under the assumptions of either Gaussian signaling (GS) and/or perfect channel state information at the transmitter (CSIT). In this paper, we consider a more realistic model for practical wireless communications where the transmitted signals are drawn from finite input constellations, referred to as discrete signaling (DS), and where the transmitter has imperfect CSIT. Well-known multiple access (MA) strategies for the MISO-BC based on linear precoding, namely rate-splitting multiple access (RSMA), space-division multiple access (SDMA), and power-domain non-orthogonal multiple access (NOMA), are studied under this formulation and compared with space-time block coding (STBC) techniques that exploit similar concepts for managing inter-user interference, while relying solely on statistical magnitude CSIT. The precoding and decoding formulations of each strategy are described within a unified framework, and key differences between GS and DS in managing inter-user interference are identified. Furthermore, a novel space-time RSMA (ST-RSMA) design is proposed, and the previously considered ST-NOMA scheme is generalized for more than two transmit antennas. Extensive numerical results under Rayleigh block-fading channels, based on symbol error rate (SER) and bit error rate (BER) Monte Carlo simulations, are presented across various scenarios, including different numbers of antennas, transmission rates, CSIT quality levels, and channel strength disparities among users. Based on these results, the strengths and weaknesses of each strategy are identified, and the most suitable approach for each scenario is determined.