Transceiver Technique in Large-Scale Low-Resolution MIMO Systems

Low-resolution communication systems are attracting significant attention due to their exceptionally low power consumption and cost, enabling the deployment of large numbers of fully digital transmit and receive chains. However, the extreme nonlinearity introduced by such systems necessitates a reevaluation of conventional transceiver techniques. For instance, the sharp transitions on the transmitter side present challenges in confining out-of-band emissions, while the coarse quantization on the receiver side renders many conventional detection schemes ineffective. Therefore, we propose a set of simple yet effective techniques to address these challenges, thereby enabling the advantages of low-power, low-resolution, large-scale transceivers to achieve high data rate communication with significantly improved energy efficiency. Importantly, while achieving the optimal maximum likelihood (ML) performance, which is exponential in complexity with the number of transmitter antennas, the proposed techniques are designed to maintain much lower polynomial computational complexity, which is essential for systems with hundreds or thousands of antennas. We first focus on receiver techniques in one-bit binary/quadrature phase-shift keying systems. When the channel is known at the receiver, one way to simplify the signal detection is to limit the search space to a small list. Inspired by successive interference cancellation, we propose a list creation method based on successive detection, whose performance is close to the optimal ML performance but with only cubic complexity. Moreover, we demonstrate that this method can also work in higher-resolution systems and in on-off keying systems, thus implying its broader applicability. When the channel is unknown, a simple expression of the minimum mean square error estimator of the channel is derived. In this case, in order to limit the search space, we propose a correlation-based list creation method. Together with a slightly modified iterative detection and decoding process, we demonstrate that this receiver technique enables operation within 2-4 dB of capacity with hundreds of transmitter antennas. Finally, for the transmitter side, we investigate transmission techniques under spectrum constraints. Specifically, the maximum entropy rate of the transmitted signal subject to the constraint that the fraction of power outside the allocated frequency band is smaller than a given threshold is considered, where the trade-off between the resolution of the digital-to-analog converter and the symbol rate is studied, and a lower bound of the entropy rate is derived. We further propose coding schemes based on a polar code that can shape the spectrum efficiently and effectively.