Are traditional signal processing techniques rate maximizing in quantized SU-MISO systems?

In this contribution, we provide an information theoretical analysis of coarsely-quantized downlink Single-User (SU)-Multiple Input Single Output (MISO) communication systems. We address the question of whether traditional signal processing techniques, i.e., proper signaling and channel rank transmit covariance matrices, are still optimal with respect to maximizing the data rate. We investigate the mutual information lower bound based on the Bussgang theorem, in the SU-MISO downlink scenario, where we assume 1-bit quantized Digital-To-Analog Converters (DACs) in the transmit antennas at the Base Station (BS). We prove that at low Signal-To-Noise Ratio (SNR), existing signal processing techniques maximize the data rate. However, at higher SNR we show, using counter examples, that the data rates can be improved using different signal processing techniques. These results show the potential merit of reconsidering signal processing techniques in coarsely-quantized SU-MISO downlink scenarios.