The Performance Of Box-Relaxation Decoding In Massive MIMO With Low-Resolution ADCS

We study massive multiple-input-multiple-output (MIMO) uplink systems with low-resolution analog-to-digital converters (ADCs) on each receiver antenna. For the decoding, we propose an efficient convex program, namely the boxrelaxation optimization (BRO), for which we further characterize its asymptotic bit-error rate (BER) under Gaussian channel and noise models. The derived formula is an explicit function of the ratio of receive to transmit antennas, of the SNR, and of the specifics of the quantization scheme. Hence, it can be used for tuning of these parameters such that the desired BER specifications are met. We further show the superiority of the BRO compared to standard linear detectors such as the zero-forcing (ZF) decoder. Numerical simulations corroborate our theoretical results.