Supervised-Learning-Aided Communication Framework for MIMO Systems With Low-Resolution ADCs

Abstract

This paper considers a massive multiple-inputmultiple-output (MIMO) system with low-resolution analog-todigital converters (ADCs). In this system, inspired by supervised learning, we propose a novel communication framework that consists of channel training and data detection. The underlying idea of the proposed framework is to use the input-output relations of a nonlinear system, formed by a channel and a quantization at the ADCs, for data detection. Specifically, for the channel training, we develop implicit and explicit training methods that empirically learn the conditional probability mass functions (PMFs) of the nonlinear system. For the data detection, we propose three detection methods that map a received signal vector to one of the indexes of possible symbol vectors, according to the empirical conditional PMFs learned from the channel training. We also present a low-complexity version of the proposed framework that reduces a detection complexity by using a successive-interference-cancellation (SIC) approach. In this lowcomplexity version, a symbol vector is divided into two subvectors and then these two subvectors are successively detected using SIC. When employing the proposed framework with one-bit ADCs, we derive an analytical expression for the symbol-vectorerror probability. One major observation is that the symbolvector-error probability decreases exponentially with the inverse of the number of transmit antennas, the operating signal-tonoise ratio, and the minimum distance that can increase with the number of receive antennas. Simulations demonstrate the detection error reduction of the proposed framework compared to existing detection techniques.

Supervised and Semi-Supervised Learning for MIMO Blind Detection With Low-Resolution ADCs

“…The work in [34] showed that if the transmitter employs QAM modulation and the quantization function satisfies Q b (−r) = −Q b (r) ∀r ∈ R, then the length of the training sequence can be reduced to T t = KL t /4. In Proposition 1 below, we generalize this result for any modulation scheme.…”

Supervised-Learning for Multi-Hop MU-MIMO Communications With One-Bit Transceivers

“…There have been increasing research interests in exploiting machine learning tools to address the nonlinearity of a MIMO system with low-resolution ADCs. By treating an end-to-end nonlinear MIMO system with low-resolution ADCs as an autoencoder, a supervised-learning aided communication framework was proposed in [17]. Specifically, it empirically learns the nonlinear channel (i.e., the conditional probability mass functions (PMFs)) by sending pilot symbols (or known data symbols) repeatedly.…”

“…In this section, we propose a novel communication framework for the nonlinear multi-hop MU-MIMO relay channel by harnessing an end-to-end supervised-learning technique. We first present a simple model that can be a good approximation of the complicated nonlinear multi-hop MU-MIMO relay channel by exploiting a coding theoretical framework developed in our prior works [14], [17]. Then, we explain how to learn the model parameters using a simple training strategy and to detect the data symbols using the trained model parameters.…”

“…When using a large number of antennas at the relay and the BS, the fabrication cost and the power consumption significantly increase. To diminish the power consumption and the cost, the use of cheaper and more energy-efficient hardware components including low precision ADCs and DACs has been considered as a promising approach [10]- [17]. Motivated by this approach, this paper focuses on a multi-hop MU-MIMO relay channel, in which information symbols of users are delivered to the BS with one-bit ADCs via layered and distributed relays, each with one-bit transceiver.…”

Deep Learning Based Channel Estimation for Massive MIMO With Mixed-Resolution ADCs

“…In [15], channel correlation is exploited by deep convolutional neural network (CNN) to improve the channel estimation accuracy and to reduce the computational complexity for millimeter wave massive MIMO systems. In [16], a supervised learning based successive interference cancellation is developed for MIMO detection with low-resolution ADCs. More results on DL for physical layer communications can be found in [12].…”