Machine Learning-Based Channel Prediction in Massive MIMO With Channel Aging

Abstract

To support the ever increasing number of devices in massive multiple-input multiple-output (mMIMO) systems, an excessive amount of overhead is required for conventional orthogonal pilot-based channel estimation schemes. To circumvent this fundamental constraint, we design a machine learning (ML)-based time-division duplex scheme in which channel state information (CSI) can be obtained by leveraging the temporal channel correlation. The presence of the temporal channel correlation is due to the stationarity of the propagation environment across time. The proposed ML-based predictors involve a pattern extraction implemented via a convolutional neural network, and a CSI predictor realized by an autoregressive (AR) predictor or an autoregressive network with exogenous inputs recurrent neural network. Closed-form expressions for the user uplink and downlink achievable spectral efficiency and average per-user throughput are provided for the ML-based time division duplex schemes. Our numerical results demonstrate that the proposed ML-based predictors can remarkably improve the prediction quality for both low and high mobility scenarios, and offer great performance gains on the per-user achievable throughput.

Convolutional-Type Neural Networks for Fading Channel Forecasting

“…Hansen et al (1999) [42]) or machine learning-based clustering or classification schemes (e.g. Ahrens et al (2019) [43] or Yuan et al (2020) [25]), so that the forecasts can be delivered by combining outputs from predictors that are trained for the most closely matching channel classes. This idea resembles that of [25] where a CNN is used as a channel classifier that assigns a pre-trained predictor with matching channel characteristics to the observation series.…”

“…On the issue of fading channel prediction, standard predictive neural networks with full connections have been evaluated in various contexts in single-subcarrier settings, e.g. Ding and Hirose (2014) [21], Liao et al (2018) [22], Jiang andSchotten (2019, 2020) [23], [24], and Yuan et al (2020) [25], with [25] employing an extra CNN channel classifier to identify patterns in the autocorrelation function of the channel prior to applying the actual predictor. Overall, while the Kalman filtering scheme is based on linear models and parametrised probability assumptions, predictive neural networks incorporate nonlinearities and no specific interpretation (such as Kalman gain or conditional state variance) is imposed on their model parameters.…”

“…In [25], Yuan et al present a channel forecasting scheme consisting of a channel classifier followed by a pre-trained predictor. While the article mainly focuses on the channel classification part (using a CNN) along with the supporting channel estimation scheme, the included predictors are interesting in their own right and can be evaluated as singlesubcarrier predictors on the dataset used in this paper.…”

“…While the article mainly focuses on the channel classification part (using a CNN) along with the supporting channel estimation scheme, the included predictors are interesting in their own right and can be evaluated as singlesubcarrier predictors on the dataset used in this paper. The two proposed predictors in [25] employ an AR model on the one hand and a modified RNN forming a NARX model as introduced in Lin et al (1996) [44] on the other hand. While the AR approach using Kalman filtering scheme has already been considered above, the NARX RNN is adapted to the present setting as described below.…”

Deep Learning-Based Secure MIMO Communications with Imperfect CSI for Heterogeneous Networks

“…Modulation recognition algorithm was considered in [27], where a convolutional neural network followed by a long short-term memory as the classifier was adopted to improve the robustness for modulation recognition. A CNN-based channel prediction scheme was designed in [28] in massive MIMO systems under channel aging effects, where an autoregressive network was used to model the temporal channel correlation of the wireless channel. Considering perfect free space optical communications, a pilot independent deep learning-based channel estimator was proposed in [29,30].…”