Neural Network MIMO Detection for Coded Wireless Communication with Impairments

@article{Sholev2020NeuralNM,
  title={Neural Network MIMO Detection for Coded Wireless Communication with Impairments},
  author={Omer Sholev and Haim H. Permuter and Eilam Ben-Dror and Wenliang Liang},
  journal={2020 IEEE Wireless Communications and Networking Conference (WCNC)},
  year={2020},
  pages={1-8}
}
In this paper, a neural network based Multiple-Input-Multiple-Output (MIMO) algorithm is presented. The algorithm is specifically designed to be integrated in a coded MIMO-OFDM system, and is based upon projected gradient descent iterations. We combine our model as a part of a modern coded MIMO-OFDM system, and we compare its performance with common MIMO detectors on simulated data, as well as on field data. We also investigated our model’s performance in the presence of several common… 
View on IEEE
ee.bgu.ac.il
5 Citations
Highly Influential Citations
1
Background Citations
1
Methods Citations
1
Results Citations
1

5 Citations

A Study on MIMO Channel Estimation by 2D and 3D Convolutional Neural Networks

This paper investigates the performance of various CNN architectures for a diverse data set generated according to 5G NR standard, and in particular the influence of spatial correlation, Doppler and reference signal resource allocation.

Data-Driven Precoded MIMO Detection Robust to Channel Estimation Errors

Numerical results show that severe CSIT degradation impedes the correct detection when a conventional detector is used, but the proposed learning-based detectors can achieve good detection performance even under severeCSIT deterioration, and can yield 4–8 dB power gain for BER values lower than 10−4 when compared to the classic linear minimum mean square error (MMSE) detector.

Low PAPR Waveform Design for OFDM Systems Based on Convolutional Autoencoder

This paper introduces the architecture of a convolutional autoencoder (CAE) for the task of peak-to-average power ratio (PAPR) reduction and waveform design, for orthogonal frequency division…

Joint high-dimensional soft bit estimation and quantization using deep learning

EQ-Net is introduced: a deep learning approach for joint soft bit estimation and quantization in high-dimensional multiple-input multiple-output (MIMO) systems and is robust to distributional shifts when used for soft bit quantization and is competitive with state-of-the-art deep learning approaches when faced with channel estimation errors insoft bit estimation.

EQ-Net: A Unified Deep Learning Framework for Log-Likelihood Ratio Estimation and Quantization

EQ-Net is introduced: a data-driven framework that solves both tasks of log-likelihood ratio (LLR) estimation and quantization and achieves state-of-the-art results on both tasks when compared to previous methods.

References

SHOWING 1-10 OF 36 REFERENCES

A Model-Driven Deep Learning Network for MIMO Detection

Numerical results show that the proposed approach can improve the performance of the iterative algorithm significantly under Rayleigh and correlated MIMO channels.

Deep Mimo Detection Using ADMM Unfolding

A low-complexity deep neural network (DNN)based multiple-input-multiple-output (MIMO) detector for the BPSK and QPSK constellation cases is presented and it is shown that for a 160×160 MIMO system, the DNN design can attain nearly optimal bit-error rate performance.

Efficient MIMO Detection with Imperfect Channel Knowledge - A Deep Learning Approach

A deep learning based efficient MIMO detection approach using a neural network to directly get a mapping function of received signals, channel matrix and transmitted bit streams and shows that better robustness can be achieved in condition of imperfect channel knowledge compared with conventional algorithms.

A Partial Learning-Based Detection Scheme for Massive MIMO

Theoretical analyses and simulation results showed that the proposed PL-based detection scheme can achieve low BER with low computational complexity.

Implementation Methodologies of Deep Learning-Based Signal Detection for Conventional MIMO Transmitters

The DNN structure for one-tap MIMO channel can achieve the optimal maximum likelihood detection performance, and furthermore, the CNN and RNN structures for multipath fading channel can detect the transmitted signal properly.

Multilevel MIMO Detection with Deep Learning

It is shown how real multilevel modulation symbols can be detected via deep neural networks via DNN, and a twin-network neural structure is proposed for detection.

Fixing the Complexity of the Sphere Decoder for MIMO Detection

A new detection algorithm for uncoded multiple input-multiple output (MIMO) systems based on the complex version of the sphere decoder (SD) is presented, overcoming the two main problems of the SD from an implementation point of view: its variable complexity and its sequential nature.

Improving Massive MIMO Belief Propagation Detector with Deep Neural Network

The results show that, compared to other state-of-the-art detectors, the DNN detectors can achieve lower bit error rate (BER) with improved robustness against various antenna configurations and channel conditions at the same level of complexity.

Large MIMO Systems

There is in-depth coverage of algorithms for large MIMO signal processing, based on meta-heuristics, belief propagation and Monte Carlo sampling techniques, and suited for large-scale signal detection, precoding and LDPC code designs.

Fifty Years of MIMO Detection: The Road to Large-Scale MIMOs

This treatise provides a recital on the historic heritages and novel challenges facing LS-MIMOs from a detection perspective, and an extensive review of the representative MIMO detection methods conceived during the past 50 years is presented, and relevant insights are inferred for the sake of designing complexity-scalable MIMo detection algorithms that are potentially applicable to LS-mIMO systems.