Deep neural network-aided Gaussian message passing detection for ultra-reliable low-latency communications

Ultra-reliable low-latency communications (URLLC) is a key technology in 5G supporting real-time multimedia services, which requires a low-cost signal recovery technology in the physical layer. A kind of well-known low-complexity signal detection is message passing algorithm (MPA) based on factor gr...

Full description

Saved in:
Bibliographic Details
Main Authors: Guo, Jie, Song, Bin, Chi, Yuhao, Jayasinghe, Lahiru, Yuen, Chau, Guan, Yong Liang, Du, Xiaojiang, Guizani, Mohsen
Other Authors: School of Electrical and Electronic Engineering
Format: Article
Language:English
Published: 2021
Subjects:
Online Access:https://hdl.handle.net/10356/151659
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-151659
record_format dspace
spelling sg-ntu-dr.10356-1516592021-07-01T08:07:44Z Deep neural network-aided Gaussian message passing detection for ultra-reliable low-latency communications Guo, Jie Song, Bin Chi, Yuhao Jayasinghe, Lahiru Yuen, Chau Guan, Yong Liang Du, Xiaojiang Guizani, Mohsen School of Electrical and Electronic Engineering Engineering::Electrical and electronic engineering Ultra-reliable Low-latency Communications Deep Neural Network Ultra-reliable low-latency communications (URLLC) is a key technology in 5G supporting real-time multimedia services, which requires a low-cost signal recovery technology in the physical layer. A kind of well-known low-complexity signal detection is message passing algorithm (MPA) based on factor graph. However, reliability and robustness of MPA are deteriorated when there are cycles in factor graph. To address this issue, we propose two novel Gaussian message passing (GMP) algorithms with the aid of deep neural network (DNN), in which the network architectures consist of two DNNs associated with detections for mean and variance of the signal. Particularly, the network architecture is constructed by transforming the factor graph and message update functions of the original GMP algorithm from node-type into edge-type. Then, weights and bias parameters are assigned in the network architecture. With the aid of deep learning methods, the optimal weights and bias parameters are obtained. Numerical results demonstrate that two proposed DNN-aided GMP algorithms can significantly improve the convergence of original GMP algorithm and also achieve robust performances in the cases without prior information. This work has been supported by the National Natural Science Foundation of China (No. 61772387, 61802296, 61750110529), China Postdoctoral Science Foundation Grant (No. 2017M620438), the Fundamental Research Funds for the Central Universities (JB180101), Fundamental Research Funds of Ministry of Education and China Mobile (MCM20170202), and also supported by the ISN State Key Laboratory . 2021-07-01T08:07:44Z 2021-07-01T08:07:44Z 2019 Journal Article Guo, J., Song, B., Chi, Y., Jayasinghe, L., Yuen, C., Guan, Y. L., Du, X. & Guizani, M. (2019). Deep neural network-aided Gaussian message passing detection for ultra-reliable low-latency communications. Future Generation Computer Systems, 95, 629-638. https://dx.doi.org/10.1016/j.future.2019.01.041 0167-739X 0000-0002-9859-8228 0000-0002-9307-2120 https://hdl.handle.net/10356/151659 10.1016/j.future.2019.01.041 2-s2.0-85060922018 95 629 638 en Future Generation Computer Systems © 2019 Elsevier B.V. All rights reserved.
institution Nanyang Technological University
building NTU Library
continent Asia
country Singapore
Singapore
content_provider NTU Library
collection DR-NTU
language English
topic Engineering::Electrical and electronic engineering
Ultra-reliable Low-latency Communications
Deep Neural Network
spellingShingle Engineering::Electrical and electronic engineering
Ultra-reliable Low-latency Communications
Deep Neural Network
Guo, Jie
Song, Bin
Chi, Yuhao
Jayasinghe, Lahiru
Yuen, Chau
Guan, Yong Liang
Du, Xiaojiang
Guizani, Mohsen
Deep neural network-aided Gaussian message passing detection for ultra-reliable low-latency communications
description Ultra-reliable low-latency communications (URLLC) is a key technology in 5G supporting real-time multimedia services, which requires a low-cost signal recovery technology in the physical layer. A kind of well-known low-complexity signal detection is message passing algorithm (MPA) based on factor graph. However, reliability and robustness of MPA are deteriorated when there are cycles in factor graph. To address this issue, we propose two novel Gaussian message passing (GMP) algorithms with the aid of deep neural network (DNN), in which the network architectures consist of two DNNs associated with detections for mean and variance of the signal. Particularly, the network architecture is constructed by transforming the factor graph and message update functions of the original GMP algorithm from node-type into edge-type. Then, weights and bias parameters are assigned in the network architecture. With the aid of deep learning methods, the optimal weights and bias parameters are obtained. Numerical results demonstrate that two proposed DNN-aided GMP algorithms can significantly improve the convergence of original GMP algorithm and also achieve robust performances in the cases without prior information.
author2 School of Electrical and Electronic Engineering
author_facet School of Electrical and Electronic Engineering
Guo, Jie
Song, Bin
Chi, Yuhao
Jayasinghe, Lahiru
Yuen, Chau
Guan, Yong Liang
Du, Xiaojiang
Guizani, Mohsen
format Article
author Guo, Jie
Song, Bin
Chi, Yuhao
Jayasinghe, Lahiru
Yuen, Chau
Guan, Yong Liang
Du, Xiaojiang
Guizani, Mohsen
author_sort Guo, Jie
title Deep neural network-aided Gaussian message passing detection for ultra-reliable low-latency communications
title_short Deep neural network-aided Gaussian message passing detection for ultra-reliable low-latency communications
title_full Deep neural network-aided Gaussian message passing detection for ultra-reliable low-latency communications
title_fullStr Deep neural network-aided Gaussian message passing detection for ultra-reliable low-latency communications
title_full_unstemmed Deep neural network-aided Gaussian message passing detection for ultra-reliable low-latency communications
title_sort deep neural network-aided gaussian message passing detection for ultra-reliable low-latency communications
publishDate 2021
url https://hdl.handle.net/10356/151659
_version_ 1705151288222679040