A DNN-Based OFDM Channel Estimation Algorithm Without Training Overheads

Yujia Zhu; Rongrong Qian; Xiaoming Lv; Wenping Ren; Mathini Sellathurai

doi:10.1109/ITAIC58329.2023.10408909

A DNN-Based OFDM Channel Estimation Algorithm Without Training Overheads

Yujia Zhu, Rongrong Qian, Xiaoming Lv, Wenping Ren, Mathini Sellathurai

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

In this paper, we propose a channel estimation algorithm for OFDM systems based on a deep neural network to reduce overheads in model training. In this method, the channel estimation problem is formulated as an image repair problem, where a channel matrix containing pilot values is regarded as an incomplete picture, and then a specially designed deep neural network based on the deep image prior (DIP) is exploited to reconstruct complete and noise-removed channel images from the incomplete picture. While reducing complexity and training overheads, the method also ensures estimation accuracy. Simulation results show the superior performance and effectiveness of the proposed channel estimation algorithm.

Original language	English
Title of host publication	11th IEEE Joint International Information Technology and Artificial Intelligence Conference (ITAIC)
Publisher	IEEE
Pages	687-691
Number of pages	5
ISBN (Electronic)	9798350333664
DOIs	https://doi.org/10.1109/ITAIC58329.2023.10408909
Publication status	Published - 1 Feb 2024
Event	11th IEEE Joint International Information Technology and Artificial Intelligence Conference 2023 - Chongqing, China Duration: 8 Dec 2023 → 10 Dec 2023

Conference

Conference	11th IEEE Joint International Information Technology and Artificial Intelligence Conference 2023
Abbreviated title	ITAIC 2023
Country/Territory	China
City	Chongqing
Period	8/12/23 → 10/12/23

Keywords

OFDM
channel estimation
deep image prior
deep learning

ASJC Scopus subject areas

Artificial Intelligence
Information Systems

Access to Document

10.1109/ITAIC58329.2023.10408909

Cite this

@inproceedings{6cfb99bca8144f91b0edf581c0ae066f,

title = "A DNN-Based OFDM Channel Estimation Algorithm Without Training Overheads",

abstract = "In this paper, we propose a channel estimation algorithm for OFDM systems based on a deep neural network to reduce overheads in model training. In this method, the channel estimation problem is formulated as an image repair problem, where a channel matrix containing pilot values is regarded as an incomplete picture, and then a specially designed deep neural network based on the deep image prior (DIP) is exploited to reconstruct complete and noise-removed channel images from the incomplete picture. While reducing complexity and training overheads, the method also ensures estimation accuracy. Simulation results show the superior performance and effectiveness of the proposed channel estimation algorithm.",

keywords = "OFDM, channel estimation, deep image prior, deep learning",

author = "Yujia Zhu and Rongrong Qian and Xiaoming Lv and Wenping Ren and Mathini Sellathurai",

year = "2024",

month = feb,

day = "1",

doi = "10.1109/ITAIC58329.2023.10408909",

language = "English",

pages = "687--691",

booktitle = "11th IEEE Joint International Information Technology and Artificial Intelligence Conference (ITAIC)",

publisher = "IEEE",

address = "United States",

note = "11th IEEE Joint International Information Technology and Artificial Intelligence Conference 2023, ITAIC 2023 ; Conference date: 08-12-2023 Through 10-12-2023",

}

Zhu, Y, Qian, R, Lv, X, Ren, W & Sellathurai, M 2024, A DNN-Based OFDM Channel Estimation Algorithm Without Training Overheads. in 11th IEEE Joint International Information Technology and Artificial Intelligence Conference (ITAIC). IEEE, pp. 687-691, 11th IEEE Joint International Information Technology and Artificial Intelligence Conference 2023, Chongqing, China, 8/12/23. https://doi.org/10.1109/ITAIC58329.2023.10408909

TY - GEN

T1 - A DNN-Based OFDM Channel Estimation Algorithm Without Training Overheads

AU - Zhu, Yujia

AU - Qian, Rongrong

AU - Lv, Xiaoming

AU - Ren, Wenping

AU - Sellathurai, Mathini

PY - 2024/2/1

Y1 - 2024/2/1

N2 - In this paper, we propose a channel estimation algorithm for OFDM systems based on a deep neural network to reduce overheads in model training. In this method, the channel estimation problem is formulated as an image repair problem, where a channel matrix containing pilot values is regarded as an incomplete picture, and then a specially designed deep neural network based on the deep image prior (DIP) is exploited to reconstruct complete and noise-removed channel images from the incomplete picture. While reducing complexity and training overheads, the method also ensures estimation accuracy. Simulation results show the superior performance and effectiveness of the proposed channel estimation algorithm.

AB - In this paper, we propose a channel estimation algorithm for OFDM systems based on a deep neural network to reduce overheads in model training. In this method, the channel estimation problem is formulated as an image repair problem, where a channel matrix containing pilot values is regarded as an incomplete picture, and then a specially designed deep neural network based on the deep image prior (DIP) is exploited to reconstruct complete and noise-removed channel images from the incomplete picture. While reducing complexity and training overheads, the method also ensures estimation accuracy. Simulation results show the superior performance and effectiveness of the proposed channel estimation algorithm.

KW - OFDM

KW - channel estimation

KW - deep image prior

KW - deep learning

UR - http://www.scopus.com/inward/record.url?scp=85186080382&partnerID=8YFLogxK

U2 - 10.1109/ITAIC58329.2023.10408909

DO - 10.1109/ITAIC58329.2023.10408909

M3 - Conference contribution

SP - 687

EP - 691

BT - 11th IEEE Joint International Information Technology and Artificial Intelligence Conference (ITAIC)

PB - IEEE

T2 - 11th IEEE Joint International Information Technology and Artificial Intelligence Conference 2023

Y2 - 8 December 2023 through 10 December 2023

ER -

A DNN-Based OFDM Channel Estimation Algorithm Without Training Overheads

Abstract

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this