Graph Neural Network Assisted S-Parameter Inference and Control-Word Generation of Terahertz Reconfigurable Intelligent Surface

Zihan Ning; Tong Sun; Jizhao Li; Yanfei Ren; Chenjia Xie; Li Du; Lianggong Wen; Yuan Du

doi:10.1109/ICTA60488.2023.10364304

Graph Neural Network Assisted S-Parameter Inference and Control-Word Generation of Terahertz Reconfigurable Intelligent Surface

Zihan Ning
, Tong Sun
, Jizhao Li
, Yanfei Ren
, Chenjia Xie
, Li Du
, Lianggong Wen^*
, Yuan Du^*

^*Corresponding author for this work

School of Integrated Circuit Science and Engineering

Nanjing University
Beihang University

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

1 Scopus citations

Abstract

For large-scale terahertz (THz) reconfigurable intelligent surfaces (RIS), each unit element is supposed to be controlled by an independent voltage source. This leads to a huge solution space for both key metrics inference and control-word generation. In this paper, we propose a fast AI-assisted control-word generation scheme to reduce the computation cost of Electro-Magnetic (EM) simulations (Forward Model) and to accelerate the iteration process for control word search (Inverse Model). The results demonstrate that the Forward Model can predict the S-parameters between 100GHz and 800GHz with a minimal mean absolute error (MAE) of 0.69dB. Our method is more than 180 times faster than traditional full-wave simulation methods without training time. Additionally, the Inverse Model can generate demanded control words within 1.5 dB error requirement in less than 200 iterations.

Original language	English
Title of host publication	Proceedings of 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	41-42
Number of pages	2
ISBN (Electronic)	9798350344288
DOIs	https://doi.org/10.1109/ICTA60488.2023.10364304
State	Published - 2023
Event	2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023 - Hefei, China Duration: 27 Oct 2023 → 29 Oct 2023

Publication series

Name	Proceedings of 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023

Conference

Conference	2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023
Country/Territory	China
City	Hefei
Period	27/10/23 → 29/10/23

Keywords

AI-aided design
Control-word generation
Neural networks
Terahertz reconfigurable intelligent surfaces
s-parameter inference

Access to Document

10.1109/ICTA60488.2023.10364304

Cite this

Ning, Z., Sun, T., Li, J., Ren, Y., Xie, C., Du, L., Wen, L., & Du, Y. (2023). Graph Neural Network Assisted S-Parameter Inference and Control-Word Generation of Terahertz Reconfigurable Intelligent Surface. In Proceedings of 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023 (pp. 41-42). (Proceedings of 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICTA60488.2023.10364304

Ning, Zihan ; Sun, Tong ; Li, Jizhao et al. / Graph Neural Network Assisted S-Parameter Inference and Control-Word Generation of Terahertz Reconfigurable Intelligent Surface. Proceedings of 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023. Institute of Electrical and Electronics Engineers Inc., 2023. pp. 41-42 (Proceedings of 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023).

@inproceedings{45494c3f64704e7bbe8f208068549748,

title = "Graph Neural Network Assisted S-Parameter Inference and Control-Word Generation of Terahertz Reconfigurable Intelligent Surface",

abstract = "For large-scale terahertz (THz) reconfigurable intelligent surfaces (RIS), each unit element is supposed to be controlled by an independent voltage source. This leads to a huge solution space for both key metrics inference and control-word generation. In this paper, we propose a fast AI-assisted control-word generation scheme to reduce the computation cost of Electro-Magnetic (EM) simulations (Forward Model) and to accelerate the iteration process for control word search (Inverse Model). The results demonstrate that the Forward Model can predict the S-parameters between 100GHz and 800GHz with a minimal mean absolute error (MAE) of 0.69dB. Our method is more than 180 times faster than traditional full-wave simulation methods without training time. Additionally, the Inverse Model can generate demanded control words within 1.5 dB error requirement in less than 200 iterations.",

keywords = "AI-aided design, Control-word generation, Neural networks, Terahertz reconfigurable intelligent surfaces, s-parameter inference",

author = "Zihan Ning and Tong Sun and Jizhao Li and Yanfei Ren and Chenjia Xie and Li Du and Lianggong Wen and Yuan Du",

note = "Publisher Copyright: {\textcopyright} 2023 IEEE.; 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023 ; Conference date: 27-10-2023 Through 29-10-2023",

year = "2023",

doi = "10.1109/ICTA60488.2023.10364304",

language = "英语",

series = "Proceedings of 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "41--42",

booktitle = "Proceedings of 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023",

address = "美国",

}

Ning, Z, Sun, T, Li, J, Ren, Y, Xie, C, Du, L, Wen, L & Du, Y 2023, Graph Neural Network Assisted S-Parameter Inference and Control-Word Generation of Terahertz Reconfigurable Intelligent Surface. in Proceedings of 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023. Proceedings of 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023, Institute of Electrical and Electronics Engineers Inc., pp. 41-42, 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023, Hefei, China, 27/10/23. https://doi.org/10.1109/ICTA60488.2023.10364304

Graph Neural Network Assisted S-Parameter Inference and Control-Word Generation of Terahertz Reconfigurable Intelligent Surface. / Ning, Zihan; Sun, Tong; Li, Jizhao et al.
Proceedings of 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023. Institute of Electrical and Electronics Engineers Inc., 2023. p. 41-42 (Proceedings of 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023).

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

TY - GEN

T1 - Graph Neural Network Assisted S-Parameter Inference and Control-Word Generation of Terahertz Reconfigurable Intelligent Surface

AU - Ning, Zihan

AU - Sun, Tong

AU - Li, Jizhao

AU - Ren, Yanfei

AU - Xie, Chenjia

AU - Du, Li

AU - Wen, Lianggong

AU - Du, Yuan

PY - 2023

Y1 - 2023

N2 - For large-scale terahertz (THz) reconfigurable intelligent surfaces (RIS), each unit element is supposed to be controlled by an independent voltage source. This leads to a huge solution space for both key metrics inference and control-word generation. In this paper, we propose a fast AI-assisted control-word generation scheme to reduce the computation cost of Electro-Magnetic (EM) simulations (Forward Model) and to accelerate the iteration process for control word search (Inverse Model). The results demonstrate that the Forward Model can predict the S-parameters between 100GHz and 800GHz with a minimal mean absolute error (MAE) of 0.69dB. Our method is more than 180 times faster than traditional full-wave simulation methods without training time. Additionally, the Inverse Model can generate demanded control words within 1.5 dB error requirement in less than 200 iterations.

AB - For large-scale terahertz (THz) reconfigurable intelligent surfaces (RIS), each unit element is supposed to be controlled by an independent voltage source. This leads to a huge solution space for both key metrics inference and control-word generation. In this paper, we propose a fast AI-assisted control-word generation scheme to reduce the computation cost of Electro-Magnetic (EM) simulations (Forward Model) and to accelerate the iteration process for control word search (Inverse Model). The results demonstrate that the Forward Model can predict the S-parameters between 100GHz and 800GHz with a minimal mean absolute error (MAE) of 0.69dB. Our method is more than 180 times faster than traditional full-wave simulation methods without training time. Additionally, the Inverse Model can generate demanded control words within 1.5 dB error requirement in less than 200 iterations.

KW - AI-aided design

KW - Control-word generation

KW - Neural networks

KW - Terahertz reconfigurable intelligent surfaces

KW - s-parameter inference

UR - https://www.scopus.com/pages/publications/85182939617

U2 - 10.1109/ICTA60488.2023.10364304

DO - 10.1109/ICTA60488.2023.10364304

M3 - 会议稿件

AN - SCOPUS:85182939617

T3 - Proceedings of 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023

SP - 41

EP - 42

BT - Proceedings of 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023

Y2 - 27 October 2023 through 29 October 2023

ER -

Ning Z, Sun T, Li J, Ren Y, Xie C, Du L et al. Graph Neural Network Assisted S-Parameter Inference and Control-Word Generation of Terahertz Reconfigurable Intelligent Surface. In Proceedings of 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023. Institute of Electrical and Electronics Engineers Inc. 2023. p. 41-42. (Proceedings of 2023 IEEE International Conference on Integrated Circuits, Technologies and Applications, ICTA 2023). doi: 10.1109/ICTA60488.2023.10364304

Graph Neural Network Assisted S-Parameter Inference and Control-Word Generation of Terahertz Reconfigurable Intelligent Surface

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this