Length-Dependent Deep Neural Network Based Modeling for High-Speed Channels
This article presents a length-dependent deep neural network (LD-DNN) based channel modeling methodology to predict the frequency response of high-speed channels. The proposed method significantly enhances the model accuracy and design efficiency while considering the channel length dependence that...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10385090/ |
| _version_ | 1827378255779332096 |
|---|---|
| author | Hung Khac Le Soyoung Kim |
| author_facet | Hung Khac Le Soyoung Kim |
| author_sort | Hung Khac Le |
| collection | DOAJ |
| description | This article presents a length-dependent deep neural network (LD-DNN) based channel modeling methodology to predict the frequency response of high-speed channels. The proposed method significantly enhances the model accuracy and design efficiency while considering the channel length dependence that was neglected in previous modeling approaches. We define the concept of the electrical length to model the length and frequency dependence, then further leverage the activation function to capture the multiple reflection effects to improve accuracy. Additionally, we model the insertion loss resonance induced by crosstalk that can seriously deteriorate signal integrity. As a result, by adopting the proposed model which can predict the S-parameters as a function of length, the need for performing additionally 3D electromagnetic simulations when adjusting the channel length can be eliminated. Various high-speed channel cases are tested to validate the accuracy of the proposed method. The modeling accuracy is less than 4% for different high-speed channel structures with run times of less than 1.4 second per design. |
| first_indexed | 2024-03-08T12:53:23Z |
| format | Article |
| id | doaj.art-64e42e29eb534e26b84a0e516564c94c |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-03-08T12:53:23Z |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-64e42e29eb534e26b84a0e516564c94c2024-01-20T00:02:21ZengIEEEIEEE Access2169-35362024-01-01127624763610.1109/ACCESS.2024.335184310385090Length-Dependent Deep Neural Network Based Modeling for High-Speed ChannelsHung Khac Le0https://orcid.org/0000-0002-8466-5645Soyoung Kim1https://orcid.org/0000-0001-8901-3649Department of Electrical and Computer Engineering, College of Information and Communication Engineering, Sungkyunkwan University, Suwon, South KoreaCollege of Information and Communication Engineering, Sungkyunkwan University, Suwon, South KoreaThis article presents a length-dependent deep neural network (LD-DNN) based channel modeling methodology to predict the frequency response of high-speed channels. The proposed method significantly enhances the model accuracy and design efficiency while considering the channel length dependence that was neglected in previous modeling approaches. We define the concept of the electrical length to model the length and frequency dependence, then further leverage the activation function to capture the multiple reflection effects to improve accuracy. Additionally, we model the insertion loss resonance induced by crosstalk that can seriously deteriorate signal integrity. As a result, by adopting the proposed model which can predict the S-parameters as a function of length, the need for performing additionally 3D electromagnetic simulations when adjusting the channel length can be eliminated. Various high-speed channel cases are tested to validate the accuracy of the proposed method. The modeling accuracy is less than 4% for different high-speed channel structures with run times of less than 1.4 second per design.https://ieeexplore.ieee.org/document/10385090/Activation functiondeep neural network (DNN)electrical lengthhigh-speed channelmultiple reflectionresonance |
| spellingShingle | Hung Khac Le Soyoung Kim Length-Dependent Deep Neural Network Based Modeling for High-Speed Channels IEEE Access Activation function deep neural network (DNN) electrical length high-speed channel multiple reflection resonance |
| title | Length-Dependent Deep Neural Network Based Modeling for High-Speed Channels |
| title_full | Length-Dependent Deep Neural Network Based Modeling for High-Speed Channels |
| title_fullStr | Length-Dependent Deep Neural Network Based Modeling for High-Speed Channels |
| title_full_unstemmed | Length-Dependent Deep Neural Network Based Modeling for High-Speed Channels |
| title_short | Length-Dependent Deep Neural Network Based Modeling for High-Speed Channels |
| title_sort | length dependent deep neural network based modeling for high speed channels |
| topic | Activation function deep neural network (DNN) electrical length high-speed channel multiple reflection resonance |
| url | https://ieeexplore.ieee.org/document/10385090/ |
| work_keys_str_mv | AT hungkhacle lengthdependentdeepneuralnetworkbasedmodelingforhighspeedchannels AT soyoungkim lengthdependentdeepneuralnetworkbasedmodelingforhighspeedchannels |