SI/PI-Database of PCB-Based Interconnects for Machine Learning Applications

Schierholz, Morten; Sanchez-Masis, Allan; Carmona-Cruz, Allan; Duan, Xiaomin; Roy, Kallol; Yang, Cheng Fu; Rimolo-Donadío, Renato; Schuster, Christian

doi:10.1109/access.2021.3061788

Cited by 22 publications

(3 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, artificial neural networks (ANN) have been used to predict the S-parameter of high-speed interconnects based on an open SI/PI-Database [1], [3]. However, the data-driven neural networks [3] can lead to negative insertion loss, which violates the passive characteristics that were demonstrated in this paper.…”

Section: Introductionmentioning

confidence: 80%

“…It is well known that the insertion loss (IL) is defined as [4] IL = −20 log 10 |S 21 | dB (1) where S 21 represents the transmission S-parameter of highspeed interconnects from port 1 to port 2. We perform our study on an open SI/PI dataset, which has transmission lines on 11 cavity PCB with two 10x10 via-arrays [1]. Fig.…”

Section: Background and Motivationmentioning

confidence: 99%

“…Machine learning (ML), especially deep learning is gaining greater attention due to its success in various cognitive applications. Recently, machine learning approaches have been applied for electromagnetic simulation and modeling, which has shown great promise [1]. Deep neural networks can be trained to build accurate models for nonlinear electromagnetic responses with different design parameters based on the simulated and experimental data.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Physics-Enforced Modeling for Insertion Loss of Transmission Lines by Deep Neural Networks

Chen¹,

Tan²

2021

Preprint

View full text Add to dashboard Cite

In this paper, we investigate data-driven parameterized modeling of insertion loss for transmission lines with respect to design parameters. We first show that direct application of neural networks can lead to non-physics models with negative insertion loss. To mitigate this problem, we propose two deep learning solutions. One solution is to add a regulation term, which represents the passive condition, to the final loss function to enforce the negative quantity of insertion loss. In the second method, a third-order polynomial expression is defined first, which ensures positiveness, to approximate the insertion loss, then DeepONet neural network structure, which was proposed recently for function and system modeling, was employed to model the coefficients of polynomials. The resulting neural network is applied to predict the coefficients of the polynomial expression. The experimental results on an open-sourced SI/PI database of a PCB design show that both methods can ensure the positiveness for the insertion loss. Furthermore, both methods can achieve similar prediction results, while the polynomial-based DeepONet method is faster than DeepONet based method in training time.

show abstract

Section: Introductionmentioning

confidence: 80%

Section: Background and Motivationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation