Abstract

While machine learning is becoming a demanding request in every corner of modern technology development, we are trying to see if we could make computational electromagnetic algorithms compatible to machine learning methods. In this paper, we introduce two efforts in line with this direction: solving method of moments (MoM) can be seen as a training training process. Consequently, the artificial neural network (ANN) could be used to solve MoM naturally through training. Amazon Web Service (AWS) can be used as the computations platform to utilize the existing hardware and software resources for machine learning. Another effort regarding to the nonlinear IO of ICs can be modeled through ANN. Hence, a behavior model with growing accuracy can be obtained for the signal integrity and power integrity analysis. It can be further hybridized into discontinuous Galerkin time domain (DGTD) method for CEM characterizations. Benchmarks are provided to demonstrate the feasibility of the proposed methods.

Department(s)

Electrical and Computer Engineering

Comments

National Natural Science Foundation of China, Grant GRF 17207114

Keywords and Phrases

Artificial Neural Network; CEM; DGTD; Machine Learning; MoM

International Standard Book Number (ISBN)

978-153861241-5

Document Type

Article - Conference proceedings

Document Version

Citation

File Type

text

Language(s)

English

Rights

© 2024 The Institute of Engineering and Technology, All rights reserved.

Publication Date

17 Oct 2018

Download

Full Text Link

Share

 
COinS