.png){kind=icon}
.png){kind=icon}
Return
Deep Learning-Based Iterative Residual Encoder–Decoder Solution for Electromagnetic Modeling Over a Broad Frequency Band
.png){kind=icon}
PRE
AI
DOI：10.1109/LAWP.2025.3597920.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In this letter, we propose a deep learning-based iterative residual encoder–decoder method (IRED), which provides an efficient deep learning framework for electromagnetic modeling over a broad frequency band. The method leverages the connection between traditional iterative method and iterative neural networks, making it more versatile by processing the imaginary and real parts of complex numbers in two distinct channels. In the neural network, each iteration layer consists of an encoder–decoder neural network designed to solve the residual at each iterative step. In a given region, the input of the IRED combines the initial guess of the total field at any frequency across a broadband with the contrast distribution of the scatterer, while the corresponding output is the total electromagnetic field. The effectiveness of this method is validated through the 2-D volume integral equation. Numerical results demonstrate that the method operates effectively over a broadband, achieving higher accuracy and faster efficiency compared to traditional algorithms.
Keywords:
Broad frequency band
deep learning (DL)
electromagnetic modeling (EM)
iterative residual encoder–decoder (IRED)
volume integral equation (VIE)
Journal
IF:
4.8
Papers:
1.0W
Citations:
2.8W