TY - JOUR
T1 - Implementing the Fast Full-Wave Electromagnetic Forward Solver Using the Deep Convolutional Encoder-Decoder Architecture
AU - Yao, He Ming
AU - Jiang, Lijun
AU - Ng, Michael
N1 - This work was supported in part by the Research Grants Council of Hong Kong under Grant GRF 17207114 and Grant GRF 17210815; in part by the Asian Office of Aerospace Research and Development (AOARD) under Grant FA2386-17-1-0010; in part by NSFC under Grant 61271158; in part by Hong Kong UGC under Grant AoE/P–04/08; in part by HKRGC GRF under Grant 12306616, Grant 12200317, Grant 12300218, Grant 12300519, Grant 17201020, and Grant 17209918; and in part by the Fellowship Award from the Research Grants Council of the Hong Kong Special Administrative Region, China, under Grant HKU PDFS2122-7S05.
Publisher Copyright:
IEEE.
PY - 2023/1
Y1 - 2023/1
N2 - In this communication, a novel deep learning (DL)-based solver is proposed for the electromagnetic forward (EMF) process. It is based on the complex-valued deep convolutional neural networks (DConvNets) comprising an encoder network and a corresponding decoder network with pixel-wise regression layer. The encoder network takes the incident EM wave and the contrast (permittivity) distribution of the object as the input. It channels the processed data into the corresponding decoder network to predict the total EM field due to the scatter of the input incident EM wave. The training of the proposed DConvNets is done using the simple synthetic dataset. Due to its strong approximation capability, the proposed DConvNets can realize the prediction of EM field. Hence, the proposed DL-based EMF solver acts as a 'inhomogeneous' transformation-the unknown EM field in the objective domain is obtained through the transformation from the information of the incident EM field and the distribution of contrasts (permittivities). Compared with conventional methods, the EMF problem can be solved with higher accuracy and significantly reduced CPU time. Numerical examples have demonstrated the feasibility of this newly proposed approach. This newly proposed DL-based EMF solver presents a new alternative to electromagnetic computation approaches.
AB - In this communication, a novel deep learning (DL)-based solver is proposed for the electromagnetic forward (EMF) process. It is based on the complex-valued deep convolutional neural networks (DConvNets) comprising an encoder network and a corresponding decoder network with pixel-wise regression layer. The encoder network takes the incident EM wave and the contrast (permittivity) distribution of the object as the input. It channels the processed data into the corresponding decoder network to predict the total EM field due to the scatter of the input incident EM wave. The training of the proposed DConvNets is done using the simple synthetic dataset. Due to its strong approximation capability, the proposed DConvNets can realize the prediction of EM field. Hence, the proposed DL-based EMF solver acts as a 'inhomogeneous' transformation-the unknown EM field in the objective domain is obtained through the transformation from the information of the incident EM field and the distribution of contrasts (permittivities). Compared with conventional methods, the EMF problem can be solved with higher accuracy and significantly reduced CPU time. Numerical examples have demonstrated the feasibility of this newly proposed approach. This newly proposed DL-based EMF solver presents a new alternative to electromagnetic computation approaches.
KW - Convolutional neural network
KW - deep learning (DL)
KW - electromagnetic forward (EMF) process
KW - real time
UR - http://www.scopus.com/inward/record.url?scp=85141650156&partnerID=8YFLogxK
U2 - 10.1109/TAP.2022.3216920
DO - 10.1109/TAP.2022.3216920
M3 - Journal article
AN - SCOPUS:85141650156
SN - 0018-926X
VL - 71
SP - 1152
EP - 1157
JO - IEEE Transactions on Antennas and Propagation
JF - IEEE Transactions on Antennas and Propagation
IS - 1
ER -