Physics-Informed Supervised Residual Learning for Electromagnetic Modeling

Shan, Tao; Zeng, Jinhong; Song, Xiaoqian; Guo, Rui; Li, Maokun; Yang, Fan; Xu, Shenheng

doi:10.1109/tap.2023.3245281

Cited by 20 publications

(4 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, exploiting the native graphic processing unit (GPU) implementations of GNNs [11], GEM can solve Maxwell's equations up to 40 times faster than conventional FDTD based on central processing unit (CPU) parallelization while yielding exactly the same results; interestingly, we also observe that GEM is at least twice as fast as state-of-the-art FDTD implementations that exploit advanced optimizations and parallelization, e.g., grid chunking [7], even if our solution does not yet adopt those same techniques. Contrary to some recent trends in the literature [12]- [17], it is shown that it's not necessary to identify appropriate input features or deep learning architectures aimed at training datadriven CEM models. Instead, FDTD equations can be intrinsically transformed into a GEM model that fundamentally outperforms the legacy implementation of the full-wave solver in terms of computational performance, without any need for training.…”

Section: Introductionmentioning

confidence: 78%

“…In this direction, the use of convolutional neural networks (CNNs) and sequence models, such as long short-term memory networks has been explored [12]- [16]. Despite being able to capture spatiotemporal correlations, the scalability of such models as standalone solvers of Maxwell's equations is innately limited.…”

Section: Related Workmentioning

confidence: 99%

“…It is paramount to remark that GEM does not require training, as its structure is already defined to perfectly mimic the numerical solution to Maxwell equations. In this sense, GEM is innately superior to all data-driven approaches proposed for CEM [12]- [19], since its predictions do not introduce any error, and it employs very simple operations to infer the electromagnetic fields; specifically, two matrix multiplications and two aggregations. Indicatively, in Figs.…”

Section: A Proof Of Gem Conceptmentioning

confidence: 99%

See 2 more Smart Citations

Stochastic Evaluation of Indoor Wireless Network Performance with Data-Driven Propagation Models

Bakirtzis

Wassell

Fiore

et al. 2022

GLOBECOM 2022 - 2022 IEEE Global Communications Conference

View full text Add to dashboard Cite

Computational electromagnetics (CEM) is employed to numerically solve Maxwell's equations, and it has very important and practical applications across a broad range of disciplines, including biomedical engineering, nanophotonics, wireless communications, and electrodynamics. The main limitation of existing CEM methods is that they are computationally demanding. Our work introduces a leap forward in scientific computing and CEM by proposing an original solution of Maxwell's equations that is grounded on graph neural networks (GNNs) and enables the high-performance numerical resolution of these fundamental mathematical expressions. Specifically, we demonstrate that the update equations derived by discretizing Maxwell's partial differential equations can be innately expressed as a two-layer GNN with static and pre-determined edge weights. Given this intuition, a straightforward way to numerically solve Maxwell's equations entails simple message passing between such a GNN's nodes, yielding a significant computational time gain, while preserving the same accuracy as conventional transient CEM methods. Ultimately, our work supports the efficient and precise emulation of electromagnetic wave propagation with GNNs, and more importantly, we anticipate that applying a similar treatment to systems of partial differential equations arising in other scientific disciplines, e.g., computational fluid dynamics, can benefit computational sciences.

show abstract

Section: Introductionmentioning

confidence: 78%

Section: Related Workmentioning

confidence: 99%

Section: A Proof Of Gem Conceptmentioning

confidence: 99%

See 1 more Smart Citation

Stochastic Evaluation of Indoor Wireless Network Performance with Data-Driven Propagation Models

Bakirtzis

Wassell

Fiore

et al. 2022

GLOBECOM 2022 - 2022 IEEE Global Communications Conference

View full text Add to dashboard Cite

show abstract

“…Such integration has been shown its accuracy, reliability, and interpretability of predictions in various applications [4]- [9]. When DL is taken as a numerical solver, including a computational electromagnetic (EM) one, NNs are typically regarded as function approximators [10]- [14], functional approximators [15]- [20] or operator approximators [21], [22]. In this work, we are interested in taking NNs as function approximators where the EM data is regarded as the function with respect to the spatial coordinate.…”

Section: Introductionmentioning

confidence: 99%

Universal Approximation Theorem and Deep Learning for the Solution of Frequency Domain Electromagnetic Scattering Problems

Wang,

Pan

2024

Preprint

View full text Add to dashboard Cite

Unlike the universal approximation theorems for functions mapping from a real-valued (RV) vector to a RV number or from a complex-valued (CV) vector to a CV number, in the field of electromagnetism, we need to approximate functions mapping from a RV vector to a CV number when we consider the electric field as a function of the spatial coordinate in the frequency domain. Typically, CV numbers contain phase information. When such phase information is handled properly, the performance of the neural networks (NNs) can be improved. This work derives a universal approximation theorem for functions mapping from a RV vector to a CV number. A deep NN, named as HV-DL, is designed accordingly, which consists of a RV input layer, a RV module containing two branches, a CV module and a CV output layer. The proposed universal approximation theorem is verified by numerical experiments on the HV-DL solution of the two-dimensional (2-D) electric field integral equation (EFIE). To integrate the underlying physics of electromagnetic scattering into the proposed HV-DL, the loss function is computed according to the EFIE.

show abstract

A comprehensive review of model compression techniques in machine learning

Dantas,

Sabino da Silva,

Cordeiro

et al. 2024

Appl Intell

View full text Add to dashboard Cite

This paper critically examines model compression techniques within the machine learning (ML) domain, emphasizing their role in enhancing model efficiency for deployment in resource-constrained environments, such as mobile devices, edge computing, and Internet of Things (IoT) systems. By systematically exploring compression techniques and lightweight design architectures, it is provided a comprehensive understanding of their operational contexts and effectiveness. The synthesis of these strategies reveals a dynamic interplay between model performance and computational demand, highlighting the balance required for optimal application. As machine learning (ML) models grow increasingly complex and data-intensive, the demand for computational resources and memory has surged accordingly. This escalation presents significant challenges for the deployment of artificial intelligence (AI) systems in real-world applications, particularly where hardware capabilities are limited. Therefore, model compression techniques are not merely advantageous but essential for ensuring that these models can be utilized across various domains, maintaining high performance without prohibitive resource requirements. Furthermore, this review underscores the importance of model compression in sustainable artificial intelligence (AI) development. The introduction of hybrid methods, which combine multiple compression techniques, promises to deliver superior performance and efficiency. Additionally, the development of intelligent frameworks capable of selecting the most appropriate compression strategy based on specific application needs is crucial for advancing the field. The practical examples and engineering applications discussed demonstrate the real-world impact of these techniques. By optimizing the balance between model complexity and computational efficiency, model compression ensures that the advancements in AI technology remain sustainable and widely applicable. This comprehensive review thus contributes to the academic discourse and guides innovative solutions for efficient and responsible machine learning practices, paving the way for future advancements in the field. Graphical abstract

show abstract

Physics-Informed Supervised Residual Learning for Electromagnetic Modeling

Cited by 20 publications

References 56 publications

Stochastic Evaluation of Indoor Wireless Network Performance with Data-Driven Propagation Models

Stochastic Evaluation of Indoor Wireless Network Performance with Data-Driven Propagation Models

Universal Approximation Theorem and Deep Learning for the Solution of Frequency Domain Electromagnetic Scattering Problems

A comprehensive review of model compression techniques in machine learning

Contact Info

Product

Resources

About