Physics-Driven Neural Network for Solving Electromagnetic Inverse Scattering Problems
Du Y. Liu Z. Matkerim B. Li C. Zong Y. Qi B. Kou J.
2026Institute of Electrical and Electronics Engineers Inc.
IEEE Transactions on Antennas and Propagation
2026#74Issue 21945 - 1956 pp.
In recent years, deep-learning-based methods have been introduced for solving inverse scattering problems (ISPs), but most of them heavily rely on large training datasets and suffer from limited generalization capability. In this article, a new solving scheme is proposed where the solution is iteratively updated through a physics-driven neural network (PDNN), the hyperparameters of which are optimized by minimizing the loss function which incorporates the constraints from measured scattered fields and the prior information about scatterers. Unlike data-driven neural network solvers, PDNN does not require an offline training dataset. Its weights are iteratively updated based solely on the measured incident and scattered fields, similar in philosophy to conventional inverse algorithms but enhanced by neural network flexibility. Thus, the generalization issue is eliminated. Moreover, to accelerate imaging, a subregion enclosing the scatterers is automatically identified and used to restrict the computational domain. Numerical and experimental results demonstrate that the proposed scheme achieves high reconstruction accuracy and strong stability, even for complex and lossy scatterers.
Dielectric scatterers , inverse scattering imaging (ISP) , neural network , physics-driven
Text of the article Перейти на текст статьи
Northwestern Polytechnical University, School of Electronics and Information, Xian, 710072, China
Al-Farabi Kazakh National University, Department of Computer Science, Almaty, 050040, Kazakhstan
Xian Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Department of Advanced Optical Instrument Research, Xian, 710119, China
Northwestern Polytechnical University
Al-Farabi Kazakh National University
Xian Institute of Optics and Precision Mechanics
10 лет помогаем публиковать статьи Международный издатель
Книга Публикация научной статьи Волощук 2026 Book Publication of a scientific article 2026