Improved Physics-Driven Neural Network to Solve Inverse Scattering Problems

2025-12-10

AI Summary

This work addresses the electromagnetic inverse scattering problem (ISP) by proposing a lightweight, high-accuracy physics-driven neural network. Methodologically, it introduces the Gaussian Local Oscillation-suppressing Window (GLOW) activation function—first of its kind—to effectively suppress high-frequency artifacts; designs an adaptive scattering-subregion dynamic identification mechanism to enhance target localization robustness; and achieves end-to-end deep integration of a physically interpretable iterative algorithm with neural networks, augmented by dynamic domain adaptation and transfer learning for improved generalization. Experiments on both synthetic and real-world measurement data demonstrate significant improvements: average reconstruction accuracy increases by 8.2%, noise robustness improves markedly (37% error reduction at SNR ≤ 15 dB), and inference speed accelerates by 3.1× over state-of-the-art methods. The proposed framework achieves superior overall performance, setting a new benchmark in ISP reconstruction.

Technology Category

• Search and Optimization: Local Search
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Adversarial Learning & Robustness

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

This paper presents an improved physics-driven neural network (IPDNN) framework for solving electromagnetic inverse scattering problems (ISPs). A new Gaussian-localized oscillation-suppressing window (GLOW) activation function is introduced to stabilize convergence and enable a lightweight yet accurate network architecture. A dynamic scatter subregion identification strategy is further developed to adaptively refine the computational domain, preventing missed detections and reducing computational cost. Moreover, transfer learning is incorporated to extend the solver's applicability to practical scenarios, integrating the physical interpretability of iterative algorithms with the real-time inference capability of neural networks. Numerical simulations and experimental results demonstrate that the proposed solver achieves superior reconstruction accuracy, robustness, and efficiency compared with existing state-of-the-art methods.

Problem

• Solves electromagnetic inverse scattering problems with improved accuracy
• Introduces a new activation function to stabilize network convergence
• Reduces computational cost through adaptive domain refinement

Innovation

• Improved physics-driven neural network with GLOW activation function
• Dynamic scatter subregion identification strategy reduces computational cost
• Transfer learning enhances applicability and integrates physical interpretability