PIBNet: a Physics-Inspired Boundary Network for Multiple Scattering Simulations

Rémi Marsal, Stéphanie Chaillat · 2025

Read PIBNet: a Physics-Inspired Boundary Network for Multiple Scattering Simulations by Rémi Marsal, Stéphanie Chaillat (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-02049.txt.

Introduction

Waves bouncing off several objects at once, multiple scattering, are costly to simulate across a whole volume. Rémi Marsal and Stéphanie Chaillat work in 2025 with the boundary element method, which needs values only on the objects' surfaces, and build a physics-inspired network around it, called PIBNet.

In the boundary approach, one first solves an integral equation for the wave's trace on each boundary. The full solution everywhere else then follows cheaply from a boundary integral formula. A network designed around that structure inherits the method's efficiency. Faster scattering simulations serve acoustics, electromagnetics, and any field where waves meet cluttered geometry.

Catalog record

CREATORRémi Marsal, Stéphanie Chaillat
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE41 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2512-02049

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

PIBNet: a Physics-Inspired Boundary Network for Multiple Scattering Simulations Rémi Marsal, Stéphanie Chaillat Abstract The boundary element method (BEM) provides an efficient numerical framework for solving multiple scattering problems in unbounded homogeneous domains, since it reduces the discretization to the domain boundaries, thereby condensing the computational complexity. The procedure first consists in determining the solution trace on the boundaries of the domain by solving a boundary integral equation, after which the volumetric solution can be recovered at low computational cost with a boundary integral representation. As the first step of the BEM represents the main computational bottleneck, we introduce PIBNet, a learning-based approach designed to approximate the solution trace. The method leverages a physics-inspired graph-based strategy to model obstacles and their long-range interactions efficiently. Then, we introduce a novel multiscale graph neural network architecture for simulating the multiple scattering. To train and evaluate our network, we present a benchmark consisting of several datasets of different types of multiple scattering problems. The results indicate that our approach not only surpasses existing state-of-the-art learning-based methods on the considered tasks but also exhibits superior generalization to settings with an increased number of obstacles. github.com/ENSTA-U2IS-AI/pibnet 1 Introduction Multiple scattering is defined by 28 as ”the interaction of fields with two or more obstacles”, as illustrated in fig. 1 . The main challenge in such problems lies in the combined effect of the number of obstacles and their separation distances, as these factors govern the complexity of the multiple reflections. The boundary element method (BEM) ( 4 ) is an efficient numerical method for solving linear partial differential equations (PDEs) such as the one involved in multiple scattering. Unlike the finite element method (FEM), the BEM reformulates the PDE into a boundary integral equation (BIE) with unknowns restricted to the problem boundaries. After solving the BIE, the solution in the volumetric domain is evaluated using the boundary integral representation. Thus, by reducing the dimensionality of the problem, the BEM can offer significant gains in computational efficiency and accuracy compared to the FEM for wave propagation problems in unbounded domains. Within the BEM, the two computational stages have very different costs: solving the BIE on the boundaries is significantly more expensive than evaluating the boundary integral representation to recover the volumetric solution. To reduce the overall computational cost, several learning-based approaches have been proposed to simulate multiple scattering ( 13 ; 30 ) . Taking advantage of the recent development in neural network architectures for solving PDEs, these approaches rely either on discretizing the solution domain or on neural fields. Discretization-based methods ( 31 ; 46 ) handle complex geometries and boundary conditions but inherit the computational cost of traditional solvers. Neural field approaches ( 33 ) represent the solution as a continuous function, allowing inference at arbitrary points, but for complex geometries, they still rely on conditioning over discretized domains ( 38 ; 2 ; 1 ) . Due to these limitations, learning-based multiple-scattering methods have been restrained to two-dimensional problems only. Inspired by methods based on the BEM that learn the boundary solution instead of solving the BIE to alleviate the BEM bottleneck ( 25 ; 11 ) , we propose, PIBNet. It is a learning-based method to simulate PDEs that can be solved with the BEM i.e. linear and semi-linear PDEs. More specifically, in this paper, we address multiple scattering due the inherent challenges of such problems. We first present a benchmark including simulation datasets that focus on 3D exterior scattering problems for both Helmholtz and Laplace problems under Dirichlet or Neumann boundary conditions, as well as meaningful metrics to evaluate the estimated solution. Neural fields, which have frequently been used to solve similar tasks ( 25 ; 11 ) , are not suitable here due to the presence of multiple disjoint obstacles. Since efficiently capturing long-range interactions between obstacles is crucial, a natural approach is to explore transformer-based methods developed for solving PDEs or point cloud processing. However, as shown numerically in section 6.2 , these approaches exhibit limited performance. To improve the accuracy, we propose PIBNet, a method in which distant interactions are explicitly represented as edges in graphs. To avoid the computational cost of adding all possible interaction edges, we propose a physics-informed strategy for connecting distant points, combined with a U-Net–style multi-scale graph neural network (GNN) based on MeshGraphNet ( 12 ) that approximate the boundary solution to multiple scattering problems. This approach outperforms existing state-of-the-art learning-based methods dedicated to solving PDEs. Finally, we show the ability of our architecture to generalize to environments with up to three times more obstacles than those seen in the training set. In summary, we propose the following contributions: 1. A benchmark of datasets of exterior Laplace and Helmholtz problems with multiple disjoint obstacles obtained using the BEM. 2. PIBNet, our method that combines a physics-inspired strategy for modeling long-range interactions as graphs and a new multiscale GNN architecture for estimating the boundary solution of the multiple scattering problems. (a) (b) (c) Figure 1: Illustrations of the resulting field u (dashed arrows) from the scattering of an incoming wave u_{\mathrm{inc}} (solid arrows) by two obstacles \Omega_{1} and \Omega_{2} (a). In (b), the obstacles are closer to each other and in (c), there is a third obstacle \Omega_{3} . This induces more complex interactions between u_{\mathrm{inc}} and obstacles. 2 Preliminaries on the Boundary Element Method This section provides a comprehensive overview of the boundary element method (BEM) ( 4 ) through the problem of multiple scattering ( 28 ) of an incident wave u_{\text{inc}} . Let \Omega=\bigcup_{i=1}^{n}\Omega_{i}\in\mathbb{R}^{3} be the union of n closed bounded sets \Omega_{i} , 1\leq i\leq n , representing obstacles that do not intersect. We assume that each set \Omega_{i} has a Lipschitz-continuous and piecewise-smooth boundary \Gamma_{i} and let \Gamma=\bigcup_{i=1}^{n}\Gamma_{i} . The homogeneous Helmholtz equation is given by: \left\{\begin{aligned} \mathcal{L}u&=0&&\text{in }\mathbb{R}^{3}\setminus\Omega\\ u&=-u_{\text{inc}}&&\text{on }\Gamma\\ \end{aligned}\right. (1) where \mathcal{L}=\Delta+k^{2} with k the wavenumber. We assume the Sommerfeld radiation condition: \lim_{|\mathbf{x}|\rightarrow\infty}|\mathbf{x}|\big(\frac{\partial}{\partial|\mathbf{x}|}-\mathrm{i}k\big)u(\mathbf{x})=0 is satisfied, which ensures that no energy is radiated from infinity. Therefore, the total field is given by u_{\text{tot}}=u+u_{\text{inc}} . The BEM relies on a reformulation of eq. 1 as a boundary integral equation. The key ingredient in this reformulation is the Green’s function G , defined as the solution of \mathcal{L}G=\delta where \delta denotes the Dirac delta function. For our problem, the variational form of the boundary integral equation can be defined as follows: \int_{\Gamma}u(\mathbf{x})q(\mathbf{x})d\mathbf{x}=\int_{\Gamma\times\Gamma}G(\mathbf{x}-\mathbf{y})q(\mathbf{x})p(\mathbf{y})d\mathbf{x}d\mathbf{y} (2) where q is a test function and p is the unknown. The solution of this equation gives only the trace of the density p . The second step of the method consists in applying the boundary integral representation to compute the scattered field in the volume: u(\mathbf{x})=\mathcal{S}(p)(\mathbf{x}),\quad\mathbf{x}\in\mathbb{R}^{3}\setminus\Omega (3) where \mathcal{S}(p)(\mathbf{x})=\int_{\Gamma}G(\mathbf{x}-\mathbf{y})p(\mathbf{y})d\mathbf{y} is the single layer potential operator. After discretization of \Gamma , eq. 2 becomes a fully populated linear system, with storage and solution complexities of O(N^{2}) and O(N^{3}) , respectively. Instead of direct solvers, iterative solvers such as GMRES ( 36 ) are often employed for BIEs. The number of iterations needed for convergence provides insight into the intrinsic difficulty of the problem, such as the level of multiple scattering occurring between obstacles in the given configurations. Fast algorithms like the Fast Multipole Method or Hierarchical

Read the complete text