McMg: A Learned Phase-Space Multi-channel Multigrid Preconditioner for Helmholtz Equations
Jiwei Jia, Xinliang Liu, Juntao Wang, Jinchao Xu, [2pt] School of Mathematics, Jilin University, Changchun 130012, China · 2026
Read McMg: A Learned Phase-Space Multi-channel Multigrid Preconditioner for Helmholtz Equations by Jiwei Jia, Xinliang Liu, Juntao Wang, Jinchao Xu, [2pt] School of Mathematics, Jilin University, Changchun 130012, China (2026) 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-2606-30495.txt.
Introduction
Jiwei Jia, Xinliang Liu, Juntao Wang, and Jinchao Xu build a learned preconditioner for wave equations that defeat classical solvers. The Helmholtz equation governs time-harmonic waves in acoustics and electromagnetics. At high frequencies in mixed media it turns indefinite. Phase errors pollute accuracy across the domain. Scalar coarse grids discard the local phase information in errors. McMg keeps that information in extra channels.
Coarsening shrinks physical space while channels store wave packets. Each coarse node carries amplitude, phase, direction, and scattering data. Learned transfer operators and adaptive stencils move data between levels. Medium-dependent smoothers draw coefficients from the wave speed. With the medium fixed, the whole cycle stays linear. Nonlinear features are computed once and cached for reuse. Models trained on small domains transfer to larger ones directly. Progressive fine-tuning adds coarse levels for the biggest problems. Fewer iterations and less time beat strong classical baselines.
Catalog record
| CREATOR | Jiwei Jia, Xinliang Liu, Juntao Wang, Jinchao Xu, [2pt] School of Mathematics, Jilin University, Changchun 130012, China |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 87 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.30495
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
McMg: A Learned Phase-Space Multi-channel Multigrid Preconditioner for Helmholtz Equations Jiwei Jia, Xinliang Liu, Juntao Wang, Jinchao Xu, [2pt] School of Mathematics, Jilin University, Changchun 130012, China Abstract Solving heterogeneous Helmholtz equations at high wavenumbers remains challenging because the discretized operator is indefinite, pollution degrades phase accuracy, and scalar coarse-grid correction can discard the local phase and propagation-direction information carried by oscillatory errors. We propose Multi-channel Multigrid (McMg), a learned phase-space multigrid preconditioner for heterogeneous Helmholtz equations. Rather than predicting the solution directly, McMg maps residuals to corrections within an iterative framework. Its central idea is to coarsen physical space while retaining unresolved local wave information in the channel dimension: each coarse node carries a learned packet of amplitude, phase, direction, and scattering coefficients rather than a single scalar unknown. The architecture combines linear multi-channel transfer operators with locally adaptive stencils, neural PDE operators, and medium-dependent smoothers whose coefficients are generated from the wave speed. For a fixed medium, the V-cycle is linear in the residual; nonlinear physical features are computed once in a setup phase and cached, so each online iteration reduces to convolutions with fixed coefficients. We further study generalization across scales. Models trained on small domains transfer directly to larger domains and higher effective wavenumbers, and a Layer-by-Layer Progressive Finetuning (LLPF) strategy improves large-domain scalability by adding new coarse levels while finetuning only the newly introduced parameters. Numerical experiments on high-frequency, high-contrast, and large-scale three-dimensional problems demonstrate that McMg requires substantially fewer iterations and less wall-clock time than strong classical baselines, while consistently outperforming existing neural preconditioners. 1 Introduction The Helmholtz equation, which governs time-harmonic wave propagation, plays a fundamental role in a wide range of applications, including acoustics, electromagnetics, geophysics, and medical imaging. Despite its importance, computing accurate numerical solutions remains notoriously challenging, particularly in the presence of heterogeneous media and high wavenumbers. At high wavenumbers, standard discretizations suffer from the pollution effect, with a wavenumber-dependent phase error that accumulates across the domain 3 . In multilevel settings this difficulty is amplified: using the same fixed stencil on progressively coarser grids often yields larger dispersion errors, leading to phase mismatches between levels 44 . Meanwhile, the discretized Helmholtz operator is strongly indefinite and may be close to singular in regimes of practical interest, which breaks the assumptions behind classical multigrid. Standard relaxations then cease to be reliable smoothers, and coarse-grid correction may no longer approximate the dominant error components. As a result, multigrid iterations can stagnate or diverge, with the failure often more pronounced in heterogeneous media 15 . These challenges have motivated a broad class of Helmholtz-specific solvers and preconditioners. Within the multigrid family, shifted-Laplacian methods 14 use a complex-shifted operator to add damping, improving spectral properties and yielding effective Krylov preconditioners. Wave-ray multigrid 5 ; 32 augments standard multigrid with ray-based corrections to represent oscillatory components that cannot be resolved on very coarse grids. Beyond multigrid, the Convergent Born Series (CBS) 36 constructs a modified Born expansion based on a damped reference medium and a suitable preconditioner, yielding a provably convergent iteration even for large, strongly scattering inhomogeneous media. Domain decomposition methods 7 ; 26 couple local subdomain solves through inter-subdomain transfer mechanisms, enabling efficient sweeping-type solvers and scalable implementations. Despite these advances, achieving robustness at high wavenumbers and in strongly heterogeneous media without sacrificing efficiency remains an open problem 15 . A useful way to state the multigrid obstruction is that scalar coarsening is not well matched to the phase-space structure of high-frequency waves. After relaxation, a Helmholtz error is not generally smooth in x ; in the high-frequency regime it is more naturally described by a local WKB/geometrical-optics ansatz a(x)e^{i\phi(x)} , with wave vector \nabla\phi(x) near the characteristic shell |\xi|\approx k(x) 2 ; 13 . Coarsening only the physical coordinate x therefore discards local direction, phase, and scattering information that remains essential for correction. This suggests a different coarse object: a coarse node should carry several local wave-packet or phase-space coefficients, not a single scalar degree of freedom 40 . Recently, there has been growing interest in leveraging deep learning for solving partial differential equations (PDEs). Physics-Informed Neural Networks (PINNs) 39 incorporate the governing equations into the training objective and have been applied to a variety of forward and inverse problems, but they typically require retraining for each new parameter configuration and thus generalize poorly across parametrized PDE families. In contrast, neural operator methods 1 aim to learn mappings between infinite-dimensional function spaces. Representative architectures include DeepONet 33 , which combines a branch network encoding the input function and a trunk network encoding spatial coordinates, and the Fourier Neural Operator (FNO) 29 , which performs global spectral convolutions to capture long-range dependencies. Subsequent extensions further broaden applicability, such as Geo-FNO 28 for irregular geometries, and MgNO 19 with a multigrid-inspired hierarchical design; attention-based operator learning has also been explored, e.g., Galerkin-type attention 6 . Although neural operator methods overcome the generalization limitations of PINNs and enable direct inference, they still face challenges in achieving high accuracy and ensuring solution reliability. This issue stems from the well-known spectral bias of neural networks 38 ; 31 , which tend to learn low-frequency components efficiently while struggling to capture high-frequency modes. To address this, recent studies have proposed Neural Solvers or Preconditioners 27 ; 41 ; 49 ; 10 ; 46 ; 43 ; 23 ; 50 . Instead of acting as end-to-end predictors, these methods embed neural networks as learned components within classical iterative frameworks, allowing for progressive error reduction. Notable examples include WAVE-ADR-NS 10 , an efficient multigrid algorithm inspired by the Wave-Ray method 32 that learns to eliminate “characteristic” error components specific to the high-wavenumber Helmholtz equation. Similarly, MGCFNN 46 constructs a hierarchical AI solver by combining a neural multigrid architecture with a FNO at the coarse level. A closely related family is hybrid methods 49 ; 24 ; 25 , which interleave classical iterative solvers with neural operators, using the network to capture the low-frequency error that the conventional smoother struggles to eliminate and thereby accelerate its convergence. While promising, many neural solvers for Helmholtz equations still face practical limitations. Although they reduce the iteration count relative to classical methods, their accuracy and convergence often degrade at high wavenumbers and in strongly heterogeneous media. Moreover, the computational cost of evaluating relatively heavy networks at each iteration can offset these gains, leaving the overall solving time as slow as, or even slower than, that of classical methods. In this work, we develop Multi-channel Multigrid (McMg), a learned phase-space multigrid preconditioner for heterogeneous Helmholtz equation. The method coarsens physical space as in a standard multigrid hierarchy, but keeps a finite-dimensional representation of unresolved phase, direction, and scattering information in the channel dimension. Our objective is to address a key limitation of existing neural preconditioners: whether efficient solution performance can be retained for high-frequency, high-contrast, and large-scale Helmholtz problems, as required in realistic industrial applications. Our main contributions are summarized as follows: • We introduce McMg, a learned phase-space multi-channel multigrid preconditioner for heterogeneous Helmholtz problems. The method is formulated as an iterative error-correction scheme rather than an end-to-end neural predictor, and learns an approximation to the inverse operator that can be used as