Pith. sign in

REVIEW 3 major objections 3 minor 64 references

Limitations of Nyquist Criteria in the Discretization of 2D Electromagnetic Integral Equations at High Frequency: Spectral Insights into Pollution Effects

T0 review · 3 major / 3 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The boundary element method can suffer its own numerical pollution at high frequency.

desk verdict A real and interesting BEM pollution result for fixed unknowns per wavelength, but the headline (ka)^{1/3} current-error rate rests on a shaky mode count; the qualitative finding survives and deserves a serious referee. read the letter →

arxiv 2505.20942 v1 pith:N5Q3QS66 submitted 2025-05-27 cs.CE

Viviana Giunzioni , Adrien Merlini , Francesco P. Andriulli This is my paper
classification cs.CE MSC 65N3845E1078M15 PACS 02.60.Nm41.20.Jb42.25.Fx
keywords boundaryelementmethodnumericalpollutionhighfrequencyelectricfieldintegralequationspectralaliasinghypersingularoperatorCalderónpreconditioningfiltering
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper challenges the common belief that the boundary element method (BEM) delivers bounded solution accuracy at fixed points per wavelength. Through a spectral analysis of integral operators on a circular cylinder, it shows that the transverse-electric electric-field integral equation (TE-EFIE) discretized with pyramid basis functions has a current error that grows as the cube root of the electrical size, (ka)^{1/3}. It further shows that the same spectral aliasing mechanism degrades the well-conditioned Calderón combined field integral equation (CCFIE) when operator products are discretized as matrix products. The paper proposes a spectral filtering strategy that zeroes the hypersingular operator's eigenvalues beyond a cutoff mode index, restoring bounded and frequency-independent error.

What carries the argument

The central object is the spectral relative error between continuous integral operators and their Galerkin discretizations on a uniform circular mesh. Its two components are the projection error (from the basis's finite bandwidth) and the aliasing error (from the periodic summation of the continuous operator's eigenvalues across the discrete spectrum). The paper tracks how the aliasing error accumulates in sums and products of operators, and how that accumulation depends on the operator-specific eigenvalue asymptotics in the hyperbolic (q << ka), transition (q ~ ka), and elliptic (q >> ka) spectral regions. The hypersingular operator N_k has eigenvalues decaying as (ka)^{-1/3} in the transition region, which makes its aliasing error grow as (ka)^{1/3}; this growth is what drives the TE-EFIE pollution.

What would settle it

A direct numerical check could compute the TE-EFIE current error at fixed n_λ over a wide range of ka (e.g., ka from 100 to 4000) on a circle and fit the slope of log r_L2 versus log(ka); if the slope departs from 1/3 and instead saturates or follows a (ka)^{1/3} times a logarithmic factor, the mode-counting assumption and the resulting rate would be falsified.

Watch

Extended reading notes

Core claim

For a perfectly conducting circular cylinder, the paper derives closed-form expressions for the spectral, current, and scattering errors of BEM-discretized integral equations. It finds that in the high-frequency regime, where the number of unknowns per wavelength is kept fixed, the aliasing spectral error of the hypersingular operator grows as (ka)^{1/3} in the transition region (indices q near ka). This growth transfers to the current and scattering errors of the TE-EFIE, which increase asymptotically as (ka)^{1/3}. The same spectral aliasing mechanism, acting through products of discretized operator matrices, causes the TE-CCFIE, despite its frequency-bounded solution operator, to also exhibit an error that increases at a rate at most (ka)^{1/3}. An ideal spectral filter that truncates the hypersingular operator's eigenvalues beyond q_lim = floor((n_λ - 1 - ε)ka) eliminates the growing aliasing component and yields bounded, frequency-independent errors for the filtered TE-EFIE and TE-CCFIE.

Load-bearing premise

The derivation of the (ka)^{1/3} current-error growth assumes that the number of hypersingular-operator eigenvalues whose modulus decays as (ka)^{-1/3} scales as ka itself, i.e., that there are O(ka) modes in the transition region; this counting of modes is stated without proof.

Editorial extensions

If this is right

  • The TE-EFIE at fixed points per wavelength is not quasi-optimal in the high-frequency limit; its current error grows as (ka)^{1/3} away from resonances.
  • The well-conditioned TE-CCFIE does not fully escape pollution: discretizing operator compositions as products of matrices reintroduces the aliasing error growth, so its current and scattering errors increase at a rate between constant and (ka)^{1/3}.
  • Spectral filtering of the hypersingular operator, zeroing eigenvalues beyond a cutoff that scales with ka, restores bounded current and scattering errors for both the TE-EFIE and TE-CCFIE.
  • The scattering error of the filtered TE-EFIE goes to zero in the high-frequency limit because the aliasing error is exactly null in the hyperbolic and transition regions after filtering.
  • The analysis provides a quantitative explanation for the resonance peaks seen in TE-EFIE and TE-CCFIE current errors, attributing them to aliasing contributions at resonant modes of the hypersingular operator.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same spectral aliasing mechanism should degrade higher-order or spline BEM discretizations of the hypersingular operator, though the growth rate might be modified by the basis's spectral decay; the paper's pyramid basis (with Fourier coefficients decaying as F_q ~ (sin(pi q/N)/(pi q/N))^2) is the concrete case analyzed.
  • The (ka)^{1/3} growth rate is tied to the Airy-type transition asymptotics of Bessel functions; a boundary layer of width O((ka)^{1/3}) in mode index would imply a different aggregate error, so the counting of O(ka) modes in the transition region is the step a reader should scrutinize.
  • A practical, geometry-agnostic filtering scheme would need a local approximation of the cutoff q_lim; the paper's ideal filter relies on the exact spectral decomposition of the circle, so an extension to non-canonical scatterers would test the mechanism's generality.
  • The filtering idea could be implemented through modified Green's functions that suppress the high-order eigenmodes of the hypersingular operator without explicit diagonalization, which would connect to existing operator-filtering approaches.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 3 minor

Summary. The paper analyzes the spectral behavior of Galerkin boundary element discretizations of two-dimensional electromagnetic integral equations on a circular perfectly conducting cylinder in the high-frequency regime with a fixed number of unknowns per wavelength. Using closed-form expressions for the eigenvalues of the continuous boundary integral operators and their matrix counterparts (Eqs. (22)-(27)), it derives high-frequency asymptotics for the spectral, current, and scattering errors of the EFIE, MFIE, and Calderón-combined (CCFIE) formulations in both TM and TE polarizations. The main claims are that the TE-EFIE exhibits a pollution effect with current error growing as O((ka)^{1/3}), that the well-conditioned TE-CCFIE nevertheless suffers from error growth due to the discretization of operator products, and that an ideal spectral filter on the hypersingular operator restores frequency-bounded errors. Numerical experiments with n_λ = 4 and 8 points per wavelength are presented for the circular cylinder.

Significance. The paper addresses an important practical question: whether BEM with a fixed points-per-wavelength rule of thumb remains accurate at high frequency. The proposed spectral machinery is transparent and gives closed-form, parameter-free expressions for the errors, and the filtering strategy is a concrete, falsifiable proposal. If the quantitative claims are correct, the paper would establish a new form of pollution in a well-conditioned integral equation and offer a cure. The paper is also honest about the cylindrical restriction. However, the central quantitative claim of O((ka)^{1/3}) growth rests on a single unsupported mode-counting assertion in Section 4.2, and the numerical scattering results reported in Section 7.3 (approximate (ka)^{1/9} growth) do not support the predicted exponent. The qualitative mechanism and the filtering benefit appear credible and are the strongest parts of the paper.

major comments (3)
  1. [Section 4.2] The last paragraph of Section 4.2 asserts that 'the number of eigenvalues of N_k decaying in modulo as (ka)^{-1/3} is proportional to the frequency.' This assertion is the key step that converts the spectral error growth of the hypersingular operator into the claimed current-error rate r_L2(Γ) = O((ka)^{1/3}) via Eq. (81). It is, however, not supported by the asymptotic expansions the paper itself uses in Section 3.2.2. From Eqs. (39)-(42), the modulus |λ_N,q| is O((ka)^{-1/3}) only in the transition layer |q - ka| = O((ka)^{1/3}) around the turning point; for q well below ka, Eq. (35) gives |λ_N,q| = O(1), and for q well above ka, Eq. (51) gives |λ_N,q| ~ q/(2ka). Hence the number of modes with |λ_N,q| ~ (ka)^{-1/3} is O((ka)^{1/3}), not O(ka). Recomputing the sums in (81) with M = O((ka)^{1/3}) gives a numerator of order O((ka)^{1/3}) and a denominator of order O(1), so r_L2(Γ) = O((ka)^{1/6}), not O((ka)^{1/3}). The same correction propagates to the rates claimed for r_H^s and r_H^s_k in the same paragraph and to the scattering-error prediction O((ka)^{1/3}) in Section 5.2; for the scattering measure of Eq. (105), the revised leading-order rate would be O(1). The authors must either provide a derivation of the stronger O(ka) mode count or revise the predicted exponents and the associated numerical comparisons.
  2. [Section 7.3] Section 7.3 reports that the scattering error of the TE-EFIE and TE-CCFIE increases 'at a rate approximately equal to (ka)^{1/9}, compatible with the expectation of O((ka)^{1/3}) (Section 5.2).' A fitted exponent of 1/9 is not compatible with a predicted exponent of 1/3; in an asymptotic analysis the fitted slope should approach the predicted value as the frequency increases. This discrepancy is a direct indicator that the asymptotic rates in Sections 4.2 and 5.2 are not confirmed by the numerical experiments, and it is not resolved by the narrative in Section 7.3. The authors should fit the current and scattering error slopes over a wide range of ka and compare them with the theoretical rates; if the theory is corrected as in the previous comment, the numerical results may in fact be consistent with a lower exponent, which should be stated honestly.
  3. [Section 4.2] The phrase 'spectral shape invariance in frequency' is introduced in Section 4.2 without definition and is used as the justification for the O(ka) mode count. Since this assertion is load-bearing for the central quantitative claims, a precise formulation of the invariance (for example, in terms of the scaled variable (q - ka)/(ka)^{1/3}) and a derivation of the resulting mode count are required. Alternatively, the paper could compute the mode count directly from the uniform asymptotic expansions cited in Eqs. (39)-(42).
minor comments (3)
  1. [Eqs. (39)-(42)] The symbols Σ_α, Σ_β, Σ_γ, and Σ_δ are introduced as 'real quantities' without explicit definitions; please provide their definitions or point to the corresponding equations in the cited reference [1].
  2. [Abstract] The abstract promises 'rigorous spectral analysis'; the manuscript should align this wording with the level of derivation provided, particularly for the mode-counting step in Section 4.2.
  3. [Section 7.3] The statement that the measured rate 'approximately equal to (ka)^{1/9}' is 'compatible with the expectation of O((ka)^{1/3})' is logically misleading; a quantitative comparison with a fitted exponent and a stated frequency range would be more informative.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular reduction; the spectral derivation is self-contained, with numerical validation as an internal consistency check rather than an independent calibration.

full rationale

The derivation chain is self-contained and non-circular. The discrete eigenvalues in Eq. (27) follow exactly from the circulant structure of the Galerkin matrices on a uniform circular mesh, with F_q defined in Eq. (28); the spectral relative error (29)-(31), the composition rules (32)-(33), and the current and scattering error coefficients (78)-(80), (99)-(102) are algebraic consequences of those definitions, not fitted quantities. The high-frequency asymptotics in Section 3 use standard Bessel and Hankel expansions, and Section 4.2 carries those rates through the sums in (81)-(83). No parameter is fitted to the numerical results; the Calderón shift constant 0.4 is imported from prior literature and mainly affects CCFIE conditioning, not the TE-EFIE rate. The numerical comparisons in Section 7 are consistency checks: the 'predicted' curves evaluate the exact alias-sum formula (27), while the 'numerical' curves come from an independent MoM implementation, so agreement verifies the spectral model rather than calibrating a free constant. Self-citations (preliminary results [33], CCFIO stabilization [3,22], filtering [48]) are background or independently verifiable, e.g., the numerical conditioning results in Fig. 7. The potential concern about the O(ka) count of transition modes in Section 4.2, whose standard Airy-type width would be O((ka)^{1/3}), and the reported (ka)^{1/9} scattering increase in Section 7.3, bear on asymptotic correctness rather than on circularity; they do not arise from a parameter fit and do not reduce any prediction to its input.

Assumptions & free parameters 2 free parameters · 4 assumptions · 1 invented entities

The central derivation relies on the known spectral decomposition of the integral operators on a circle, the exact aliasing formula for the discrete matrices, and standard Bessel asymptotics. The only ad hoc assumption that is load-bearing for the claimed error rates is the counting of transition-region modes in Section 4.2. The proposed filter is a constructive entity without independent external evidence. No physical free parameters are fitted in this paper.

free parameters (2)
  • Calderón complex wavenumber shift coefficient = 0.4 in k̃ = k - j 0.4 k^{1/3} a^{-2/3}
    Used in the CCFIE preconditioner (Eqs. 11-12, 18-19). The coefficient is imported from prior literature (Antoine, Darbas, Boubendir-Turc) where it was empirically chosen as optimal. It is not fitted in this paper and does not affect the TE-EFIE analysis.
  • Filter cutoff margin ε = small positive, ε > 0
    Appears in q_lim = floor((n_λ - 1 - ε)ka) in Eq. (107). Any small positive ε suppresses the aliasing in the transition region; the exact value is not optimized and the argument is asymptotic.
assumptions (4)
  • standard math The operators S, D, D*, N on a circle are simultaneously diagonalized by Fourier modes, with eigenvalues (22)-(24).
    Standard spectral theory of boundary integral operators on circular domains, cited to Hsiao-Kleinman and Warnick-Chew. Invoked throughout Sections 2-3.
  • domain assumption The discrete BEM matrix eigenvalues are exactly given by the aliasing formula (27), assuming exact quadrature.
    Equation (27) is stated as exact for matrices obtained in infinite precision/infinite accuracy. The numerical implementation uses 100-point Gauss-Legendre quadrature, making the assumption approximate.
  • standard math The high-frequency behavior of the operator eigenvalues in the three spectral regions is governed by the cited Bessel function asymptotics.
    Large-argument and large-order expansions from Abramowitz-Stegun and NIST are used in Section 3.2 without proof.
  • ad hoc to paper Spectral shape invariance in frequency for the lossless TE-EFIE: the number of eigenvalues of N_k decaying in modulus as (ka)^{-1/3} is proportional to ka.
    Invoked in Section 4.2 to count O(ka) terms in the transition-region sums. This is the load-bearing counting step and is not derived from the Bessel asymptotics, which suggest an O((ka)^{1/3}) boundary layer.
invented entities (1)
  • Ideally filtered hypersingular operator N_k^F
    purpose: Cures the pollution by zeroing the eigenvalues of the hypersingular operator beyond a cutoff q_lim, eliminating the aliasing contributions that cause the (ka)^{1/3} error growth.
    Defined in Eq. (106) with cutoff (107). Its beneficial effect is demonstrated only within the paper's own spectral model and numerical experiments; no independent physical or experimental evidence is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Limitations of Nyquist Criteria in the Discretization of 2D Electromagnetic Integral Equations at High Frequency: Spectral Insights into Pollution Effects." pith.science (2026). https://pith.science/paper/N5Q3QS66

@misc{pith2026250520942,
  author       = {Pith},
  title        = {Pith review of: Limitations of Nyquist Criteria in the Discretization of 2D Electromagnetic Integral Equations at High Frequency: Spectral Insights into Pollution Effects},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/N5Q3QS66}},
  note         = {Machine review of arXiv:2505.20942}
}
read the original abstract

The use of boundary integral equations in modeling boundary value problems-such as elastic, acoustic, or electromagnetic ones-is well established in the literature and widespread in practical applications. These equations are typically solved numerically using boundary element methods (BEMs), which generally provide accurate and reliable solutions. When the frequency of the wave phenomenon under study increases, the discretization of the problem is typically chosen to maintain a fixed number of unknowns per wavelength. Under these conditions, the BEM over finite-dimensional subspaces of piecewise polynomial basis functions is commonly believed to provide a bounded solution accuracy. If proven, this would constitute a significant advantage of the BEM with respect to finite element and finite difference time domain methods, which, in contrast, are affected by numerical pollution. In this work, we conduct a rigorous spectral analysis of some of the most commonly used boundary integral operators and examine the impact of the BEM discretization on the solution accuracy of widely used integral equations modeling two-dimensional electromagnetic scattering from a perfectly electrically conducting cylinder. We consider both ill-conditioned and well-conditioned equations, the latter being characterized by solution operators bounded independently of frequency. Our analysis, which is capable of tracking the effects of BEM discretization on compositions and sums of different operators, reveals a form of pollution that affects, in different measures, equations of both kinds. After elucidating the mechanism by which the BEM discretization impacts accuracy, we propose a solution strategy that can cure the pollution problem thus evidenced. The defining strength of the proposed theoretical model lies in its capacity to deliver deep insight into the root causes of the phenomenon.

Figures

Figures reproduced from arXiv: 2505.20942 by the authors.

Figure 1
Figure 1. Predicted values (from formulae (31)) of projection and aliasing spectral error in the transition region for diverse operators for varying 𝑘𝑎 with 𝑛𝜆 = 4. 7.1. Spectral error results In the first set of numerical results, we analyze the high-frequency behavior of the spectral relative error of the four standard operators (31), to confirm that the trends predicted from the asymptotic expansions of the Bessel function… view at source ↗
Figure 2
Figure 2. Projection and aliasing spectral error in the transition region for diverse operators for varying [PITH_FULL_IMAGE:figures/full_fig_p017_2.png] view at source ↗
Figure 3
Figure 3. Spectral error of the CCFIO in the transition region for varying [PITH_FULL_IMAGE:figures/full_fig_p017_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Comparison between predicted (p.) and numerical (n.) results for the [PITH_FULL_IMAGE:figures/full_fig_p018_4.png]
Figure 5
Figure 5. Figure 5: Comparison between predicted (p.) and numerical (n.) results for the [PITH_FULL_IMAGE:figures/full_fig_p018_5.png]
Figure 6
Figure 6. Figure 6: Comparison between predicted (p.) and numerical (n.) results for the [PITH_FULL_IMAGE:figures/full_fig_p019_6.png]
Figure 7
Figure 7. Figure 7: 𝐿2 norm of the CCFIO and its inverse, and 𝐿2 norm of the CCFIO matrix evaluated with 𝑛𝜆 = 4 and its inverse, for the TM (a) and TE (b) formulations [PITH_FULL_IMAGE:figures/full_fig_p019_7.png]
Figure 8
Figure 8. Figure 8: Comparison between predicted (p.) and numerical (n.) results for the [PITH_FULL_IMAGE:figures/full_fig_p020_8.png]
Figure 9
Figure 9. Figure 9: Comparison between predicted (p.) and numerical (n.) results for the [PITH_FULL_IMAGE:figures/full_fig_p021_9.png]
Figure 10
Figure 10. Figure 10: Comparison between predicted (p.) and numerical (n.) results for the [PITH_FULL_IMAGE:figures/full_fig_p021_10.png]
Figure 11
Figure 11. Figure 11: 𝐿2 (a), 𝐻𝑠 (b), 𝐻𝑠 𝑘 (c) current error and 𝐿2 scattering error for the TE-EFIE and TE-CCFIE built with the ideally filtered operator N𝑘 𝐹 for varying 𝑘𝑎 with 𝑛𝜆 = 4. As far as the scattering error—which is less prone to the projection spectral error than the current e…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

64 extracted references · 38 canonical work pages

  1. [1]

    Stegun.Handbook of Mathematical Functions: With Formulas, Graphs, and Mathematical Tables, volume 55

    Milton Abramowitz and Irene A. Stegun.Handbook of Mathematical Functions: With Formulas, Graphs, and Mathematical Tables, volume 55. Courier Corporation, 1964

    work page 1964

  2. [2]

    Adrian, A

    S.B. Adrian, A. Dely, D. Consoli, A. Merlini, and F.P. Andriulli. Electromagnetic integral equations: Insights in conditioning and precondi- tioning.IEEE Open Journal of Antennas and Propagation, 2:1143–1174, 2021. ISSN 2637-6431. doi: 10.1109/OJAP.2021.3121097

    work page arXiv 2021

  3. [3]

    Onthehighfrequencybehaviorandstabilizationofapreconditionedandresonance-freeformulation

    F.P.Andriulli,I.Bogaert,andK.Cools. Onthehighfrequencybehaviorandstabilizationofapreconditionedandresonance-freeformulation. In2015 International Conference onElectromagnetics in Advanced Applications (ICEAA), pages 1321–1324, Torino, Italy, September 2015. IEEE. ISBN 978-1-4799-7806-9. doi: 10.1109/ICEAA.2015.7297331

    work page arXiv 2015

  4. [4]

    Antoine and M

    X. Antoine and M. Darbas. Integral equations and iterative schemes for acoustic scattering problems. 2016

    work page 2016

  5. [5]

    Antoine, M

    X. Antoine, M. Darbas, and Y.Y. Lu. An improved surface radiation condition for high-frequency acoustic scattering problems.Computer Methods in Applied Mechanics and Engineering, 195(33-36):4060–4074, July 2006. ISSN 00457825. doi: 10.1016/j.cma.2005.07.010

    doi:10.1016/j.cma.2005.07.010 2006

  6. [6]

    A. K. Aziz, R. B. Kellogg, and A. B. Stephens. A two point boundary value problem with a rapidly oscillating solution.Numerische Mathematik, 53(1-2):107–121, January 1988. ISSN 0029-599X, 0945-3245. doi: 10.1007/BF01395880

    doi:10.1007/bf01395880 1988

  7. [7]

    Babuška and J

    I. Babuška and J. M. Melenk. The partition of unity method.International Journal for Numerical Methods in Engineering, 40(4):727–758, February 1997. ISSN 0029-5981, 1097-0207. doi: 10.1002/(SICI)1097-0207(19970228)40:4<727::AID-NME86>3.0.CO;2-N

    doi:10.1002/(sici)1097-0207(19970228)40:4 1997

  8. [8]

    Babuska, F

    I. Babuska, F. Ihlenburg, E.T. Paik, and S.A. Sauter. A Generalized Finite Element Method for solving the Helmholtz equation in two dimensions with minimal pollution. 1995

    work page 1995

Show all 64 references
  1. [9]

    I.Babuška,F.Ihlenburg,T.Strouboulis,andS.K.Gangaraj. AposteriorierrorestimationforfiniteelementsolutionsofHelmholtz’equation— Part II: Estimation of the pollution error.International Journal for Numerical Methods in Engineering, 40(21):3883–3900, November 1997. ISSN 0029-5981,...

    1997 doi

  2. [10]

    Babuška and S.A

    I.M. Babuška and S.A. Sauter. Is the pollution effect of the FEM avoidable for the Helmholtz equation considering high wave numbers? SIAMJournalonNumericalAnalysis,34(6):2392–2423,December1997. ISSN0036-1429,1095-7170. doi: 10.1137/S0036142994269186

    doi

  3. [11]

    Arefinedgalerkinerrorandstabilityanalysisforhighlyindefinitevariationalproblems.SIAMJournalonNumerical Analysis, 45(1):37–53, January 2007

    L.BanjaiandS.Sauter. Arefinedgalerkinerrorandstabilityanalysisforhighlyindefinitevariationalproblems.SIAMJournalonNumerical Analysis, 45(1):37–53, January 2007. ISSN 0036-1429, 1095-7170. doi: 10.1137/060654177

    2007 doi

  4. [12]

    Barucq, A

    H. Barucq, A. Bendali, M. Fares, V. Mattesi, and S. Tordeux. A symmetric Trefftz-DG formulation based on a local boundary element methodforthesolutionoftheHelmholtzequation.JournalofComputationalPhysics,330:1069–1092,February2017. ISSN00219991. doi: 10.1016/j.jcp.2016.09.062

    doi

  5. [13]

    On accuracy conditions for the numerical computation of waves.Journal of Computational Physics, 59(3):396–404, July 1985

    A Bayliss, C.I Goldstein, and E Turkel. On accuracy conditions for the numerical computation of waves.Journal of Computational Physics, 59(3):396–404, July 1985. ISSN 00219991. doi: 10.1016/0021-9991(85)90119-6

    1985 doi

  6. [14]

    Betcke, E

    T. Betcke, E. Van ’T Wout, and P. Gélat. Computationally efficient boundary element methods for high-frequency Helmholtz problems in unbounded domains. In D. Lahaye, J. Tang, and K. Vuik, editors,Modern Solvers for Helmholtz Problems, pages 215–243. Springer International Publ...

    2025 doi

  7. [15]

    F.Boeykens,H.Rogier,J.VanHese,J.Sercu,andT.Boonen. Rigorousanalysisofinternalresonancesin3-DhybridFE-BIEformulationsby means of the Poincaré–Steklov operator.IEEE Transactions on Microwave Theory and Techniques, 61(10):3503–3513, October 2013. ISSN 0018-9480, 1557-9670. doi: 1...

    2013

  8. [16]

    Boubendir and C

    Y. Boubendir and C. Turc. Well-conditioned boundary integral equation formulations for the solution of high-frequency electromagnetic scattering problems.Computers & Mathematics with Applications, 67(10):1772–1805, June 2014. ISSN 08981221. doi: 10.1016/j.camwa. 2014.04.003

    2014 doi

  9. [17]

    Buffa and S

    A. Buffa and S. Sauter. On the acoustic single layer potential: Stabilization and Fourier analysis.SIAM Journal on Scientific Computing, 28 (5):1974–1999, January 2006. ISSN 1064-8275, 1095-7197. doi: 10.1137/040615110

    1974 doi

  10. [18]

    Journal of Computational Physics, 341:429–446, July 2017

    S.Chaillat,M.Darbas,andF.LeLouër.Fastiterativeboundaryelementmethodsforhigh-frequencyscatteringproblemsin3Delastodynamics. Journal of Computational Physics, 341:429–446, July 2017. ISSN 00219991. doi: 10.1016/j.jcp.2017.04.020

    2017 doi

  11. [19]

    S. N. Chandler-Wilde and D. P. Hewett. Wavenumber-explicit continuity and coercivity estimates in acoustic scattering by planar screens. Integral Equations and Operator Theory, 82(3):423–449, July 2015. ISSN 0378-620X, 1420-8989. doi: 10.1007/s00020-015-2233-6

    2015 doi

  12. [20]

    IEEE Press, New York, 1995

    Weng Cho Chew.Waves and Fields in Inhomogeneous Media. IEEE Press, New York, 1995. ISBN 978-0-470-54705-2

    1995

  13. [21]

    ISSN 0025-5718, 1088-6842

    S.H.Christiansen.DiscreteFredholmpropertiesandconvergenceestimatesfortheelectricfieldintegralequation.MathematicsofComputation, 73(245):143–167, July 2003. ISSN 0025-5718, 1088-6842. doi: 10.1090/S0025-5718-03-01581-3

    2003 doi

  14. [22]

    Onthefastdirectsolution of a preconditioned electromagnetic integral equation, April 2022

    D.Consoli,C.Henry,A.Dély,L.Rahmouni,J.E.OrtizG.,T.L.Chhim,S.B.Adrian,A.Merlini,andF.P.Andriulli. Onthefastdirectsolution of a preconditioned electromagnetic integral equation, April 2022

    2022

  15. [23]

    Darbas.Préconditionneurs Analytiques de Type Calderon Pour Les Formulations Intégrales Des Problèmes de Diffraction d’ondes

    M. Darbas.Préconditionneurs Analytiques de Type Calderon Pour Les Formulations Intégrales Des Problèmes de Diffraction d’ondes. PhD thesis, Toulouse, INSA, 2004

    2004

  16. [24]

    M. Darbas. Generalized combined field integral equations for the iterative solution of the three-dimensional Maxwell equations.Applied Mathematics Letters, 19(8):834–839, August 2006. ISSN 08939659. doi: 10.1016/j.aml.2005.11.005

    2006 doi

  17. [25]

    Erroranalysisof2-DMoMforMFIE/EFIE/CFIEbasedonthecircularcylinder.IEEETransactionsonAntennas and Propagation, 53(1):321–331, January 2005

    C.P.DavisandK.F.Warnick. Erroranalysisof2-DMoMforMFIE/EFIE/CFIEbasedonthecircularcylinder.IEEETransactionsonAntennas and Propagation, 53(1):321–331, January 2005. ISSN 0018-926X. doi: 10.1109/TAP.2004.838791

    2005

  18. [26]

    Davis and K.F

    C.P. Davis and K.F. Warnick. On the physical interpretation of the Sobolev norm in error estimation. 20(2), 2005

    2005

  19. [27]

    DispersionandpollutionoftheFEMsolutionfortheHelmholtzequationinone,twoandthree dimensions.InternationalJournal forNumericalMethodsinEngineering, 46(4):471–499, October1999

    A.Deraemaeker,I.Babuska,andP.Bouillard. DispersionandpollutionoftheFEMsolutionfortheHelmholtzequationinone,twoandthree dimensions.InternationalJournal forNumericalMethodsinEngineering, 46(4):471–499, October1999. ISSN 0029-5981, 1097-0207. doi: 10.1002/(SICI)1097-0207(19991010...

    doi

  20. [28]

    Domínguez, I

    V. Domínguez, I. G. Graham, and V. P. Smyshlyaev. A hybrid numerical-asymptotic boundary integral method for high-frequency acoustic scattering.Numerische Mathematik, 106(3):471–510, May 2007. ISSN 0029-599X, 0945-3245. doi: 10.1007/s00211-007-0071-4

    2007 doi

  21. [29]

    Franca, C

    Leopoldo P. Franca, C. Farhat, A.P. Macedo, and M. Lesoinne. Residual-free bubbles for the Helmholtz equation.International Jour- nal for Numerical Methods in Engineering, 40(21):4003–4009, November 1997. ISSN 0029-5981, 1097-0207. doi: 10.1002/(SICI) 1097-0207(19971115)40:21<...

    1997 doi

  22. [30]

    Galkowski

    J. Galkowski. Lower bounds for piecewise polynomial approximations of oscillatory functions.Journal of Approximation Theory, 305: 106100, January 2025. ISSN 00219045. doi: 10.1016/j.jat.2024.106100

    2025

  23. [31]

    Galkowski and E.A

    J. Galkowski and E.A. Spence. Does the Helmholtz boundary element method suffer from the pollution effect?, September 2022

    2022

  24. [32]

    Helmholtz boundary integral methods and the pollution effect

    J Galkowski, M Rachh, and E A Spence. Helmholtz boundary integral methods and the pollution effect

  25. [33]

    Giunzioni, A

    V. Giunzioni, A. Merlini, and F.P. Andriulli. On a high-frequency analysis of some relevant integral equations in electromagnetics. In2024 InternationalConferenceonElectromagneticsinAdvancedApplications(ICEAA),pages821–824,Lisbon,Portugal,September2024.IEEE. ISBN 9798350360974...

  26. [34]

    I. G. Graham, M. Löhndorf, J. M. Melenk, and E. A. Spence. When is the error in the h-BEM for solving the Helmholtz equation bounded independently of k?BIT Numerical Mathematics, 55(1):171–214, March 2015. ISSN 0006-3835, 1572-9125. doi: 10.1007/ s10543-014-0501-5

    2015

  27. [35]

    Gustafsson

    M. Gustafsson. Degrees of freedom for radiating systems.IEEE Transactions on Antennas and Propagation, 73(2):1028–1038, February

  28. [36]

    Harrington.Field Computation by Moment Methods

    Roger F. Harrington.Field Computation by Moment Methods. Wiley, May 1993. ISBN 978-0-7803-1014-8

    1993

  29. [37]

    Hsiao and R.E

    G.C. Hsiao and R.E. Kleinman. Error analysis in numerical solution of acoustic integral equations.International Journal for Numerical Methods in Engineering, 37(17):2921–2933, September 1994. ISSN 0029-5981, 1097-0207. doi: 10.1002/nme.1620371705

    1994 doi

  30. [38]

    Number164inAppliedMathematicalSciences.Springer,Berlin,

    GeorgeC.HsiaoandWolfgangL.Wendland.BoundaryIntegralEquations. Number164inAppliedMathematicalSciences.Springer,Berlin,

  31. [39]

    Ihlenburg and I

    F. Ihlenburg and I. Babuška. Dispersion analysis and error estimation of Galerkin finite element methods for the Helmholtz equation. International Journal for Numerical Methods in Engineering, 38(22):3745–3774, November 1995. ISSN 0029-5981, 1097-0207. doi: 10.1002/nme.1620382203

    1995 doi

  32. [40]

    Ihlenburg and I

    F. Ihlenburg and I. Babuška. Finite element solution of the Helmholtz equation with high wave number part i: The h-version of the FEM. Computers & Mathematics with Applications, 30(9):9–37, November 1995. ISSN 08981221. doi: 10.1016/0898-1221(95)00144-N

    1995 doi

  33. [41]

    Ihlenburg and I

    F. Ihlenburg and I. Babuska. Finite element solution of the Helmholtz equation with high wave number part ii: The h-p version of the FEM. SIAM Journal on Numerical Analysis, 34(1):315–358, February 1997. ISSN 0036-1429, 1095-7170. doi: 10.1137/S0036142994272337

    1997 doi

  34. [42]

    Wiley, New York, 3rd ed edition, 1999

    John David Jackson.Classical Electrodynamics. Wiley, New York, 3rd ed edition, 1999. ISBN 978-0-471-30932-1

    1999

  35. [43]

    IEEE Press, Piscataway, NJ, 2nd ed edition, 2015

    Jian-Ming Jin.Theory and Computation of Electromagnetic Fields. IEEE Press, Piscataway, NJ, 2nd ed edition, 2015. ISBN 978-1-119- 10804-7 978-1-119-10808-5

    2015

  36. [44]

    Löhndorf and J.M

    M. Löhndorf and J.M. Melenk. Wavenumber-explicit hp-BEM for high frequency scattering.SIAM Journal on Numerical Analysis, 49(6): 2340–2363, January 2011. ISSN 0036-1429, 1095-7170. doi: 10.1137/100786034

    2011 doi

  37. [45]

    Sixboundaryelementsperwavelength: Isthatenough?JournalofComputationalAcoustics, 10(01):25–51, March2002

    S.Marburg. Sixboundaryelementsperwavelength: Isthatenough?JournalofComputationalAcoustics, 10(01):25–51, March2002. ISSN 0218-396X, 1793-6489. doi: 10.1142/S0218396X02001401

    doi

  38. [46]

    S. Marburg. Numerical damping in the acoustic boundary element method.Acta Acustica united with Acustica, 102(3):415–418, May 2016. ISSN 16101928. doi: 10.3813/AAA.918958

    2016 doi

  39. [47]

    S. Marburg. A pollution effect in the boundary element method for acoustic problems.Journal of Theoretical and Computational Acoustics, 26(02):1850018, June 2018. ISSN 2591-7285, 2591-7811. doi: 10.1142/S2591728518500184. 26 V. Giunzioniet al./Journal of Computational Physics (2025)

    2025 doi

  40. [48]

    Masciocchi, E

    M.E. Masciocchi, E. Citraro, A. Dély, L. Rahmouni, A. Merlini, and F.P. Andriulli. On some operator filtering strategies based on suitably modified Green’s functions, July 2023

    2023

  41. [49]

    J. R. Mautz and R. F. Harrington. H-field, E-field, and combined-field solutions for conducting bodies of revolution.Archiv Elektronik und Uebertragungstechnik, 32:157–164, April 1978

    1978

  42. [50]

    MappingpropertiesofcombinedfieldHelmholtzboundaryintegraloperators.SIAMJournalonMathematicalAnalysis,44(4): 2599–2636, January 2012

    J.M.Melenk. MappingpropertiesofcombinedfieldHelmholtzboundaryintegraloperators.SIAMJournalonMathematicalAnalysis,44(4): 2599–2636, January 2012. ISSN 0036-1410, 1095-7154. doi: 10.1137/100784072

    2012 doi

  43. [51]

    CambridgeUniv

    FrankW.J.OlverandNationalInstituteofStandardsandTechnology,editors.NISTHandbookofMathematicalFunctions. CambridgeUniv. Press [u.a.], Cambridge, 2010. ISBN 978-0-521-19225-5 978-0-521-14063-8

    2010

  44. [52]

    Peterson, S

    A.F. Peterson, S. Ray, and R. Mittra.Computational Methods for Electromagnetics. IEEE/OUP Series on Electromagnetic Wave Theory. IEEE Press ; Oxford University Press, New York : Oxford, 1998. ISBN 978-0-7803-1122-0

    1998

  45. [53]

    C.E. Shannon. Communication in the presence of noise.Proceedings of the IRE, 37(1):10–21, January 1949. ISSN 0096-8390. doi: 10.1109/JRPROC.1949.232969

    1949

  46. [54]

    Steinbach.Numerical Approximation Methods for Elliptic Boundary Value Problems: Finite and Boundary Elements

    O. Steinbach.Numerical Approximation Methods for Elliptic Boundary Value Problems: Finite and Boundary Elements. Springer, New York, 2008. ISBN 978-0-387-31312-2 978-0-387-68805-3

    2008

  47. [55]

    Thompson and P.M

    L.L. Thompson and P.M. Pinsky. A Galerkin least-squares finite element method for the two-dimensional Helmholtz equation.International JournalforNumericalMethodsinEngineering,38(3):371–397,February1995.ISSN0029-5981,1097-0207.doi: 10.1002/nme.1620380303

    doi

  48. [56]

    Wang and Y.S

    K. Wang and Y.S. Wong. Pollution-free finite difference schemes for non-homogeneous Helmholtz equation.International Journal of Numerical Analysis and Modeling, 11(4):787–815, 2014

    2014

  49. [57]

    Warnick and Weng Cho Chew.Numerical Analysis for Electromagnetic Integral Equations

    Karl F. Warnick and Weng Cho Chew.Numerical Analysis for Electromagnetic Integral Equations. Artech House, Boston, 2008. ISBN 978-1-59693-333-0

    2008

  50. [58]

    Warnick and W.C

    K.E. Warnick and W.C. Chew. Accuracy of the method of moments for scattering by a cylinder.IEEE Transactions on Microwave Theory and Techniques, 48(10):1652–1660, October 2000. ISSN 00189480. doi: 10.1109/22.873892

    doi

  51. [59]

    Onthespectrumoftheelectricfieldintegralequationandtheconvergenceofthemomentmethod.International Journal for Numerical Methods in Engineering, 51(1):31–56, May 2001

    K.F.WarnickandW.C.Chew. Onthespectrumoftheelectricfieldintegralequationandtheconvergenceofthemomentmethod.International Journal for Numerical Methods in Engineering, 51(1):31–56, May 2001. ISSN 0029-5981, 1097-0207. doi: 10.1002/nme.144

    2001 doi

  52. [60]

    Hermann Weyl. Das asymptotische Verteilungsgesetz der Eigenwerte linearer partieller Differentialgleichungen (mit einer Anwendung auf die Theorie der Hohlraumstrahlung).Mathematische Annalen, 71(4):441–479, December 1912. ISSN 0025-5831, 1432-1807. doi: 10.1007/BF01456804

    1912 doi

  53. [61]

    Zworski.Semiclassical Analysis

    M. Zworski.Semiclassical Analysis. Number v. 138 in Graduate Studies in Mathematics. American Mathematical Society, Providence, R.I,

  54. [2008]

    ISBN 978-3-540-15284-2 978-3-540-68545-6

  55. [2012]

    ISBN 978-0-8218-8320-4

  56. [2025]

    doi: 10.1109/TAP.2024.3524437

    ISSN 0018-926X, 1558-2221. doi: 10.1109/TAP.2024.3524437

    2024

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.