Pith. sign in

REVIEW 3 major objections 5 minor 47 references

Near-Field Sampling for Line Sources

T0 review · 3 major / 5 minor · reviewed 2026-08-01 · deepseek-v4-flash

Pith's one-line read Near-field sampling for line sources reaches the SVD-optimal bound using geometry alone.

desk verdict Solid geometry-based sampling scheme; the propagating-mode part is a genuine contribution, but the post-knee reactive gain leans on a parameter picked post hoc, so the outperformance claim needs robustness tests. read the letter →

arxiv 2607.20701 v2 pith:YUD2WFYZ submitted 2026-07-22 eess.SP

classification eess.SP
keywords near-fieldsamplingdegreesoffreedomDoFdensityviewlengthmutualshadowreactivemodessingularvalueslinesources
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 asks where to place a fixed number of near-field receivers so that the field between a line source and a line receiver is captured with the least error. Its answer: place one receiver per spatial degree of freedom, at equal increments of the cumulative view-length density—a quantity computed from the source/receiver geometry alone—and then add extra samples near the receiver edges according to a power-law 'reactive' density that models modes beyond the spectral knee. The paper establishes that any sampling scheme's error is bounded below by the first omitted singular value of the continuous channel operator, and its numerical tests show the proposed geometry-aware scheme tracking that bound across parallel, perpendicular, non-coplanar, and gently curved line pairs, outperforming both uniform sampling and propagating-DoF-only sampling after the knee. The practical payoff is a nonuniform sampling prescription that needs no precomputed channel matrix or SVD yet approaches what SVD-optimal sampling achieves.

What carries the argument

View-length DoF density: ρ_p(r_R) = (1/λ) ∫_{Tx} |(lhat_T×R)·(lhat_R×R)|/|R|³ dl_T, whose cumulative integral L_TR defines the 'DoF coordinate' in which uniform spacing means one sample per DoF. This density is the mechanism that maps geometry to sampling locations; it gives the total mode count, the local concentration of propagating modes, and the one-DoF placement rule (13). The reactive correction is a power-law weight q_r(f)=f^{-p}+(1−f)^{-p} in that same coordinate, used only after the knee to concentrate extra samples at the receiver ends. The error functional E(P_N)=||H(eH†eH−I)||_2=||Σ(V^H eV eV^H V−I)||_2 provides the Eckart–Young lower bound that the scheme is calibrated against.

What would settle it

Compute the channel singular values for a line pair with strong curvature or a near-90° receiver bend (the paper's own 'excessive bending' caveat), and compare the spectral transition to ⌈L_TR/λ⌉ from (11); a mismatch beyond a few modes would place the view-length count on the wrong side for 3D curves. A cleaner check: on any tested geometry, plot E(P_N) with the reactive term removed—if the gap to σ_{N+1}/σ1 fails to close as N grows, the reactive-density model is not the right correction.

Watch

Extended reading notes

Core claim

Near-field samples for line sources can be chosen directly from geometry. The paper's central claim is that the number and placement of significant spatial modes are governed by the view length between the source and receiver curves, L_TR = ∫∫ |(lhat_T×R)·(lhat_R×R)|/|R|³ dl_R dl_T, with NDoF = L_TR/λ. Sampling at points ξ_n satisfying L_TR(ξ_{n+1})−L_TR(ξ_n)=λ (one sample per degree of freedom) captures the dominant propagating subspace; beyond the NDoF knee, additional samples are placed according to the reactive density ρ_r ∝ [f(ξ)^{-p}+(1−f(ξ))^{-p}]ρ_p(ξ) with p=2/3, where f is the DoF-normalized cumulative coordinate. The sampling error E(P_N) is shown to be at least σ_{N+1}, the first

Load-bearing premise

The load-bearing assumption is that one geometric integral—the view length between the source and receiver lines—continues to count the spatial modes correctly when the lines are non-coplanar or gently curved in 3D; if that count is wrong, the knee location and every sampling point derived from it shift.

Editorial extensions

If this is right

  • For line-source/receiver geometries, the number of significant samples can be set by NDoF = L_TR/λ and the locations read off from one cumulative integral, without building the channel matrix.
  • Sampling schemes that use only the propagating DoF density will converge slowly after the spectral knee; the reactive edge density is needed to match the SVD bound.
  • The error of any N-point near-field sample set is at least the (N+1)-th singular value, so a good sampling rule should target that bound; the proposed rule does so in the tested geometries.
  • The placement rule transfers across parallel, perpendicular, rotated, non-coplanar, and moderately curved configurations, suggesting it depends only on the view-length structure, not on coordinate-system details.
  • In antenna measurement/inverse-source settings, this implies geometry-adaptive nonuniform sampling grids can replace dense uniform grids while preserving dominant modes.

Reading between the lines

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

  • If the view-length count also holds for two-dimensional surfaces (a suggested future direction), the same one-DoF placement idea would become a surface sampling rule, potentially unifying near-field sampling, beamforming, and inverse-source discretization under one geometric functional.
  • The reactive power-law exponent p=2/3 is chosen empirically from two examples; an editorial extension would be to test whether p depends on edge geometry, source length, or separation, or whether an adaptive estimator of post-knee mode energy could replace the fixed exponent.
  • The scalar Green's function idealization leaves polarization coupling and multiple field components unexplored; if the vector extension preserves the density interpretation, the sampling rule could be applied directly to full electromagnetic measurements.
  • A practical near-field measurement campaign could validate the rule by comparing reconstruction error against uniform sampling with the same sample count; the paper's numerical setup offers a ready protocol for such a test.
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 / 5 minor

Summary. The paper develops a geometry-aware sampling framework for near-field line-source systems. Starting from a scalar Green's-function channel model, it defines a sampling error functional (7) and proves that it is lower-bounded by the first neglected singular value of the dense channel (8). It then derives a propagating-mode DoF density from the view-length (mutual-shadow) integral (11), uses equal increments of the cumulative density to place one-DoF samples (13), and augments this with a heuristic reactive-mode density (23)-(25) to improve post-knee sampling. Numerical experiments for parallel, perpendicular, non-coplanar, and gently curved line pairs compare the proposed strategy against uniform, DoF-center, DoF-edge, DEIM, and the singular-value bound, reporting that the method closely tracks the SVD-optimal performance and outperforms propagating-only sampling.

Significance. If the claims are robust, the framework offers a computationally cheap, geometry-driven alternative to matrix-based sample selection (e.g., DEIM) for near-field line sampling, with a rigorous operator-side error bound. The derivation of the propagating density as the total variation of the projected-direction function (17)-(19) is clean and elegant, and the numerical validation spans a wider set of 3D geometries than is often considered. However, two load-bearing points need attention before the claims can be fully accepted: the 3D view-length formula (11) is asserted without derivation, and the reactive-density exponent p=2/3 is selected post hoc on the same benchmark used to demonstrate the method's advantage.

major comments (3)
  1. [Sec. III-B, Eq. (11), and Sec. IV] The 3D view-length formula (11) is carried over from the 2D mutual-shadow result (9) without a derivation. The manuscript itself states, 'The generalized expression (11) for curves in R^3 is validated numerically in Sec. IV.' Since (11) feeds directly into the density (14)-(20), the one-DoF rule (13), and the knee location for all non-coplanar and curved examples, its status is load-bearing. The scope caveat 'provided that the curves do not exhibit excessive bending' is heuristic. Please either derive (11) from a local-tangent/infinitesimal-segment argument and state the regime of validity, or explicitly label it as a conjecture supported by numerical evidence. The conclusion's phrase 'closed-form expressions were derived' overstates the current state.
  2. [Sec. III-C, Eqs. (23)-(25), and Fig. 5] The reactive-density exponent p=2/3 is selected post hoc on the same benchmark used to support the central outperformance claim: 'Among the tested choices, p=2/3 provides the clearest improvement and is therefore used in all examples throughout this paper.' Only p=0.5 and p=2/3 are shown, so the sensitivity to this free parameter is not established. Since the abstract's 'significantly outperforms' claim is specifically about the post-knee reactive correction, add a p-sweep for representative configurations, derive p from the edge-condition asymptotics of the Helmholtz operator, or otherwise show that the improvement is robust over a range of p. Without this, the claim could be an artifact of a favorable parameter choice.
  3. [Sec. III-A, Eq. (13) and DoF-edge/DoF-center definitions] The one-DoF condition (13) equates a cumulative view-length increment to exactly one wavelength, which for integer NDoF would imply NDoF+1 sample points (including both endpoints). The implementation instead uses N points with equal increments in the DoF coordinate, giving N-1 intervals for DoF-edge and N cells for DoF-center. The manuscript says 'the sampling intervals are adjusted... so that each cell represents approximately NDoF/N spatial DoF,' which is not the same as (13). This ambiguity affects the x-axis interpretation in Figs. 2 and 5-8 and the 'one-DoF' terminology. Please define exactly how N maps to the point set (N intervals vs. N-1 intervals, inclusion of endpoints) for each sampling rule.
minor comments (5)
  1. [Sec. II, Eq. (8)] The norm equality in (8) uses the unitarity of V^H; this step is not explicitly stated. A one-line justification would improve readability.
  2. [App. A, Eqs. (29) and (31)] The expressions (29) and (31) are view-length densities, whereas the DoF density in (20) includes a factor 1/λ. Please state explicitly whether the Appendix expressions omit this factor and note that the normalization in Sec. III-C cancels it.
  3. [General] The terms 'gently curved' (Sec. III) and 'monotone smoothly varying' (Sec. V) are used inconsistently; unify the terminology.
  4. [Fig. 4] The caption says 'following 25 reactive modes,' but the text in Sec. III-C mentions the first 25 post-knee modes. Clarify the definition.
  5. [Eqs. (5)-(6)] The reconstructed field is denoted bER in (5) but the error in (6) is written as bER - E_R; define bER explicitly as the reconstructed dense-field vector.

Circularity Check

1 steps flagged · score 5.0 of 10

Reactive exponent p=2/3 is selected post hoc on the same benchmark used to demonstrate outperformance, making the headline quantitative claim partially in-sample.

  1. fitted input called prediction [Section III-C, paragraph following Fig. 5 (Eqs. (23)-(25))]
    "Among the tested choices, p=2/3 provides the clearest improvement and is therefore used in all examples throughout this paper."

    The reactive density (25), which is the source of the post-knee outperformance claimed in the abstract, contains an exponent p that is not derived from edge-condition asymptotics but selected by inspecting the very Fig. 5 sampling-error curves that are then presented as evidence. With no p-sensitivity analysis, the p=2/3 curves demonstrate a tuned fit rather than a parameter-free prediction; the same benchmark cannot independently validate a parameter chosen to improve on it. The paper further states that the reactive-integral benchmark, built from the post-knee singular modes, gives even more improvement, confirming that p=2/3 is a fitted approximation rather than a derived consequence.

full rationale

The propagating-mode part of the derivation is not circular: Eq. (8) is an Eckart-Young lower bound, the DoF density (20) follows from the view-length formulation of [27], and the one-DoF placement rule is benchmarked against dense-channel SVD and DEIM, giving independent, falsifiable references. Self-citations [27], [37], and [41] are present, but the central sampling-error comparisons are anchored to the dense-channel singular-value spectrum, so they are not solely load-bearing. The genuine circularity is restricted to the reactive component: q_r(f) is an unmotivated power-law ansatz, and the exponent p=2/3 is chosen after examining the same Fig. 5 data used to support the abstract's 'closely approaches' and 'significantly outperforms' claims. Because p=0.5 also improves over DoF-edge sampling, the qualitative effect is not entirely manufactured, but the quantitative headline claim is partly an in-sample consequence of parameter selection. Hence score 5: partial circularity with the central propagating derivation retaining independent content.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

No new physical entities are introduced; the reactive-mode density ρ_r is a modeling term, not a new field or particle. The ledger's main weights are: (i) the 3D view-length functional (11), which is assumed and numerically spot-checked; and (ii) the fitted reactive exponent p=2/3. The propagating density and the Eckart–Young bound are standard or inherited with independent validation.

free parameters (3)
  • Reactive-density exponent p = p = 2/3
    Chosen post hoc: 'Among the tested choices, p=2/3 provides the clearest improvement and is therefore used in all examples throughout this paper' (Sec. III-C). Tuned against the same SVD lower bound used to evaluate the method.
  • Number of post-knee modes in diagnostic benchmark = 25 modes ("at least 75% of the identified reactive modes")
    Sec. III-C: the reactive-mode profile and the reactive-integral benchmark in Figs. 4-5 use the first 25 post-knee modes; the cutoff is ad hoc and affects the diagnostic benchmark shape, not the sampling rule itself.
  • Dense reference grid spacing = λ/50
    Sec. III-A: 'a spacing of λ/50... serves as the continuous-channel reference'. A discretization choice, not fitted to the target result.
assumptions (6)
  • domain assumption Scalar propagation model: E(r_R) = ∫ G(r_R, r_T) J(r_T) dl_T with scalar Green's function, one field component
    Stated in Sec. II ('we adopt the commonly used scalar propagation model'); vector polarization coupling is deferred to future work.
  • domain assumption NDoF = L_TR/λ, the view-length (mutual-shadow) degrees-of-freedom estimate
    Eq. (10), inherited from the authors' prior mutual-shadow theory [27]; it is the foundation connecting sample count to singular-value knee.
  • ad hoc to paper 3D generalization (11): for curves in R³ the view length is ∫∫ |(l̂_T×R)·(l̂_R×R)|/|R|³ dl_R dl_T
    Stated as a generalization of the 2D in-plane form (9) ('When the two lines are restricted to the same plane... reduces to the in-plane form') and only numerically validated: 'The generalized expression (11) for curves in R³ is validated numerically in Sec. IV.' No derivation is given for the 3D case.
  • domain assumption Full mutual visibility without occlusion; gently curved lines with gradually varying tangent directions
    Sec. III opening: 'Assuming full mutual visibility without occlusion'; Sec. I: observation line 'gently curved... provided that its local tangent direction varies only gradually.' The boundary of 'gently' is left heuristic.
  • domain assumption Singular-value spectrum separates into propagating, reactive, and noise-dominated regions (spectral knee)
    Used in Sec. III-C to justify isolating 'the first 25 post-knee modes'; classification from the authors' prior work [37].
  • standard math Eckart–Young theorem: best rank-N approximation error of H is σ_{N+1}
    Used to establish the lower bound E(P_N) ≥ σ_{N+1} in Eq. (8). Standard linear algebra.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Near-Field Sampling for Line Sources." pith.science (2026). https://pith.science/paper/YUD2WFYZ

@misc{pith2026260720701,
  author       = {Pith},
  title        = {Pith review of: Near-Field Sampling for Line Sources},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YUD2WFYZ}},
  note         = {Machine review of arXiv:2607.20701}
}
read the original abstract

Near-field sampling seeks to represent electromagnetic fields between transmitting and receiving regions using a minimal number of measurement points while preserving the dominant spatial modes. This paper develops a geometry-aware sampling framework based on spatial degrees of freedom (DoF). A view-length formulation is used to derive closed-form expressions for the propagating-mode DoF density for simple line-source geometries, providing both the total DoF and its local distribution. One-DoF sampling points are obtained from equal increments of the cumulative DoF density, yielding an adaptive nonuniform sampling strategy up to the knee of the singular-value spectrum. To improve the representation of the remaining modes beyond the knee, a reactive-mode density is introduced to guide the placement of additional edge samples. An operator-based sampling error functional is formulated and shown to be lower-bounded by the neglected singular values of the continuous channel operator. Numerical results demonstrate that the proposed sampling strategy closely approaches the optimal performance obtained from singular-value decomposition and significantly outperforms sampling based solely on the propagating-mode DoF density.

Figures

Figures reproduced from arXiv: 2607.20701 by the authors.

Figure 1
Figure 1. General geometry for the sampling problem. The transmitting region [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Sampling error (8) using uniform sampling, DoF-center, DoF-edge [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Illustration of monotonic and sign-changing cases of the projected [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Local singular-vector magnitude-squared averages and normalized [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Sampling-error comparison for the parallel- and perpendicular-line [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 7
Figure 7. Figure 7: Validation of the NDoF estimate (11) and the sampling-error bound (8) [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: Normalized sampling error of the proposed reactive-density method [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

47 extracted references · 1 linked inside Pith

  1. [1]

    Efficient near-field far-field transformation for nonredundant sampling representation on arbitrary surfaces in near-field antenna measurements,

    M. A. Qureshi, C. H. Schmidt, and T. F. Eibert, “Efficient near-field far-field transformation for nonredundant sampling representation on arbitrary surfaces in near-field antenna measurements,”IEEE Trans. Antennas Propag., vol. 61, no. 4, pp. 2025–2033, Dec. 2012

  2. [2]

    Antenna characterization from a small number of far-field measurements via reduced-order models,

    N. M ´ezi`eres, M. Mattes, and B. Fuchs, “Antenna characterization from a small number of far-field measurements via reduced-order models,” IEEE Trans. Antennas Propag., vol. 70, no. 4, pp. 2422–2430, Oct. 2021

  3. [3]

    Fast antenna far-field characterization via sparse spherical harmonic expansion,

    B. Fuchs, L. Le Coq, S. Rondineau, and M. D. Migliore, “Fast antenna far-field characterization via sparse spherical harmonic expansion,”IEEE Trans. Antennas Propag., vol. 65, no. 10, pp. 5503–5510, Aug. 2017

  4. [4]

    Singular value optimization in inverse electromagnetic scattering,

    A. Capozzoli, C. Curcio, and A. Liseno, “Singular value optimization in inverse electromagnetic scattering,”IEEE Antennas Wireless Propag. Lett., vol. 16, pp. 1094–1097, Dec. 2016

  5. [5]

    A two-step inverse-scattering technique in variable- exponent lebesgue spaces for through-the-wall microwave imaging: Experimental results,

    A. Randazzoet al., “A two-step inverse-scattering technique in variable- exponent lebesgue spaces for through-the-wall microwave imaging: Experimental results,”IEEE Trans. Geosci. Remote Sens., vol. 59, no. 9, pp. 7189–7200, Sep. 2021

  6. [6]

    Near-field transverse reso- lution in planar source reconstructions,

    M. A. Maisto, R. Pierri, and R. Solimene, “Near-field transverse reso- lution in planar source reconstructions,”IEEE Trans. Antennas Propag., vol. 69, no. 8, pp. 4836–4845, Aug. 2021

  7. [7]

    Inverse source problem and minimum-energy sources,

    E. A. Marengo, A. J. Devaney, and R. W. Ziolkowski, “Inverse source problem and minimum-energy sources,”J. Opt. Soc. Amer . A, vol. 17, no. 1, pp. 34–45, 2000

  8. [8]

    Reactive near-field reconstruction via Rayleigh–Sommerfeld integral: Ill-conditioning analysis and regulariza- tion,

    D.-H. Han and X.-C. Wei, “Reactive near-field reconstruction via Rayleigh–Sommerfeld integral: Ill-conditioning analysis and regulariza- tion,”IEEE Trans. Microw. Theory Techn., pp. 1–10, 2026

Show all 47 references
  1. [9]

    Inverse source in the near field: The case of a strip current,

    R. Solimene, M. A. Maisto, and R. Pierri, “Inverse source in the near field: The case of a strip current,”J. Opt. Soc. Amer . A., vol. 35, no. 5, pp. 755–763, 2018

  2. [10]

    Electromagnetic inverse scattering: Re- trievable information and measurement strategies,

    O. M. Bucci and T. Isernia, “Electromagnetic inverse scattering: Re- trievable information and measurement strategies,”Radio Sci., vol. 32, no. 6, pp. 2123–2137, 1997

  3. [11]

    An overview of near-field antenna measurements,

    A. Yaghjian, “An overview of near-field antenna measurements,”IEEE Trans. Antennas Propag., vol. 34, no. 1, pp. 30–45, Jan. 1986

  4. [12]

    R. J. Marks II,Advanced Topics in Shannon Sampling and Interpolation Theory. New York, NY , USA: Springer-Verlag, 1993

  5. [13]

    Communicating with waves between volumes: Evaluating orthogonal spatial channels and limits on coupling strengths,

    D. A. Miller, “Communicating with waves between volumes: Evaluating orthogonal spatial channels and limits on coupling strengths,”Appl. Opt., vol. 39, no. 11, pp. 1681–1699, Apr. 2000

  6. [14]

    IV light and information,

    D. Gabor, “IV light and information,” inProgress in Optics, vol. 1. Amsterdam, The Netherlands: Elsevier, 1961, pp. 109–153

  7. [15]

    On the degrees of freedom of scattered fields,

    O. M. Bucci and G. Franceschetti, “On the degrees of freedom of scattered fields,”IEEE Trans. Antennas Propag., vol. 37, no. 7, pp. 918– 926, Jul. 1989

  8. [16]

    Franceschetti,Wave Theory of Information

    M. Franceschetti,Wave Theory of Information. Cambridge, U.K.: Cambridge Univ. Press, 2017

  9. [17]

    Degrees of freedom and sampling representation of electromagnetic fields: Concepts and applications

    O. M. Bucci and M. D. Migliore, “Degrees of freedom and sampling representation of electromagnetic fields: Concepts and applications.” IEEE Antennas Propag. Mag., vol. 67, no. 3, pp. 10–22, Jun. 2025

  10. [18]

    Degrees of freedom in multiple-antenna channels: A signal space approach,

    A. S. Poon, R. W. Brodersen, and D. N. Tse, “Degrees of freedom in multiple-antenna channels: A signal space approach,”IEEE Trans. Inf. Theory, vol. 51, no. 2, pp. 523–536, Feb. 2005

  11. [19]

    Communicating with large intelligent surfaces: Fundamen- tal limits and models,

    D. Dardari, “Communicating with large intelligent surfaces: Fundamen- tal limits and models,”IEEE J. Sel. Areas Commun., vol. 38, no. 11, pp. 2526–2537, Nov. 2020

  12. [20]

    Capacity of the continuous-space electromagnetic channel,

    M. A. Jensen and J. W. Wallace, “Capacity of the continuous-space electromagnetic channel,”IEEE Trans. Antennas Propag., vol. 56, no. 2, pp. 524–531, Feb. 2008

  13. [21]

    Maximizing independent channels and efficiency in BTS array antennas via EM degrees of freedom,

    F. Puggelli, B. Biscontini, E. Martini, and S. Maci, “Maximizing independent channels and efficiency in BTS array antennas via EM degrees of freedom,”IEEE Trans. Antennas Propag., vol. 73, no. 6, pp. 3444–3458, Jun. 2025

  14. [22]

    Breaking the degrees-of-freedom limit of holo- graphic MIMO communications: A 3-D antenna array topology,

    S. S. A. Yuan et al., “Breaking the degrees-of-freedom limit of holo- graphic MIMO communications: A 3-D antenna array topology,”IEEE Trans. V eh. Technol., vol. 73, no. 8, pp. 11 276–11 288, Aug. 2024

  15. [23]

    Arendt, R

    W. Arendt, R. Nittka, W. Peter, and F. Steiner,Weyl’s Law: Spectral Properties of the Laplacian in Mathematics and Physics. John Wiley & Sons, Ltd, 2009

  16. [24]

    Representation of elec- tromagnetic fields over arbitrary surfaces by a finite and nonredundant number of samples,

    O. M. Bucci, C. Gennarelli, and C. Savarese, “Representation of elec- tromagnetic fields over arbitrary surfaces by a finite and nonredundant number of samples,”IEEE Trans. Antennas Propag., vol. 46, no. 3, pp. 351–359, Mar. 1998

  17. [25]

    Waves, modes, communications, and optics: A tutorial,

    D. A. Miller, “Waves, modes, communications, and optics: A tutorial,” Adv. Opt. Photon., vol. 11, no. 3, pp. 679–825, 2019

  18. [26]

    Electromagnetic degrees of freedom of an optical system,

    R. Piestun and D. A. Miller, “Electromagnetic degrees of freedom of an optical system,”JOSA A, vol. 17, no. 5, pp. 892–902, 2000

  19. [27]

    Shadow area and degrees of freedom for free-space communication,

    M. Gustafsson, “Shadow area and degrees of freedom for free-space communication,”IEEE J. Sel. Areas Inf. Theory, vol. 6, pp. 325–337, 2025

  20. [28]

    Degrees of freedom for radiating systems,

    ——, “Degrees of freedom for radiating systems,”IEEE Trans. Antennas Propag., vol. 73, no. 2, pp. 1028–1038, Feb. 2025

  21. [29]

    Efficient planar near-field measurements for radiation pattern evaluation by a warping strategy,

    M. A. Maisto, G. Leone, A. Brancaccio, and R. Solimene, “Efficient planar near-field measurements for radiation pattern evaluation by a warping strategy,”IEEE Access, vol. 9, pp. 62 255–62 265, Apr. 2021

  22. [30]

    Asymptotic study of the radiation operator for the strip current in near zone,

    R. Pierri and R. Moretta, “Asymptotic study of the radiation operator for the strip current in near zone,”Electronics, vol. 9, no. 6, p. 911, May, 2020

  23. [31]

    NDF of the near-zone field on a line perpendicular to the source,

    ——, “NDF of the near-zone field on a line perpendicular to the source,” IEEE Access, vol. 9, pp. 91 649–91 660, Jun. 2021

  24. [32]

    Dimension and sampling of the near-field and its intensity over curves,

    G. Leone, R. Moretta, and R. Pierri, “Dimension and sampling of the near-field and its intensity over curves,”IEEE Open J. Antennas Propag., vol. 3, pp. 412–424, Feb. 2022

  25. [33]

    Sampling approach for singular system computation of a radiation operator,

    R. Solimene, M. A. Maisto, and R. Pierri, “Sampling approach for singular system computation of a radiation operator,”J. Opt. Soc. Amer . A., vol. 36, no. 3, pp. 353–361, 2019

  26. [34]

    An intuitive approach to the optimal sampling of an electromagnetic field,

    M. D. Migliore, “An intuitive approach to the optimal sampling of an electromagnetic field,”Sensors, vol. 25, no. 24, p. 7591, 2025. 10

  27. [35]

    An introduction to compressive sampling,

    E. J. Cand `es and M. B. Wakin, “An introduction to compressive sampling,”IEEE Signal Process. Mag., vol. 25, no. 2, pp. 21–30, Mar. 2008

  28. [36]

    Optimizing sensing matrices for spherical near-field antenna measurements,

    A. Bangun and C. Culotta-L ´opez, “Optimizing sensing matrices for spherical near-field antenna measurements,”IEEE Trans. Antennas Propag., vol. 71, no. 2, pp. 1716–1724, Dec, 2022

  29. [37]

    Interpreting moment matrix blocks spectra using mutual shadow area,

    Y . Brick, F. P. Andriulli, and M. Gustafsson, “Interpreting moment matrix blocks spectra using mutual shadow area,”IEEE Trans. Antennas Propag., pp. 1–1, 2026

  30. [38]

    Ben-Israel and T

    A. Ben-Israel and T. Greville,Generalized Inverses: Theory and Appli- cations. New York, NY , USA: Wiley, 1974

  31. [39]

    R. F. Harrington,Field Computation by Moment Methods. New York, NY: Macmillan, 1968

  32. [40]

    The approximation of one matrix by another of lower rank,

    C. Eckart and G. Young, “The approximation of one matrix by another of lower rank,”Psychometrika, vol. 1, no. 3, pp. 211–218, 1936

  33. [41]

    Degrees of freedom and beamforming for large intelligent surfaces,

    J. Li, A. Saberkari, B. K. Lau, and M. Gustafsson, “Degrees of freedom and beamforming for large intelligent surfaces,”arXiv preprint arXiv:2606.19666, 2026

  34. [42]

    Reduced-order models for electromagnetic scattering problems,

    A. Hochman, J. F. Villena, A. G. Polimeridis, L. M. Silveira, J. K. White, and L. Daniel, “Reduced-order models for electromagnetic scattering problems,”IEEE Trans. Antennas Propag., vol. 62, no. 6, pp. 3150– 3162, Jun. 2014

  35. [43]

    The behavior of electromagnetic fields at edges,

    J. Meixner, “The behavior of electromagnetic fields at edges,”IEEE Trans. Antennas Propag., vol. 20, no. 4, pp. 442–446, Jul. 1972

  36. [44]

    The edge condition in electromagnetics,

    R. Hurd, “The edge condition in electromagnetics,”IEEE Trans. Anten- nas Propag., vol. 24, no. 1, pp. 70–73, Jan. 1976

  37. [45]

    Charge singularity at the corner of a flat plate,

    J. A. Morrison and J. Lewis, “Charge singularity at the corner of a flat plate,”SIAM J. Appl. Math., vol. 31, no. 2, pp. 233–250, 1976

  38. [46]

    Moment-method calculations on apertures using basis singular functions,

    T. Andersson, “Moment-method calculations on apertures using basis singular functions,”IEEE Trans. Antennas Propag., vol. 41, no. 12, pp. 1709–1716, Dec. 1993

  39. [47]

    Incorporation of edge conditions in moment method solutions,

    D. Wilton and S. Govind, “Incorporation of edge conditions in moment method solutions,”IEEE Trans. Antennas Propag., vol. 25, no. 6, pp. 845–850, Nov. 2003

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.