Pith. sign in

REVIEW 4 major objections 3 minor 44 references

Knapp-type obstructions in multilinear fractal Fourier extension

T0 review · 4 major / 3 minor · reviewed 2026-08-03 · deepseek-v4-flash

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

Pith's one-line read For fractal measures, multilinear Fourier extension estimates hold when the convolution of the measures is integrable, and fail past a sharpened threshold when the convolution is singular.

desk verdict A clean L^p-convolution sufficient condition, plus a plausible but unproven necessary range due to a q > 2r gap. read the letter →

arxiv 2602.08568 v3 pith:7J3YZ5UN submitted 2026-02-09 math.CA math.FAmath.MG

Itamar Oliveira , Ana E. de Orellana This is my paper · ORCID
classification math.CAmath.FAmath.MG MSC 28A8042B1028A7528A78
keywords multilinearFourierextensionfractalmeasuresdecayballconditionflat-capexampleconvolutionofM-linearindependencerandomself-similarsets
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 settles part of the question of when multilinear Fourier extension estimates—bounds on the L^q norm of a product of Fourier extensions of functions on fractal measures—can go beyond what linear restriction theory gives. The positive direction (Theorem 3.1) shows that the estimate holds for all p≥1, q≥2 satisfying an explicit condition as soon as the convolution of the k measures is absolutely continuous with a density in a specific Lebesgue space; this turns convolution regularity into the sufficient mechanism. The negative direction (Theorem 3.6) constructs k random fractal measures in dimension one, with prescribed Fourier decay and ball condition, for which the estimate fails whenever q lies below a threshold that is strictly larger than the previously known necessary range whenever the sum of the support dimensions is less than 1—the regime where the convolution is singular. The counterexample is a multilinear analogue of the classical flat-cap example, with the new feature that the embedded arithmetic progressions are chosen to be linearly independent, a stand-in for transversality. If correct, the paper identifies convolution integrability as the dividing line between positive and negative regimes, and shows that singular convolutions impose a genuinely more restrictive exponent range.

What carries the argument

The central object is the convolution µ1∗...∗µk of the k measures; its Lebesgue integrability exponent q(p−1)/(q(p−1)−p) is the sufficient threshold in the positive theorem. The obstruction is built from k random fractal measures whose supports contain carefully rescaled arithmetic progressions. The common differences of these progressions are chosen to be M-linearly independent—no nontrivial integer combination with coefficients smaller than M sums to zero—which guarantees that the sumset of the progressions has full cardinality, |V_1+...+V_k|=∏|V_m|. This M-linear independence is the paper's concrete manifestation of transversality in the fractal setting; it is what prevents a loss of info

What would settle it

Construct, for some k≥2, measures satisfying the ball condition and Fourier decay with total dimension below 1 for which the multilinear estimate holds at an exponent q below the threshold of Theorem 3.6—for instance by computing the quotient in the lower-bound computation for the explicit functions and finding it bounded—and the claimed necessary condition is false. Alternatively, recalculate the Fourier-decay and ball-condition estimates for the modified sets directly: if inserting the arithmetic progressions worsens the Fourier decay exponent, the construction has no admissible test measure

Watch

Extended reading notes

Core claim

On the paper's own terms, the central claim is a pair of theorems. Theorem 3.1: if the convolution µ1∗...∗µk belongs to L^{q(p−1)/(q(p−1)−p)}(R^d), then the multilinear extension estimate R^*_{µ1,...,µk}(p×...×p→q) holds for every choice of functions. Theorem 3.6: for dimension-one measures satisfying the ball condition µ_m(B(x,r))≈r^{α_m} and Fourier decay |µ̂_m(ξ)|≲|ξ|^{−β_m/2}, with parameters subject to α_{j+1}−β_{j+1}/2 ≤ α_j−β_j/2 and (α_k−β_k/2)+(k−1)(α_1−β_1/2)<1, the estimate fails whenever q < [2p(1−Σα_m)+pΣβ_m]/[(p−1)Σβ_m]. The proof of the failure is explicit: characteristic functions of small sets built on rescaled arithmetic progressions with M-linearly independent differences

Load-bearing premise

The negative result rests on the assertion that the random fractal measures with the arithmetic progressions inserted still satisfy the Fourier decay and ball-condition estimates; the paper justifies this by referring to an earlier construction rather than proving it in detail, and if that inheritance fails the counterexample measures would not exist.

Editorial extensions

If this is right

  • If Theorem 3.1 is correct, any k-tuple of compactly supported measures whose convolution has density in L^{q(p−1)/(q(p−1)−p)} satisfies the multilinear extension estimate; this yields non-trivial examples where the convolution density is integrable but unbounded, extending the class of measures covered by earlier sufficient conditions.
  • Theorem 3.6 implies that for measures satisfying the ball condition and Fourier decay with total dimension Σα_m<1, the exponent q must lie above the new threshold—a genuinely stronger restriction than the previously known necessary condition from local dimensions alone.
  • For the bilinear case with p=2, the theorem gives q ≥ 4/(α_1+α_2) − 2, a range that includes exponents not covered by applying the linear L^2 Fourier restriction estimate to each measure separately; hence multilinear estimates can hold beyond the linear theory even when the convolution is singular.
  • The paper's Example 3.9 shows that singular-convolution measures can still satisfy nontrivial bilinear estimates via the linear theory, so the new threshold is a substantive constraint rather than a vacuous one.
  • Proposition 3.8 gives a multilinear analogue of the classical 'top lid' necessary condition, q ≥ 2d/Σ dim_B supp(µ_m), valid for general measures without Fourier decay hypotheses.

Reading between the lines

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

  • The M-linear independence condition suggests a quantitative notion of 'fractal transversality': one could test whether allowing small integer relations among the common differences degrades the exponent range continuously, rather than switching it off abruptly.
  • The positive theorem's integrability exponent is probably not optimal; the paper itself notes that applying it bilinearly to trilinear estimates loses the simultaneous interaction, so a sharper sufficient condition might depend on joint L^q structure of the convolution rather than a single Lebesgue exponent.
  • The flat-cap construction is one-dimensional; a natural testable extension is to product measures in higher dimensions, where the linear-independence condition would control vector-valued digit relations and the threshold may take a different form.
  • The paper leaves open whether the necessary threshold in Theorem 3.6 is sharp for the constructed random measures; finding a matching upper bound or a better exponent would settle this.
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

4 major / 3 minor

Summary. The paper studies k-linear Fourier extension estimates for fractal measures. Its positive result (Theorem 3.1) states that if the convolution µ1*...*µk belongs to L^{q(p-1)/(q(p-1)-p)}, then the k-linear estimate R*_{µ1,...,µk}(p×...×p→q) holds; the proof is a smoothing/Hölder argument, with applications via Shmerkin–Solomyak to self-similar measures. The negative results (Theorems 3.6 and 3.7) construct Cantor-type measures with prescribed Fourier decay and ball conditions, embedding M-linearly independent arithmetic progressions so that multilinear Knapp-type test functions make the relevant ratio diverge for q below an explicit threshold. The authors claim this threshold is sharper than Trainor's necessary condition when the sum of the α_m is below 1.

Significance. If correct, the positive theorem identifies convolution integrability as a sufficient mechanism for multilinear fractal extension, and the negative construction introduces an interesting transversality condition (M-linear independence) while improving the known necessary range in singular-convolution cases. The paper's positive proof is transparent and checkable, and the examples are illuminating. However, the main negative theorem's proof covers only q ≤ 2r for a fixed r, while the statement covers all q < q*; moreover, Proposition 5.5 relies on an unproved assertion that the modified Cantor construction inherits the required estimates. These are load-bearing issues that need a major revision, though both seem plausibly repairable.

major comments (4)
  1. [§5.3, Eq. (5.17)] The proof of Theorem 3.6 establishes divergence only under the standing restriction 1 ≤ q ≤ 2r, where r = ceil(1/Σβ_m). The theorem is stated for all q < q* := [2p(1−Σα_m) + pΣβ_m]/[(p−1)Σβ_m]. For admissible parameters q* can exceed 2r; e.g., k=2, α1=0.4, α2=0.2, β1=0.1, β2=0.05, p=1.1 gives r=7, 2r=14, and q*≈69.7, leaving the whole interval (14,69.7) uncovered. Inequality (5.17) uses Hölder in the direction requiring 2r−q ≥ 0; for q > 2r it gives no lower bound, and the divergence computation through (5.20) inherits this restriction. Either restrict the theorem to q ≤ 2r or supply a separate argument for q > 2r (for instance, choosing r depending on q and verifying that the M-LI condition in Lemma 5.3 still holds for that larger r).
  2. [§5.2, Proposition 5.5] Proposition 5.5 is the only justification that the Cantor measures constructed in §5.2 satisfy the Fourier decay (5.13) and ball condition (5.14) that are hypotheses of Theorem 3.6. The proof states that the new M-LI arithmetic progressions 'do not affect' the estimates of [10, Sections 4.4–4.5], and that the Fourier decay depends only on the τ and t parameters while the ball condition depends only on t. This is asserted rather than proved. Since the insertion of the progressions W_{N,m} constrains the choice of the sets A_{N,m}, a detailed verification—or a separate lemma—that these constraints preserve the estimates is needed. Without it, the counterexample measures may fail to exist, so this is load-bearing.
  3. [§3, Theorem 3.1] The exponent s = q(p−1)/(q(p−1)−p) is undefined for p = 1 and can be negative for 1 < p < 2 when q is close to 2. The proof uses the Hölder exponent r = pq/(q(p−1)−p), which requires q > p/(p−1) = p'. The theorem should be stated with p > 1 and q > p/(p−1), and Corollaries 3.2 and 3.3 should be adjusted accordingly. As written, the hypothesis 'µ1*...*µk ∈ L^s' is not meaningful in part of the stated parameter range.
  4. [§5.4, Theorem 3.7] The proof of Theorem 3.7 is described only as 'a routine modification' of the proof of Theorem 3.6. Since Theorem 3.7 is one of the stated main results and has a different hypothesis regime (allowing every β_m < α_m rather than a fixed relation), the omitted details are not purely cosmetic. Please provide the modification explicitly or indicate precisely which steps of the proof of Theorem 3.6 change and why the lower-bound argument remains valid.
minor comments (3)
  1. [Example 3.4] The claim that µ*ν ∈ L^p for any 1 ≤ p < 2 whenever 1 < α+β < 2 is false in general. The density is c x^{1−α−β}, so near 0 integrability in L^p requires p(α+β−1) < 1. Please correct the stated L^p range.
  2. [§2 and §5] The notation [N] is defined as {0,...,N−1} in Section 2, but in Section 5 it is sometimes used as a set of cardinality N (e.g., 'A_{N,a,m} ⊆ [ψ(N)]/Ψ(N)'). This is confusing; please make the notation consistent.
  3. [Remark 3.1] The displayed formula for Trainor's necessary condition appears garbled: it should presumably be q ≥ d p' / (Σ γ_m), not q ≥ d p' (Σ γ_m). Please fix the typography and clarify the expression.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the main positive and negative results are derived from their stated hypotheses and external constructions; the only notable concern is a proof-range gap, not circularity.

full rationale

Theorem 3.1 is self-contained: the hypothesis that the convolution lies in L^{q(p−1)/(q(p−1)−p)} arises from a Hölder exponent split in the proof, and the conclusion is the multilinear extension estimate. The proof does not assume the estimate, and the exponent is dictated by duality rather than by the desired conclusion. Corollaries 3.2 and 3.3 rely on external results of Shmerkin–Solomyak, not on the authors' own prior work. Theorem 3.6 constructs measures with prescribed Fourier decay and ball condition, then exhibits test functions whose extension quotient diverges when q is below the stated threshold; the threshold is computed from the asymptotic divergence of a product, not fitted to match a known necessary condition. Proposition 5.5 delegates the key Fourier-decay and ball-condition estimates to Chen's external construction [10], with an asserted adaptation; this is reliance on prior independent work, not self-citation, and even a terse delegation does not make the argument circular. The self-citations in the paper ([8], [16], [29]) are background material and are not load-bearing for the main theorems. No uniqueness theorem or ansatz is imported from the authors' own previous papers. A separate correctness concern, noted in the skeptical analysis, is that the proof of Theorem 3.6 derives the lower bound only for 1 ≤ q ≤ 2r, whereas the theorem is stated for all q below the threshold q*, which can exceed 2r. This is a potential gap in the proof's range, not a circular dependency; it does not affect the circularity score.

Assumptions & free parameters 2 free parameters · 6 assumptions · 0 invented entities

The central claims rest on standard harmonic analysis tools (Hausdorff–Young, Hölder) plus the external Cantor-construction technology of Chen/Hambrook–Laba. The only genuinely ad hoc choices are the θ and ϑ sequences in the counterexample construction, which are existence parameters rather than empirical fits. No new physical or mathematical objects with independent falsifiable handles are introduced.

free parameters (2)
  • θ_{N,m} = Unspecified; θ_{N,m} ∈ [1/4,1/2]∪[2,4] chosen so that (5.4) holds
    Controls the number of endpoints t_{N,m} = ψ(N)^{α_m} θ_{N,m} and hence the ball condition (5.14); introduced by hand in the Cantor construction, with existence inherited from [10].
  • ϑ_{N,m} = Unspecified; ϑ_{N,m} ∈ [1/4,1/2]∪[2,4] chosen so that (5.6) holds
    Controls τ_{N,m} = ψ(N)^{α_m−β_m/2} ϑ_{N,m} and hence the Fourier decay (5.13); chosen ad hoc in the construction to make the estimates in Theorem 3.6 work.
assumptions (6)
  • standard math Hausdorff–Young inequality for q≥2
    Used at the start of the proof of Theorem 3.1 to pass from L^q norms of products of Fourier transforms to L^{q'} norms of convolutions.
  • standard math Hölder's inequality with conjugate exponents
    Central to the proof of Theorem 3.1 and to the Cauchy–Schwarz counting step in the proof of Theorem 3.6.
  • domain assumption Existence of θ_{N,m}, ϑ_{N,m} satisfying (5.4) and (5.6)
    Section 5.1.1 asserts, following Chen [10], that the sequences can be chosen with the stated product asymptotics; this is an external technical fact the construction depends on.
  • domain assumption Chen's random Cantor construction yields measures with Fourier decay and ball condition (Proposition 5.5)
    Section 5.2 states that the modifications (M-LI arithmetic progressions) do not affect the estimates in [10, Sections 4.4–4.5]; this is the load-bearing regularity assumption for the counterexample measures.
  • domain assumption Shmerkin–Solomyak absolute continuity criterion for self-similar measures
    Used in Corollary 3.3 to produce examples of measures whose convolution is absolutely continuous, via the quoted theorem from [34].
  • domain assumption Brownian image Fourier dimension property in Example 3.9
    The example assumes the image of a set under Brownian motion has Fourier dimension equal to its Hausdorff/packing dimension; this relies on standard results in geometric measure theory.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Knapp-type obstructions in multilinear fractal Fourier extension." pith.science (2026). https://pith.science/paper/7J3YZ5UN

@misc{pith2026260208568,
  author       = {Pith},
  title        = {Pith review of: Knapp-type obstructions in multilinear fractal Fourier extension},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7J3YZ5UN}},
  note         = {Machine review of arXiv:2602.08568}
}
abstract

For curved, smooth hypersurfaces, the classical Knapp example shows that the Stein--Tomas theorem, which gives linear Fourier restriction estimates, is sharp. Variants of this example combined with the geometric notion of \textit{transversality} motivate the $L^{2}$-based multilinear Fourier extension conjecture. In the fractal setting, work by Mockenhaupt, Mitsis, and Bak-Seeger extended the linear Fourier restriction estimate beyond the smooth setting, and subsequent work showed this extension to be sharp. In this article, we construct multilinear Knapp-type examples for fractal measures inspired by the works of Hambrook--{\L}aba and Chen. This yields two necessary conditions for a fractal multilinear Fourier extension estimate to hold: one in terms of the upper box dimension of the measures' supports, and another in terms of their Fourier decay and a ball condition. These conditions give a more restrictive range compared with previously known results whenever the convolution of the underlying measures is singular. In contrast, we complement this with a result in the positive direction by establishing a multilinear Fourier extension estimate for measures whose convolution lies in an $L^p$ space. This provides a rich class of examples of `transversal' self-similar measures through the work of Shmerkin and Solomyak.

Figures

Figures reproduced from arXiv: 2602.08568 by the authors.

Figure 1
Figure 1. If Pk m=1 αm < 1, the dark grey region represents the necessary conditions of Theorem 3.6 intersected with q ≥ P 2 k m=1 αm from Proposition 3.8, since (H1) implies dimBsupp(µm) = αm. Notice that the latter are more restrictive than the conditions from Proposition 5.3 of [39] (represented by the union of the dark and light grey regions) in a setting with k measures µ1, . . . , µk satisfying (H1) and (H2). A simple c… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

44 extracted references · 2 linked inside Pith

  1. [1]

    J. Bennett. Aspects of multilinear Harmonic Analysis related to transversality, inHarmonic Analysis and Partial Differential Equations, Contemp. Math.,612, pp. 1–28 (2014)

    2014

  2. [2]

    Bennett, N

    J. Bennett, N. Bez, Higher order transversality in harmonic analysis.RIMS Kokyuroku Bessatsu,B88, pp. 75–103 (2021)

    2021

  3. [3]

    Bennett, A

    J. Bennett, A. Carbery, T. Tao, On the multilinear restriction and Kakeya conjectures.Acta Math.,126, pp. 261–302 (2006)

    2006

  4. [4]

    Bejenaru, The almost optimal multilinear restriction estimate for hypersurfaces with curvature: the case ofn−1 hypersurfaces inR n−1.Int

    I. Bejenaru, The almost optimal multilinear restriction estimate for hypersurfaces with curvature: the case ofn−1 hypersurfaces inR n−1.Int. Math. Res. Not. IMRN2022, No. 20, pp. 16363–16404 (2022)

    2022

  5. [5]

    Bourgain, Besicovitch type maximal operators and applications to Fourier analysis.Geom

    J. Bourgain, Besicovitch type maximal operators and applications to Fourier analysis.Geom. Funct. Anal., 1(2), (1991). 23 24 I. Oliveira and A. E. de Orellana

    1991

  6. [6]

    Bak and A

    J.-G. Bak and A. Seeger. Extensions of the Stein–Tomas Theorem.Math. Res. Lett.,18(4), pp. 767–781 (2011)

    2011

  7. [7]

    B´ ar´ any, K´ aroly Simon, and Boris Solomyak.Self-similar and self-affine sets and measures

    B. B´ ar´ any, K´ aroly Simon, and Boris Solomyak.Self-similar and self-affine sets and measures. Volume 276 of Mathematical Surveys and Monographs, AMS, (2023)

    2023

  8. [8]

    Carnovale, J

    M. Carnovale, J. M. Fraser, A. E. de Orellana.L 2 restriction estimates from the Fourier spectrum. Preprint, available at: arXiv:2412.14896

    arXiv

Show all 44 references
  1. [9]

    X. Chen. A Fourier restriction theorem based on convolution powers.Proc. Amer. Math. Soc.142, No. 11, pp. 3897–3901 (2014)

    2014

  2. [10]

    X. Chen. Sets of Salem type and sharpness of theL 2-Fourier restriction theorem.Trans. Amer. Math. Soc. 368(3), 1959–1977 (2016)

    1959

  3. [11]

    Demeter.Fourier Restriction, Decoupling, and Applications

    C. Demeter.Fourier Restriction, Decoupling, and Applications. Cambridge University Press; (2020)

    2020

  4. [12]

    K. J. Falconer.Fractal Geometry: Mathematical Foundations and Applications, John Wiley & Sons, Hobo- ken, NJ, 3rd. ed., (2014)

    2014

  5. [13]

    Fefferman,Inequalities for strongly singular convolution operators.Acta Math.,124, pp

    C. Fefferman,Inequalities for strongly singular convolution operators.Acta Math.,124, pp. 9–36 (1970)

    1970

  6. [14]

    Ferrante,Different perspectives on the Mizohata–Takeuchi conjecture, University of Birmingham PhD thesis 2024

    M. Ferrante,Different perspectives on the Mizohata–Takeuchi conjecture, University of Birmingham PhD thesis 2024

    2024

  7. [15]

    Fraser, K

    R. Fraser, K. Hambrook, D. Ryou. Fourier restriction and well-approximable numbers.Math. Ann.391, pp. 4233–4269 (2025)

    2025

  8. [16]

    J. M. Fraser and A. E. de Orellana. Fourier restriction for the additive Brownian sheet. Preprint, available at: arXiv:2601.05802

  9. [17]

    Guth, Restriction estimates using polynomial partitioning II.Acta Math.,221, No

    L. Guth, Restriction estimates using polynomial partitioning II.Acta Math.,221, No. 2, pp. 371–413 (2018)

    2018

  10. [18]

    Hambrook, I

    K. Hambrook, I. Laba. On the sharpness of Mockenhaupt’s restriction theorem.Geom. Funct. Anal.23, No. 4, pp. 1262–1277 (2013)

    2013

  11. [19]

    Hambrook, I

    K. Hambrook, I. Laba. Sharpness of the Mockenhaupt-Mitsis-Bak-Seeger restriction theorem in higher dimensions.Bull. Lond. Math. Soc.48(5), pp. 757–770 (2016)

    2016

  12. [20]

    Klainerman and M

    S. Klainerman and M. Machedon, Space-time estimates for null forms and the local existence theorem, Comm. Pure Appl. Math.46, No. 9, pp. 1221–1268 (1993)

    1993

  13. [21]

    Klainerman and M

    S. Klainerman and M. Machedon, Finite energy solutions of the Yang-Mills equations inR 3+1.Ann. Math. (2),142, No. 1, 39–119 (1995)

    1995

  14. [22]

    Klainerman and M

    S. Klainerman and M. Machedon, Estimates for null forms and the spacesH s,δ.Int. Math. Res. Not., No. 17, pp. 853–865 (1996)

    1996

  15. [23]

    I. Laba, M. Pramanik. Arithmetic progressions in sets of fractional dimension.Geom. Funct. Anal.,19, No. 2, pp. 429–456 (2009)

    2009

  16. [24]

    Liang and M

    Y. Liang and M. Pramanik. Fourier dimension and avoidance of linear patterns.Adv. Math.,399(108252), (2022)

    2022

  17. [25]

    Mattila.Fourier analysis and Hausdorff dimension

    P. Mattila.Fourier analysis and Hausdorff dimension. Cambridge Studies in Advanced Mathematics,150, Cambridge, (2015)

    2015

  18. [26]

    Mattila.Geometry of Sets and Measures in Euclidean Spaces: Fractals and Rectifiability

    P. Mattila.Geometry of Sets and Measures in Euclidean Spaces: Fractals and Rectifiability. Cambridge University Press; (1995)

    1995

  19. [27]

    T. Mitsis. A Stein–Tomas restriction theorem for general measures.Publ. Math. Debrecen,60, pp. 89–99 (2002)

    2002

  20. [28]

    Mockenhaupt

    G. Mockenhaupt. Salem sets and restriction properties of Fourier transforms.Geom. Funct. Anal.,10, pp. 1579–1587 (2000)

    2000

  21. [29]

    Muscalu and I

    C. Muscalu and I. Oliveira A new approach to the Fourier extension problem for the paraboloid.Analysis and PDE,17, No. 8, pp. 2841–2921 (2024)

    2024

  22. [30]

    Nazarov, Y Peres, and P

    F. Nazarov, Y Peres, and P. Shmerkin. Convolutions of cantor measures without resonance.Isr. J. Math. 187, pp. 93–116 (2012)

    2012

  23. [31]

    Oliveira e Silva

    D. Oliveira e Silva. Extremizers for Fourier restriction inequalities: Convex arcs.JAMA124, pp. 337–385 (2014)

    2014

  24. [32]

    L. Olsen. A Multifractal Formalism.Adv. Math..116, pp. 82–196 (1995)

    1995

  25. [33]

    K. Salem. On singular monotonic functions whose spectrum has a given Hausdorff dimension.Ark. Mat., 1, pp. 353–365 (1951)

    1951

  26. [34]

    Shmerkin and B

    P. Shmerkin and B. Solomyak. Absolute continuity of self-similar measures.Trans. Am. Math. Soc.,368, No. 7, pp. 5125–5151 (2016)

    2016

  27. [35]

    Stein,Harmonic analysis: real-variable methods, orthogonality, and oscillatory integrals, Princeton Material Series 43

    E. Stein,Harmonic analysis: real-variable methods, orthogonality, and oscillatory integrals, Princeton Material Series 43. Princeton University Press, New Jersey, (1993)

    1993

  28. [36]

    R. S. Strichartz, Restrictions of Fourier transforms to quadratic surfaces and decay of solutions of wave equations.Duke Math. J.44(3) pp. 705–714 (1977)

    1977

  29. [37]

    Tao, A sharp bilinear restriction estimate on paraboloids.Geom

    T. Tao, A sharp bilinear restriction estimate on paraboloids.Geom. Funct. Anal.,16, No. 6, pp. 1359–1384 (2003)

    2003

  30. [38]

    Tao,Recent progress on the Restriction conjecture

    T. Tao,Recent progress on the Restriction conjecture. Park City proceedings, (2003)

    2003

  31. [39]

    Trainor, (2019)

    C. Trainor, (2019). Multilinear restriction estimates on fractal sets (T). University of British Columbia. 24 25 Multiple convolutions and multilinear fractal Fourier extension estimates

    2019

  32. [40]

    Tomas, A Restriction theorem for the Fourier transform.Bull

    P. Tomas, A Restriction theorem for the Fourier transform.Bull. Amer. Math. Soc., (81), (1975)

    1975

  33. [41]

    T. Tao, A. Vargas, L. Vega, A bilinear approach to the Restriction and Kakeya conjectures.J. Amer. Math. Soc.,11(4), 967–1000 (1998)

    1998

  34. [42]

    H. Wang, S. Wu, Restriction estimates using decoupling theorems and two-ends Furstenberg inequalities. Preprint, available at arXiv:2411.08871

    arXiv

  35. [43]

    Math.153, 661–698 (2001)

    T, Wolff, A sharp bilinear cone restriction estimate,Ann. Math.153, 661–698 (2001)

    2001

  36. [44]

    Zygmund, On Fourier coefficients and transforms of functions of two variables.Studia Math., 50, (1974)

    A. Zygmund, On Fourier coefficients and transforms of functions of two variables.Studia Math., 50, (1974). I. Oliveira, School of Mathematics, The Watson Building, University of Birmingham, Edgbaston, Birmingham, B15 2TT, England. Email address:i.oliveira@bham.ac.uk, oliveira....

    1974

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.