A Rellich-type theorem for the Helmholtz equation in a junction of stratified media

Anne-Sophie Bonnet-Ben Dhia (POEMS); Christophe Hazard (POEMS); Sarah Al Humaikani (POEMS); Sonia Fliss (POEMS)

REVIEW 3 major objections 3 minor 20 references

A Rellich-type theorem for the Helmholtz equation in a junction of stratified media

T0 review · 3 major / 3 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read A Rellich-type uniqueness theorem holds for junctions of three stratified half-planes: the only square-integrable Helmholtz solution is zero.

desk verdict The theorem is real, the proof is solid, and the stress-test's boundedness counterexample rests on a sign error in the eigenfunction exponent. read the letter →

arxiv 2504.17345 v1 pith:MYJYSJUD submitted 2025-04-24 math.AP math.SP

Sarah Al Humaikani (POEMS) , Anne-Sophie Bonnet-Ben Dhia (POEMS) , Sonia Fliss (POEMS) , Christophe Hazard (POEMS) This is my paper · ORCID

classification math.APmath.SP MSC 35A0535J0542A3878A50

keywords HelmholtzequationstratifiedmediaopenwaveguidesRellichuniquenesstheoremgeneralizedFouriertransformtrappedmodesembeddedeigenvaluesjunctionofhalf-planes

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

The paper proves a uniqueness theorem for the two-dimensional Helmholtz equation in unbounded domains formed by gluing three stratified half-planes, where each half-plane has a wavenumber that varies only across its own boundary direction and becomes constant outside a compact region. Theorem 1.1 states that if the branch angles are at least $\pi/2$, then any square-integrable solution to $-\Delta u-k^2u=0$ is identically zero, with no boundary condition imposed on the notch between branches. This matters because such domains model junctions of open optical or acoustic waveguides, so the result rules out trapped modes and, spectrally, embedded eigenvalues in the essential spectrum. The proof follows the Rellich strategy: represent the solution in each half-plane through a modal decomposition, then use analytic continuation in the spectral variable to force every modal coefficient of the whole solution to vanish.

What carries the argument

The load-bearing object is the generalized Fourier transform $\mathcal{F}$, the unitary diagonalization of the transverse operator $A=-\partial_x^2-k^2$ on $L^2(\mathbb{R})$, whose generalized eigenfunctions $\Psi_\pm(\lambda,x)$ are scattering waves of the stratified half-plane with reflection and transmission coefficients $R_\pm(\lambda)$, $T_\pm(\lambda)$. Applied to the boundary trace of $u$, it yields the half-plane representation $u(x,y)=\sum_\pm\int_{\mathbb{R}_+}\hat\varphi_\pm(\lambda)\Psi_\pm(\lambda,x)e^{-\sqrt{\lambda}\,y}\rho_\pm(\lambda)\,d\lambda$, in which $\lambda$ is a spectral parameter. The paper then proves that the northern trace's spectral components extend meromorphically or analytically to a domain $D_n$ outside certain obstruction curves $\Lambda_{n,w}$, $\Lambda_{n,e}^\pm$, and uses the isolated zeros principle to conclude that they vanish.

What would settle it

Construct or compute a nonzero $u\in L^2(\Omega)$ satisfying $-\Delta u-k^2u=0$ distributionally in a three-half-plane junction with branch angles at least $\pi/2$; because the theorem claims none exists, any such example found by a high-resolution finite-element solve with perfectly matched layers would settle the claim false. A cheaper check is to verify numerically that the northern spectral components $\hat\varphi_{n,\pm}$ vanish to machine precision on a nontrivial parameter interval, since failure there would break the analytic-continuation conclusion.

Watch

Extended reading notes

Core claim

The central discovery is that the classical Rellich uniqueness theorem, known for homogeneous exterior domains, survives when the medium is inhomogeneous but stratified in three half-planes meeting at angles at least $\pi/2$. In the paper's own terms: if $u\in L^2(\Omega)$, $k\in L^\infty(\Omega)$ is built from three transverse stratifications, and $-\Delta u-k^2u=0$ holds distributionally in $\Omega$, then $u=0$. The proof uses a half-plane representation of $u$ in each branch, derived from a generalized Fourier transform that diagonalizes the transverse operator $-\partial_x^2-k^2$. Square-integrability kills the travelling-mode coefficients, and the vanishing of the Fourier data on an interval extends by analyticity to the whole spectral set, so the evanescent coefficients vanish as well; unique continuation then carries $u=0$ into the entire junction.

Load-bearing premise

The proof imports rather than proves the unitary diagonalization and inversion formulas of the generalized Fourier transform for stratified media; if those transforms were not unitary or did not recover $L^2$ functions by the stated inversion, the half-plane representation of $u$ would lose its foundation.

Editorial extensions

If this is right

There are no trapped modes at a junction of open waveguides whenever every angle between branches is at least $\pi/2$.
The associated self-adjoint operator has no embedded eigenvalues in its essential spectrum $[0,+\infty)$, and the uniqueness holds no matter what material occupies the notch between the branches.
The conclusion covers general bounded stratifications, including cases where the transverse operator has $L^2$ eigenfunctions; those discrete transverse modes do not contribute to the half-plane representation.
Only positivity of the squared wavenumber in the northern half-plane is needed: $k^2$ may be negative in the southern parts of the west and east branches, so non-propagating regions there cannot rescue a trapped mode.
In the three-dimensional slab geometry $\Omega\times\mathbb{R}$, the result forbids guided waves of the form $\hat u(x,y)e^{i\xi z}$ whenever $\xi^2<\min(k_{n,+}^2,k_{n,-}^2)$.

Reading between the lines

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

If the paper's conjecture is right, the restriction to angles at least $\pi/2$ is purely technical; a natural numerical test is to search for trapped modes in Y-junctions with one acute branch, where the analyticity argument no longer applies.
The obstruction curves $\Lambda_{n,w}$ and $\Lambda_{n,e}^\pm$ behave like traces of scattering resonances, and as the branch angles tend to $\pi/2$ they collapse onto the continuous-spectrum branch cut; tracking the first resonance pole as a function of branch angle could show whether it ever crosses onto the physical sheet.
The same spectral decomposition could in principle handle junctions with more than three half-planes or half-spaces, though the paper notes that its one-dimensional integrability lemma does not extend directly to higher dimensions.
Since the theorem rules out $L^2$ eigenfunctions, any trapped-mode candidate in such a junction would have to be a resonance with complex frequency, and its decay rate could be compared with Landis-type bounds.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

The result is genuinely new: a Rellich-type uniqueness theorem for the Helmholtz equation in junctions of three stratified half-planes, with no boundary condition required. The two-layer case is worked out in full detail, with explicit eigenfunctions, analyticity checks, and integrability estimates; the general stratification case is convincingly reduced to a meromorphic-continuation argument in the appendix plus an imported diagonalization theorem from Ott's thesis. The paper also states its limitations honestly: angles below π/2, three dimensions, and the possible role of trapped modes in closed waveguides are all flagged as open or excluded. No overclaiming.

The stress-test concern does not survive reading the paper. Proposition 2.1(i) is correct: for Ψ_-, the branch on x>0 is T_- e^{iβ_+ x}, not T_- e^{-iβ_+ x}. When β_+ is imaginary, iβ_+ is real negative, so the exponential decays rather than grows. The counterexample flips the sign. The second concern about analytic continuation for non-real λ is also a non-issue, since Proposition 2.1(ii) only claims boundedness on compact subsets of D×R, which follows from continuity.

Soft spots are minor. The general-stratification diagonalization theorem is imported rather than proved; for a self-contained paper that is a gap, but the two-layer case is fully justified and the external source is standard in this niche. The analytic continuation in the general-angle case lives on a domain with branch curves, and the argument only needs vanishing on a neighborhood of (0,∞), which is sufficient. The proof is somewhat dense and would benefit from a short summary of where each external input enters, but that is exposition, not substance.

The citation pattern looks healthy: self-citations to [3] are limited to the proof scheme and the conical-domain antecedent, and the generalized Fourier transform is credited to Ott and to Titchmarsh's framework. No fitting, no invented entities, no hidden parameters.

Who is this for? Researchers in mathematical wave propagation, open waveguides, and spectral theory. This is a solid, useful contribution that deserves serious referee time. I would cite it and would bring it to a reading group. Recommend sending it to peer review with no hesitation.

Referee Report

3 major / 3 minor

Summary. The paper proves a Rellich-type uniqueness theorem for the two-dimensional Helmholtz equation in unbounded domains formed as junctions of three stratified half-planes, with the angles between branches at least pi/2. The proof has two main ingredients: a generalized Fourier transform that diagonalizes the transverse operator -d_x^2 - k(x)^2 in each stratified half-plane, and an analytic continuation argument in the spectral variable that forces the spectral traces of an L^2 solution to vanish. The two-layer case is treated with explicit generalized eigenfunctions, and the general stratified case is handled through a diagonalization theorem imported from [12] and a meromorphic continuation result proved in Appendix A. The stated byproduct is the absence of trapped modes at junctions of open waveguides under the same angle condition.

Significance. The result is a natural and valuable extension of Rellich's uniqueness theorem to non-homogeneous unbounded media, with concrete consequences for the spectral theory of open waveguide junctions. The paper is clearly structured, and the two-layer calculation is instructive and mostly self-contained. However, the proof as written relies on a uniform boundedness statement for generalized eigenfunctions that is false in the total-reflection regime; this is a load-bearing gap in the analytic continuation arguments. If the gap is repaired, the theorem is likely to stand, but the current manuscript cannot be accepted without substantial revision.

major comments (3)

[Sec. 2.1, Proposition 2.1(i), Eq. (2.8)] The asserted bound |Psi_+(lambda,x)| <= 2 for all (lambda,x) in Lambda_+ times R is false. For example, take k_-=2, k_+=1, and lambda=-3 in Lambda_-=(-4,-1). Then beta_+=i sqrt(2) by (2.6), and (2.5) gives Psi_-(lambda,x)=T_- e^{-i beta_+ x}=T_- e^{sqrt(2) x} for x>0. Since |T_-|=2/sqrt(3), |Psi_-(5)| approx 1357 >> 2. The bound fails exactly on the interval -k_-^2 < lambda < -k_+^2 where one branch is evanescent. This invalidates the justification of (2.11) for phi in L^1(R) in Remark 2.4, the integrability claim in Remark 2.6, and the Fubini estimate in the proof of Corollary 2.7.
[Sec. 2.1, Proposition 2.1(ii) and Sec. 4.2, Proposition 4.4] The statement that the analytic continuations of Psi_+(lambda,x) are bounded in any compact set of D times R is also false for non-real lambda. When Im beta_+ > 0, the term e^{-i beta_+ x} is unbounded as x -> +infty, even on compact subsets of D. The same problem appears in Proposition 4.4, where the meromorphic continuations are claimed to be bounded in any compact set of (D\P) times R. This matters because the Morera-theorem arguments in Section 3 require a locally uniform bound in the spatial variable, not merely pointwise analyticity in lambda for fixed x.
[Sec. 3.1 and Sec. 3.2, analytic continuation of bphi_n^+] The core analyticity argument for bphi_n^+ is not justified by the stated estimates. In the treatment of bphi_{n,w}(lambda), the proof uses (2.8) to bound |Psi_n^+(lambda,x)| and |Psi_w^+(mu,x_w)|, but those bounds are false when one of the transverse wavenumbers is imaginary. For complex lambda in D_n, the exponential growth of the generalized eigenfunction in the direction of the evanescent branch is not controlled by the half-plane representation in the written proof, because the decay e^{-sqrt(mu) y_w} can be overcome by the growth of Psi_n^+. A repair likely requires restricting the initial identities to the interval where both beta_- and beta_+ are real and then using analytic continuation of the semiexplicit expressions, or explicitly choosing the decaying branch for evanescent modes; but as written, the claim that bphi_n^+ has a meromorphic continuation in D_n is not established.

minor comments (3)

[Sec. 1.1] The phrase 'provided the unique continuation principle in R^2\Omega holds true' is imprecise: the equation is not imposed in R^2\Omega. The intended argument is that u vanishes in H_n, and then unique continuation inside each connected half-plane H_w and H_e propagates the vanishing. This should be stated explicitly.
[Sec. 2.1, Figure 4] In Figure 4, the entry 'Undefined' for Psi_+ on the interval (-k_-^2,-k_+^2) is correct because this interval lies outside Lambda_+, but the caption or the figure legend should clarify that the table displays the behavior only on the domain of definition of each function.
[Sec. 3.2] The notation Lambda_{n,w} is used for a curve in the spectral parameter plane, which is easily confused with the half-lines Lambda_+ and Lambda_- used earlier. A different symbol, such as Gamma_{n,w}, would improve readability.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the main theorem is derived from externally cited generalized Fourier transform results and an independent analytic-continuation argument; the self-citation to [3] is only schematic and not load-bearing.

full rationale

The proof of Theorem 1.1 does not fit parameters, rename a known result, or define its objects in terms of the target conclusion. The half-plane representation (2.16) is derived from the generalized Fourier transform diagonalization theorem, cited externally from [12] and [15] (Theorem 2.3 for two-layer media, Theorem 4.1 for general stratifications), and the analytic-continuation step uses Morera's theorem together with the vanishing of the transformed traces on intervals (2.17). The conclusion bφ±_n = 0 follows from analytic continuation and the isolated zeros principle, not from assuming Theorem 1.1. The self-citation to [3] is used only to recall the proof scheme and to identify homogeneous or conical special cases where Theorem 1.1 could already be deduced; it is not an input to the general stratified-junction proof. The cited external theorems do not contain the junction Rellich statement, so they provide independent support. The skeptical concern about the boundedness estimate (2.8) in the total-reflection regime is a mathematical-validity issue rather than a circularity issue: even if that estimate needed repair, the argument would not reduce to its own inputs. Overall, the derivation is self-contained against external benchmarks, with only a minor, non-load-bearing self-citation to the authors' earlier conical-domain result.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

No free parameters, fitted constants, or invented entities; the paper is a pure mathematical proof. The main external input is the generalized Fourier transform diagonalization theorem, cited from prior work rather than reproved in full.

assumptions (5)

standard math Theorem 2.3 (and Theorem 4.1 in the general case): the generalized Fourier transform F is unitary from L2(R) to the spectral space and diagonalizes A = -partial_x^2 - k^2.
Imported from [12] and [15]; the half-plane representations (2.16) and (4.6) depend on this theorem.
standard math Unique continuation principle holds in R^2 \ Omega.
Used after showing u = 0 in H_n to conclude u = 0 in all of Omega (Sections 3.1, 3.2, 4.3).
standard math The transverse operator A has at most finitely many negative eigenvalues, and its continuous spectrum is [-max(k^2), +infty), as summarized from [16].
Used in Section 4.1 to define the spectral space L2(Lambda_+,rho_+) x L2(Lambda_-,rho_-) x C^N and to justify that eigencomponents of the trace vanish (Proposition 4.3).
standard math Morera's theorem and the isolated zeros principle for analytic functions.
Used throughout Sections 3 and 4.3 to convert local vanishing into global vanishing of the modal coefficients.
domain assumption The stratification interfaces in each half-plane lie outside the intersections of half-planes, which forces |theta_j| >= pi/2.
Needed so k is well-defined on overlaps; the paper derives this condition in Section 1.1 and it is part of the theorem's hypotheses.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A Rellich-type theorem for the Helmholtz equation in a junction of stratified media." pith.science (2026). https://pith.science/paper/MYJYSJUD

@misc{pith2026250417345,
  author       = {Pith},
  title        = {Pith review of: A Rellich-type theorem for the Helmholtz equation in a junction of stratified media},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MYJYSJUD}},
  note         = {Machine review of arXiv:2504.17345}
}

abstract

We prove that there are no non-zero square-integrable solutions to a two-dimensional Helmholtz equation in some unbounded inhomogeneous domains which represent junctions of stratified media. More precisely, we consider domains that are unions of three half-planes, where each half-plane is stratified in the direction orthogonal to its boundary. As for the well-known Rellich uniqueness theorem for a homogeneous exterior domain, our result does not require any boundary condition. Our proof is based on half-plane representations of the solution which are derived through a generalization of the Fourier transform adapted to stratified media. A byproduct of our result is the absence of trapped modes at the junction of open waveguides as soon as the angles between branches are greater than $\pi$/2.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

20 extracted references · 17 canonical work pages

[3]

Bonnet-Ben Dhia, S

A.-S. Bonnet-Ben Dhia, S. Fliss, C. Hazard, and A. Tonnoir , A Rellich type theorem for the Helmholtz equation in a conical domain , Comptes Rendus Mathematique, 354 (2016), pp. 27–32, https://doi.org/10.1016/j.crma.2015.10.015

work page doi:10.1016/j.crma.2015.10.015 2016
[12]

J. Ott , Halfspace Matching: a Domain Decomposition Method for Scattering by 2D Open Waveguides, PhD thesis, Karlsruher Institut f¨ ur Technologie (KIT), 2017, https://doi.org/ 10.5445/IR/1000070898

work page arXiv 2017
[1]

Y. M. Berezansky, Z. G. Sheftel, and G. F. Us , Functional Analysis, II, vol. 86 of Opera- tor Theory: Advances and Applications, Birkh¨ auser Basel, 1996, https://doi.org/10.1007/ 978-3-0348-9024-3

work page 1996
[2]

Bonnet-Ben Dhia, G

A.-S. Bonnet-Ben Dhia, G. Dakhia, C. Hazard, and L. Chorfi , Diffraction by a defect in an open waveguide: a mathematical analysis based on a modal radiation condition , SIAM Journal on Applied Mathematics, 70 (2009), pp. 677–693, https://doi.org/10.1137/ 080740155

work page 2009
[4]

Bonnet-Ben Dhia, B

A.-S. Bonnet-Ben Dhia, B. Goursaud, and C. Hazard , Mathematical analysis of the junc- tion of two acoustic open waveguides , SIAM Journal on Applied Mathematics, 71 (2011), pp. 2048–2071, https://doi.org/10.1137/100811374

work page doi:10.1137/100811374 2011
[5]

DeBi `evre and D

S. DeBi `evre and D. W. Pravica , Spectral analysis for optical fibres and stratifled fluids II: Absence of eigenvalues , Communications in Partial Differential Equations, 17 (1992), pp. 69–97, https://doi.org/10.1080/03605309208820835

work page doi:10.1080/03605309208820835 1992
[6]

Dyatlov and M

S. Dyatlov and M. Zworski, Mathematical Theory of Scattering Resonances, American Math- ematical Society, 2019. 26 S. AL HUMAIKANI, A.-S. BONNET-BEN DHIA, S. FLISS, AND C. HAZARD

work page 2019
[7]

Fern´andez-Bertolin, D

A. Fern´andez-Bertolin, D. Stan, and L. Roncal, Landis’ conjecture: a survey, 2024, https: //arxiv.org/abs/2412.00788

arXiv 2024

Show all 20 references

[8]

Hazard, On the absence of trapped modes in locally perturbed open waveguides , IMA Jour- nal of Applied Mathematics, 80 (2015), pp

C. Hazard, On the absence of trapped modes in locally perturbed open waveguides , IMA Jour- nal of Applied Mathematics, 80 (2015), pp. 1049–1062, https://doi.org/10.1093/imamat/ hxu046

2015 doi
[9]

Kirsch, An Introduction to the Mathematical Theory of Inverse Problems , Springer New York, 2011, https://doi.org/10.1007/978-1-4419-8474-6

A. Kirsch, An Introduction to the Mathematical Theory of Inverse Problems , Springer New York, 2011, https://doi.org/10.1007/978-1-4419-8474-6

2011 doi
[10]

Krejˇciˇrik and V

D. Krejˇciˇrik and V. Lotoreichik , Quasi-conical domains with embedded eigenvalues , Bul- letin of the London Mathematical Society, 56 (2024), pp. 2969–2981, https://doi.org/10. 1112/blms.13113

2024
[11]

Linton and P

C. Linton and P. McIver , Embedded trapped modes in water waves and acoustics , Wave Motion, 45 (2007), pp. 16–29, https://doi.org/10.1016/j.wavemoti.2007.04.009

2007 doi
[13]

Pagneux, Trapped modes and edge resonances in acoustics and elasticity , in Dynamic Localization Phenomena in Elasticity, Acoustics and Electromagnetism, R

V. Pagneux, Trapped modes and edge resonances in acoustics and elasticity , in Dynamic Localization Phenomena in Elasticity, Acoustics and Electromagnetism, R. V. Craster and J. Kaplunov, eds., Springer Vienna, 2013, pp. 181–223, https://doi.org/10.1007/ 978-3-7091-1619-7 5

2013
[14]

Rellich, ¨Uber das asymptotische verhalten der l¨ osungen vonδu+λu = 0 in unendlichen ge- bieten, Jahresber

F. Rellich, ¨Uber das asymptotische verhalten der l¨ osungen vonδu+λu = 0 in unendlichen ge- bieten, Jahresber. Dtsch. Math.-Ver., 53 (1943), pp. 57–65, http://eudml.org/doc/146328

1943
[15]

Santosa and R

F. Santosa and R. Magnanini , Wave propagation in a 2-d optical waveguide , SIAM Journal on Applied Mathematics, 61 (2001), pp. 1237–1252, https://doi.org/10.1137/ S0036139999360489

2001
[16]

Schechter, Operator Methods in Quantum Mechanics, Elsevier North Holland, Inc., 1981

M. Schechter, Operator Methods in Quantum Mechanics, Elsevier North Holland, Inc., 1981

1981
[17]

Titchmarsh, Eigenfunction expansions associated with second order differential equations , Oxford Clarendon Press, 1946

E. Titchmarsh, Eigenfunction expansions associated with second order differential equations , Oxford Clarendon Press, 1946

1946
[18]

E. V. Vesalainen, Rellich type theorems for unbounded domains, Inverse Problems and Imag- ing, 8 (2014), pp. 865–883, https://doi.org/10.3934/ipi.2014.8.865

2014 doi
[19]

Weder, Absence of eigenvalues of the acoustic propagator in deformed wave guides , The Rocky Mountain Journal of Mathematics, (1988), pp

R. Weder, Absence of eigenvalues of the acoustic propagator in deformed wave guides , The Rocky Mountain Journal of Mathematics, (1988), pp. 495–503

1988
[20]

Weder, Spectral and Scattering Theory for Wave Propagation in Perturbed Stratified Media, vol

R. Weder, Spectral and Scattering Theory for Wave Propagation in Perturbed Stratified Media, vol. 87 of Applied Mathematical Sciences, Springer-Verlag New York, 1991

1991

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.

[1] [3]

Bonnet-Ben Dhia, S

A.-S. Bonnet-Ben Dhia, S. Fliss, C. Hazard, and A. Tonnoir , A Rellich type theorem for the Helmholtz equation in a conical domain , Comptes Rendus Mathematique, 354 (2016), pp. 27–32, https://doi.org/10.1016/j.crma.2015.10.015

work page doi:10.1016/j.crma.2015.10.015 2016

[2] [12]

J. Ott , Halfspace Matching: a Domain Decomposition Method for Scattering by 2D Open Waveguides, PhD thesis, Karlsruher Institut f¨ ur Technologie (KIT), 2017, https://doi.org/ 10.5445/IR/1000070898

work page arXiv 2017

[3] [1]

Y. M. Berezansky, Z. G. Sheftel, and G. F. Us , Functional Analysis, II, vol. 86 of Opera- tor Theory: Advances and Applications, Birkh¨ auser Basel, 1996, https://doi.org/10.1007/ 978-3-0348-9024-3

work page 1996

[4] [2]

Bonnet-Ben Dhia, G

A.-S. Bonnet-Ben Dhia, G. Dakhia, C. Hazard, and L. Chorfi , Diffraction by a defect in an open waveguide: a mathematical analysis based on a modal radiation condition , SIAM Journal on Applied Mathematics, 70 (2009), pp. 677–693, https://doi.org/10.1137/ 080740155

work page 2009

[5] [4]

Bonnet-Ben Dhia, B

A.-S. Bonnet-Ben Dhia, B. Goursaud, and C. Hazard , Mathematical analysis of the junc- tion of two acoustic open waveguides , SIAM Journal on Applied Mathematics, 71 (2011), pp. 2048–2071, https://doi.org/10.1137/100811374

work page doi:10.1137/100811374 2011

[6] [5]

DeBi `evre and D

S. DeBi `evre and D. W. Pravica , Spectral analysis for optical fibres and stratifled fluids II: Absence of eigenvalues , Communications in Partial Differential Equations, 17 (1992), pp. 69–97, https://doi.org/10.1080/03605309208820835

work page doi:10.1080/03605309208820835 1992

[7] [6]

Dyatlov and M

S. Dyatlov and M. Zworski, Mathematical Theory of Scattering Resonances, American Math- ematical Society, 2019. 26 S. AL HUMAIKANI, A.-S. BONNET-BEN DHIA, S. FLISS, AND C. HAZARD

work page 2019

[8] [7]

Fern´andez-Bertolin, D

A. Fern´andez-Bertolin, D. Stan, and L. Roncal, Landis’ conjecture: a survey, 2024, https: //arxiv.org/abs/2412.00788

arXiv 2024