The boundary control approach to inverse spectral theory

S.A. Avdonin; V.S. Mikhaylov

REVIEW 2 major objections 3 minor 35 references

The boundary control approach to inverse spectral theory

T0 review · 2 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 control operator unifies inverse spectral theory on the half-line.

desk verdict A valuable synthesis with a clean local GL equation, but the advertised a.e. spectral representation is false as stated for q∈L^1_loc. read the letter →

arxiv 2505.23329 v1 pith:JEUC4LTK submitted 2025-05-29 math.AP math.SP

S.A. Avdonin , V.S. Mikhaylov This is my paper

classification math.APmath.SP MSC 34B2034E0534L2534E4047B2081Q10

keywords inversespectraltheorySchrödingeroperatorboundarycontrolmethodGelfand–LevitanequationsresponsefunctionmeasureTitchmarsh–Weylm-functionconnecting

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

For the Schrödinger operator $-\partial_x^2+q(x)$ on the half-line with Dirichlet condition at $x=0$, this paper establishes that one object, the boundary control operator $C_T$, carries the entire inverse spectral problem. The operator is built only from the response function $r(t)$ on $[0,2T]$, the boundary data of the associated wave equation. The paper proves the almost-everywhere spectral formula $r(t)=\int \frac{\sin(\sqrt{\lambda}t)}{\sqrt{\lambda}} d\sigma(\lambda)$ with $\sigma$ the regularized spectral measure, and shows that a local Gelfand–Levitan equation for the kernel of the inverse control operator recovers the potential by $q(y)=2\frac{d}{dy}V(y,y)$. If these claims are correct, the Gelfand–Levitan theory, Krein's method, Simon's $A$-amplitude theory, Remling's de Branges-space approach, and the boundary control method are equivalent formulations of one underlying structure, with boundary control providing the local and physically motivated route.

What carries the argument

The central object is the connecting operator $C_T$ on $L^2(0,T)$, defined by $\langle C_T f,g\rangle=\langle u^f(\cdot,T),u^g(\cdot,T)\rangle$; it equals $(W_T)^*W_T$, is positive definite and boundedly invertible, and its kernel is $c_T(t,s)=p(2T-t-s)-p(t-s)$ with $p(t)=\frac12\int_0^{|t|}r(s)\,ds$, so it is known from the response function alone. The argument is carried by the kernels $w(x,s)$ and $V(y,t)$ of the forward and inverse control operators, by the spectral measure $\rho$ and its regularization $\sigma(\lambda)=\rho(\lambda)-\frac{2}{3\pi}\lambda^{3/2}$, and by sine-transform identities that convert the spectral representation of $C_T$ into the local Gelfand–Levitan equation (2.64). The machinery turns inverse spectral data into a linear Fredholm equation whose kernel is local because of the finite speed of wave propagation.

What would settle it

Pick a confining locally integrable potential such as $q(x)=x^2$ on the half-line, compute its response function $r(t)$ from the wave equation (2.1), and compute the right-hand side of (2.42) from its spectral measure; if the integral diverges or disagrees with $r(t)$ on a set of positive measure, Theorem 2 fails as stated. Alternatively, test numerically whether the kernel integral in (2.35) converges uniformly for a potential whose spectral measure has eigenvalues growing faster than $\lambda^{1/2}$; a counterexample would falsify the claimed $q\in L^1_{\mathrm{loc}}$ generality.

Watch

Extended reading notes

Core claim

The central discovery is that the dynamical inverse data of the boundary control method determine and are determined by the spectral data of the half-line Schrödinger operator through explicit integral identities. The connecting operator $C_T=(W_T)^*W_T$, where $W_T$ maps boundary controls to the wave field at time $T$, is expressed directly in terms of the response function $r$ by $c_T(t,s)=p(2T-t-s)-p(t-s)$ with $p(t)=\frac12\int_0^{|t|}r(s)\,ds$; Theorem 1 represents its kernel as $\int \sin\sqrt{\lambda}(T-t)\sin\sqrt{\lambda}(T-s)/\lambda\,d\sigma(\lambda)$. From this the authors derive the almost-everywhere spectral representation of $r$ (Theorem 2), and then a local Gelfand–Levitan equation (Theorem 3) whose solution $V$ yields $q(y)=2\frac{d}{dy}V(y,y)$. The same apparatus reproduces, by simple changes of variables, Krein's integral equation and Remling's two equations, and identifies the positivity of $C_T$ with the existence of a locally integrable potential having the given response data.

Load-bearing premise

The load-bearing assumption is that the spectral measure has the same high-energy cumulative growth as the free Schrödinger operator, namely $\frac{2}{3\pi}\lambda^{3/2}$, so that subtracting that term leaves a remainder for which the oscillatory integrals converge; this is built into the definition of the regularized measure $\sigma$, but it is not automatic for arbitrary locally integrable potentials, and the differentiation step producing the formula for $r(t)$ is not fully justified in the paper.

Editorial extensions

If this is right

The classical Gelfand–Levitan equation becomes local: the potential on $[0,T]$ is recovered from response data on $[0,2T]$ by solving (2.64), without global spectral information.
The response function has a direct spectral meaning: $r(t)$ equals the regularized sine transform of the spectral measure for almost every $t$, so dynamical and spectral data are interchangeable.
The positivity of $C_T$ characterizes admissible inverse data: given $r\in L^1(0,2T)$, there exists a unique locally integrable $q$ on $[0,T]$ with that response exactly when $C_T$ constructed from $r$ is positive definite.
Because Krein's equation, Remling's equations, and the boundary control equations reduce to one another by changes of variables, a solver for any one of them applies to all the inverse spectral settings.
The recovery is linear: after constructing $C_T$ from the data, the potential is read off from derivatives of the solution of a linear Fredholm equation, not from nonlinear optimization.

Reading between the lines

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

The equivalence suggests that the numerical conditioning of inverse spectral reconstruction is governed by the Fredholm equation (2.64), so the positive definiteness of $C_T$ could serve as a data-consistency certificate before running any reconstruction.
The local character coming from finite propagation speed implies the method can be applied window-by-window, reconstructing $q$ on successive intervals from overlapping response segments, which would matter for long cables or layered media where only partial boundary data are available.
The same operator identity $(I+K)^*C_T(I+K)=I$ is likely to transfer to other boundary conditions and matrix-valued potentials, with an analogous positivity condition as the existence criterion.
Since the boundary control method is dimension-independent in principle, a higher-dimensional analogue of the local Gelfand–Levitan equation could give a constructive route from boundary response to an unknown coefficient without global spectral asymptotics; this is a testable extension, not a claim of the paper.

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.

Desk Editor's Note

read the letter

Dear colleague,

You should know two things about this paper. First, the advertised almost-everywhere spectral representation (Theorem 2) is not established for the stated class q∈L^1_loc and is in fact false as stated: the free spectral subtraction (2/(3π))λ^{3/2} is only the correct asymptotic for a restricted class of potentials, and for confining potentials like q(x)=x² the signed measure σ has a negative component of unbounded variation, making the integral ∫ sin(√λ t)/√λ dσ(λ) divergent. The differentiation step in (2.46)–(2.47) is not justified for such σ. The authors' own truncation remark (2.48) shows they know the fix, but it is not in the theorem statement. Second, the local Gelfand–Levitan equation (2.64) and the derivation in Section 2.5 are clean and stand on their own, independent of the spectral representation.

What the paper does well: it lays out the equivalence of the Krein, Remling, Simon and BC formulations, and it shows that the BC operator C_T provides a common architecture. The derivation of the operator identity (I+K)^* C_T (I+K)=I and the Volterra argument leading to the local GL equation is a genuine contribution. That part is worth taking seriously.

The soft spot is exactly Theorem 2 and Lemma 2. Lemma 2 claims uniform convergence for q∈L^1_loc, but the same counterexample q(x)=x² makes Φ_n(t,t) diverge as n→∞ because the free subtraction leaves a non-cancelling √λ density. So the proof of Theorem 1, which relies on Lemma 2, also needs the stricter hypothesis. This is not a cosmetic issue; it changes the admissible potential class from "locally integrable" to something like "potentials with free spectral asymptotics" (e.g., integral conditions on q).

My take: the paper's synthesis and the local GL equation deserve a serious referee, but Section 2.4 needs a major revision. If the authors restrict the potential class or systematically use the truncated spectral function, the a.e. representation may be salvageable. As written, the central new theorem is false.

Recommendation: send to peer review, but flag Theorem 2 as needing significant repair. I'd use this paper in a reading group to discuss why spectral asymptotics matter; I would not cite Theorem 2 in its current form, though Section 2.5 is likely citable once the hypotheses are fixed.

Best.

Referee Report

2 major / 3 minor

Summary. The paper develops the boundary control method for the one-dimensional Schrödinger operator −∂_x^2+q(x) on L^2(R_+) with Dirichlet boundary condition and q∈L^1_loc(R_+). It derives spectral representations for the kernel c_T of the connecting operator and for the response function r in terms of the regularized spectral measure σ(λ)=ρ(λ)−(2/(3π))λ^{3/2} (Theorems 1 and 2), and it derives a local Gelfand–Levitan equation (Theorem 3). The stated aim is to show that the Gelfand–Levitan, Krein, Simon, Remling, and boundary control approaches are equivalent reformulations of one underlying structure, with a strengthened almost-everywhere version of Remling's distributional formula (1.17).

Significance. If the theorems were valid, the paper would provide a genuinely useful unification of the major one-dimensional inverse spectral theories, together with a new local proof of the Gelfand–Levitan equation. The operator-identity derivation in Section 2.5 and the Goursat estimates in the Appendix are coherent and appear correct. However, the advertised strengthening of Remling's formula to pointwise almost-everywhere convergence is the central new claim, and it is not established for the stated class q∈L^1_loc; this is a load-bearing gap rather than a cosmetic issue.

major comments (2)

[§2.4, Eqs. (2.45)–(2.47)] The differentiation of (2.46) under the signed spectral measure is not justified for q∈L^1_loc. For such potentials the signed measure dσ=dρ−d((2/(3π))λ^{3/2}) need not have bounded variation, so the Lebesgue differentiation theorem cannot be applied and the integral in (2.42) may not exist. For instance, for q(x)=x^2, the spectral function grows like a multiple of λ, so dσ contains a negative component with density of order −√λ, making the integral in (2.42) behave like ∫ sin(√λ t)dλ, which diverges. Thus Theorem 2 as stated is false, not merely missing a technical hypothesis. The truncation suggested in (2.48) is not part of the theorem statement or proof.
[§2.4, Lemma 2 and Theorem 1] The same hypothesis problem affects the regularized spectral representation of c_T. Lemma 2 and Theorem 1 assert uniform convergence of the truncated integrals (2.32) and (2.35) for all q∈L^1_loc, but this relies on the free spectral asymptotics (2/(3π))λ^{3/2} being the correct subtraction. For generic locally integrable potentials, including confining potentials, the actual spectral function has different growth, so σ is not a signed measure of bounded variation and the asserted uniform convergence is not obtained. The proof should either state the additional conditions on q under which the subtraction is valid, or reformulate both theorems locally using a truncated spectral function as in the remark after (2.48).

minor comments (3)

[§2.4, after (2.48)] The sentence beginning "...in we can replace the formulas (2.35), (2.42) the regularized spectral function..." is grammatically incomplete and should be rewritten.
[Proof of Theorem 2] The proof ends with "the last formula proves the statement of the proposition"; this should say "theorem".
[Theorem 3, Eq. (2.64)] The recovery formula q(y)=2 d/dx V(y,y) should read q(y)=2 d/dy V(y,y); as written the derivative variable is ambiguous.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the core derivations are self-contained identities from the BC-method definitions, and the self-citations used are independent prior results.

full rationale

The paper's central derivations do not reduce to their own inputs. The operator C_T is defined dynamically via the inner product of waves, and Proposition 3 expresses its kernel through the response function; the operator identity (2.57), (I+K)^* C_T (I+K) = I, is a direct consequence of C_T = (W_T)^* W_T and (W_T)^{-1} = (I+K)J_T, not an assumed form of the Gelfand-Levitan equation. The local Gelfand-Levitan equation (2.64) is obtained by algebraic manipulation of this identity together with the representation of C_T, so it is derived rather than presupposed. Theorem 1's spectral representation of c_T follows from the spectral theorem, the identity (2.31), and the sin-transform identity (2.39); the proof does not use the target representation of r(t). Theorem 2 then differentiates (2.46) to obtain the a.e. representation of r(t); whether that differentiation is justified for the full class q in L^1_loc is a genuine mathematical correctness issue (the skeptic's example q(x)=x^2 suggests the stated theorem may fail as written, and the truncation remark near (2.48) is not incorporated into the theorem hypotheses), but an unjustified or false technical step is not circularity. The paper does cite the authors' earlier work, notably [7] for the relation A(t) = -2r(2t), and that relation is load-bearing for the claimed equivalence of positivity conditions; however, it is an independently published result with stated assumptions rather than an unverified premise whose conclusion is identical to the present claims. No fitted parameters are renamed as predictions, no known result is merely re-labelled, and no uniqueness theorem is imported from the authors' prior work to force the chosen approach. The derivation chain is self-contained once the potential class and the cited lemmas are accepted; the main unresolved issues are technical hypotheses, not circularity.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No parameters are fitted and no new entities are postulated. The central assumptions are analytic: the free spectral asymptotic encoded in σ and background theorems on spectral functions and Goursat problems. The paper's results stand or fall on these, not on any numerical or physical input.

assumptions (4)

domain assumption The spectral measure ρ of the half-line operator satisfies the free asymptotic normalization ρ(λ) = (2/(3π))λ^{3/2} + o(λ^{3/2}), so that σ(λ)=ρ(λ)-(2/(3π))λ^{3/2} makes the oscillatory integrals converge.
Invoked in (1.2), Lemma 2, and Theorems 1 and 2. Not stated as an explicit hypothesis for q∈L^1_loc; false for confining potentials, so the theorem statements need this or a truncation-based repair.
standard math Levitan's asymptotic theorem [30] on uniform convergence of the sequence Ψ_n in (2.33).
Lemma 2 relies on this cited theorem for the uniform convergence that underpins Theorem 1; the paper does not reprove it.
standard math Spectral theorem and transformation operator identities (2.27), (2.31) for q∈L^1_loc.
Used throughout Section 2.4 to pass from dynamical inner products to spectral integrals.
standard math Finite propagation speed and the Volterra invertibility of W_T (Proposition 2) hold for q∈L^1_loc.
This gives the local nature of the response function and the control operator construction.

how reviews work

0 comments

Cite this review

Pith. "Pith review of The boundary control approach to inverse spectral theory." pith.science (2026). https://pith.science/paper/JEUC4LTK

@misc{pith2026250523329,
  author       = {Pith},
  title        = {Pith review of: The boundary control approach to inverse spectral theory},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JEUC4LTK}},
  note         = {Machine review of arXiv:2505.23329}
}

read the original abstract

We establish connections between different approaches to inverse spectral problems: the classical Gelfand--Levitan theory, the Krein method, the Simon theory, the approach proposed by Remling and the Boundary Control method. We show that the Boundary Control approach provides simple and physically motivated proofs of the central results of other theories. We demonstrate also the connections between the dynamical and spectral data and derive the local version of the classical Gelfand--Levitan equations.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

35 extracted references · 35 canonical work pages

[1]

Avdonin S A and Belinskiy B P 2005 On the basis properties of the functions arising in the boundary control problem of a string with a variable tension,Discrete and Continuous Dynamical Systems (Supplement Volume)40–49

work page 2005
[2]

Avdonin S A and Belishev M I 1996 Boundary control and dynamical inverse problem for nonselfadjoint Sturm-Liouville operator (BC-method)Control Cybernet25 no 3 429–440

work page 1996
[3]

USSR Sbornik7 287–310

Avdonin S A, Belishev M I and Ivanov S A 1992 Matrix inverse problem for the equation utt −u xx +Q(x)u= 0Math. USSR Sbornik7 287–310

work page 1992
[4]

Avdonin S and Kurasov P 2008 Inverse problems for quantum treesInverse Problems and Imaging2 no 1 1–21

work page 2008
[5]

Avdonin S, Lenhart S and Protopopescu V 2002 Solving the dynamical inverse problem for the Schr¨ odinger equation by the boundary control methodInverse Problems18 no 2 349–361

work page 2002
[6]

Inverse Ill-Posed Probl13 no 3-6 317–330

Avdonin S, Lenhart S and Protopopescu V 2005 Determining the potential in the Schr¨ odinger equation from the Dirichlet to Neumann map by the boundary control methodJ. Inverse Ill-Posed Probl13 no 3-6 317–330

work page 2005
[7]

Avdonin S, Mikhaylov V and Rybkin A 2007 The boundary control approach to the Titchmarsh-Weylm−functionComm. Math. Phys.275 no 3 791–803

work page 2007
[8]

Letters22 no 11 1705–1709 19

Avdonin S and Pandolfi L 2009 Boundary control method and coefficient identification in the presence of boundary dissipationApplied Math. Letters22 no 11 1705–1709 19

work page 2009

Show all 35 references

[9]

Belishev M I 1987 An approach to multidimensional inverse problems for the wave equation Dokl. Akad. Nauk SSSR297 no 3 524–527 (in Russian) Belishev M I 1988Soviet Math. Dokl.36 no 3 481–484 (Engl. Transl.)

1987
[10]

Sb.180 no 5 584– 602, 720 (in Russian) Belishev M I 1990Math

Belishev M I 1989 Wave bases in multidimensional inverse problemsMat. Sb.180 no 5 584– 602, 720 (in Russian) Belishev M I 1990Math. USSR-Sb.67 no 1 23–42 (Engl. Transl.)

1989
[11]

Belishev M I 1997 Boundary control in reconstruction of manifolds and metrics (the BC method)Inverse Problems13 no 5 R1–R45

1997
[12]

Belishev M I 2007 Recent progress in the boundary control methodInverse Problems23 no 5 R1–R67

2007
[13]

Belishev M I and Blagoveshchenskii A S 1992 Multidimensional analogues of equations of Gelfand-Levitan-Krein type in an inverse problem for the wave equationConditionally well- posed problems in mathematical physics and analysis50–63,Ross. Akad. Nauk Sib. Otd., Inst. Mat., Nov...

1992
[14]

Belishev M I and Ivanov S A 1999 Characterization of data in the dynamic inverse problem for a two-velocity systemZap. Nauchn. Sem. S.-Peterburg. Otdel. Mat. Inst. Steklov. (POMI) 259,Kraev. Zadachi Mat. Fiz. i Smezh. Vopr. Teor. Funkts.30 19–45 296 (in Russian) Belishev M I a...

1999
[15]

Belishev M I and Katchalov A P 1992 Boundary control and quasiphotons in a problem of the reconstruction of a Riemannian manifold from dynamic dataZap. Nauchn. Sem. S.- Peterburg. Otdel. Mat. Inst. Steklov. (POMI)203,Mat. Voprosy Teor. Rasprostr. Voln.22 21–50 174 (in Russian)...

1992
[16]

Multidimensional inverse problemsComput

Belishev M I and Kurylev Ya V 1991 Boundary control, wave field continuation and inverse problems for the wave equation. Multidimensional inverse problemsComput. Math. Appl.22 no 4-5 27–52

1991
[17]

Blagoveschenskii A S 1971 On a local approach to the solution of the dynamical inverse problem for an inhomogeneous stringTrudy MIAN115 28–38 (in Russian)

1971
[18]

Nauk SSSR

Gel’fand I M and Levitan B M 1951 On the determination of a differential equation from its spectral functionIzvestiya Akad. Nauk SSSR. Ser. Mat.15 309–360 (in Russian) Gel’fand I M and Levitan B M 1955Amer. Math. Soc. Transl. (2)1 253–304

1951
[19]

Gesztesy F 2007 Inverse spectral theory as influenced by Barry Simon Spectral theory and mathematical physics: a Festschrift in honor of Barry Simon’s 60th birthdayProc. Sympos. Pure Math.76 Part 2 741–820Amer. Math. Soc. Providence RI

2007
[20]

General real potential and the connection to the spectral measureAnn

Gesztesy F and Simon B 2000 A new approach to inverse spectral theory, II. General real potential and the connection to the spectral measureAnn. of Math. (2)152 no 2 593–643

2000
[21]

Gopinath B and Sondhi M M 1970 Determination of the shape of the human vocal tract from acoustical measurementsBell Syst. Tech. J.July 1970 1195–1214

1970
[22]

Gopinath B and Sondhi M M 1971 Inversion of the Telegraph Equation and the Synthesis of Nonuniform LinesProceedings of the IEEE59 no 3 383–392

1971
[23]

Katchalov A, Kurylev Ya and Lassas M 2001 Inverse Boundary Spectral Problems (Chapman Hall/CRC: Boca Raton FL

2001
[24]

Krein M G 1953 A transmission function of a second order one-dimensional boundary value problemDokl. Akad. Nauk. SSSR88 no 3 405–408

1953
[25]

Krein M G 1954 On the one method of effective solving the inverse boundary value problem Dokl. Akad. Nauk. SSSR94 no 6 987–990

1954
[26]

Levitan B M, 1987 Inverse Sturm-Liouville Problems (Utrecht: The Netherlands)

1987
[27]

Nauk19 no 2 (116) 3–63

Levitan B M and Gasymov M G 1964 Determination of a differential equation by two of its spectraUspehi Mat. Nauk19 no 2 (116) 3–63

1964
[28]

Nauk SSSR Ser

Levitan B M 1952 On the asymptotic behavior of the spectral function of a self-adjoint differential equation of the second orderIzvestiya Akad. Nauk SSSR Ser. Mat.16 325–352

1952
[29]

Nauk SSSR

Levitan B M 1953 On the asymptotic behavior of the spectral function of a self-adjoint differential equation of the second order and on expansion in eigenfunctionsIzvestiya Akad. Nauk SSSR. Ser. Mat.17 331–364

1953
[30]

II.Izvestiya Akad

Levitan B M 1955 On the asymptotic behavior of a spectral function and on expansion in eigenfunctions of a self-adjoint differential equation of second order. II.Izvestiya Akad. Nauk SSSR. Ser. Mat.19 33–58 20 SERGEI A VDONIN AND VICTOR MIKHAYLOV

1955
[31]

Part II: Linear differential operators in Hilbert space (New York: Frederick Ungar Publishing Co.)

Naimark M A 1968 Linear differential operators. Part II: Linear differential operators in Hilbert space (New York: Frederick Ungar Publishing Co.)

1968
[32]

Z.245 no 3 597–617

Remling C 2003 Inverse spectral theory for one-dimensional Schr¨ odinger operators: theA functionMath. Z.245 no 3 597–617

2003
[33]

Remling C 2002 Schr¨ odinger operators and de Branges spacesJ. Funct. Anal.196 no 2 323–394

2002
[34]

Fundamental formalismAnnals of Mathematics150 1029–1057

Simon B 1999 A new approach to inverse spectral theory, I. Fundamental formalismAnnals of Mathematics150 1029–1057

1999
[35]

Tikhonov A N and Samarskii A A 1963 Equations of Mathematical Physics (New York: Pergamon Press) Department of Mathematics and Statistics, University of Alaska F airbanks, PO Box 756660, F airbanks, AK 99775 Email address:saavdonin@alaska.edu, Victor.Mikhaylov@iecn.u-nancy.fr

1963

Pith tools

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

[1] [1]

Avdonin S A and Belinskiy B P 2005 On the basis properties of the functions arising in the boundary control problem of a string with a variable tension,Discrete and Continuous Dynamical Systems (Supplement Volume)40–49

work page 2005

[2] [2]

Avdonin S A and Belishev M I 1996 Boundary control and dynamical inverse problem for nonselfadjoint Sturm-Liouville operator (BC-method)Control Cybernet25 no 3 429–440

work page 1996

[3] [3]

USSR Sbornik7 287–310

Avdonin S A, Belishev M I and Ivanov S A 1992 Matrix inverse problem for the equation utt −u xx +Q(x)u= 0Math. USSR Sbornik7 287–310

work page 1992

[4] [4]

Avdonin S and Kurasov P 2008 Inverse problems for quantum treesInverse Problems and Imaging2 no 1 1–21

work page 2008

[5] [5]

Avdonin S, Lenhart S and Protopopescu V 2002 Solving the dynamical inverse problem for the Schr¨ odinger equation by the boundary control methodInverse Problems18 no 2 349–361

work page 2002

[6] [6]

Inverse Ill-Posed Probl13 no 3-6 317–330

Avdonin S, Lenhart S and Protopopescu V 2005 Determining the potential in the Schr¨ odinger equation from the Dirichlet to Neumann map by the boundary control methodJ. Inverse Ill-Posed Probl13 no 3-6 317–330

work page 2005

[7] [7]

Avdonin S, Mikhaylov V and Rybkin A 2007 The boundary control approach to the Titchmarsh-Weylm−functionComm. Math. Phys.275 no 3 791–803

work page 2007

[8] [8]

Letters22 no 11 1705–1709 19

Avdonin S and Pandolfi L 2009 Boundary control method and coefficient identification in the presence of boundary dissipationApplied Math. Letters22 no 11 1705–1709 19

work page 2009