SearcharxivSearch

arXiv subjects

Mark M. Malamud

Publications and source records attributed to Mark M. Malamud.

At least 19 recordsLinked to original sources

Real-valued spectral shift functions for contractions and dissipative operators

In recent joint papers the authors of this note solved a famous problem remained open for many years and proved that for arbitrary contractions with trace class difference there exists an integrable spectral shift function, for which an analogue of the Lifshits--Krein trace formula holds. Similar results were also obtained for pairs of dissipative operators. Note that in contrast with the case of self-adjoint and unitary operators it may happen that there is no {\it real-valued} integrable spectral shift function. In this note we announce results that give sufficient conditions for the existence of an integrable real-valued spectral shift function in the case of pairs of contractions. We also consider the case of pairs of dissipative operators.

math.FA

On the trace formulas and completeness property of root vectors systems for $2 \times 2$ Dirac type operators

The paper is concerned with the completeness property of the system of root vectors of a boundary value problem for the following $2 \times 2$ Dirac type equation $$ L y = -i B^{-1} y' + Q(x) y = λy , \quad y= {\rm col}(y_1, y_2), \quad x \in [0,1], $$ $$ B = {\rm diag}(b_1, b_2), \quad b_1 < 0 < b_2, \quad\text{and}\quad Q \in W_1^n[0,1] \otimes \mathbb{C}^{2 \times 2}, $$ subject to general non-regular two-point boundary conditions $C y(0) + D y(1) = 0$. If $b_2 = -b_1 = 1$ this equation is equivalent to the one dimensional Dirac equation. We establish asymptotic expansion of the characteristic determinant of this boundary value problem. This expansion directly yields new completeness result for the system of root vectors of such boundary value problem with non-regular and even degenerate boundary conditions. We also present several explicit completeness results in terms of the values $Q^{(j)}(0)$ and $Q^{(j)}(1)$. In the case of degenerate boundary conditions and analytic $Q(\cdot)$, the criterion of completeness property is established.

math.SP

On transformation operators and Riesz basis property of root vectors system for $n \times n$ Dirac type operators. Application to the Timoshenko beam model

The paper is concerned with the following $n\times n$ Dirac type equation$$Ly=-iB(x)^{-1}(y'+Q(x)y)=λy, \quad B(x)=B(x)^*,\quad y={\rm col}(y_1,\ldots,y_n),\quad x\in[0,\ell],$$ on a finite interval $[0,\ell]$. Here $Q$ is a summable potential $n\times n$ matrix function and $B$ is an invertible self-adjoint diagonal bounded matrix function. If $n=2m$ and $B(x)={\rm diag}(-I_m,I_m)$, this equation is equivalent to Dirac equation of order $n$. We show the existence of triangular transformation operators for such equation under additional uniform separation conditions on the entries of the matrix function $B$. Here we apply this result to study direct spectral properties of the boundary value problem (BVP) associated with the above equation subject to the general boundary conditions $U(y)=Cy(0)+Dy(\ell)=0,{\rm rank}(C\ D)=n$. We apply this result to show that the deviation of the characteristic determinants of this BVP and the unperturbed BVP (with $Q=0$) is a Fourier transform of some summable function, which in turn yields asymptotic behavior of the spectrum in the case of regular boundary conditions. Namely, $λ_m=λ_m^0+o(1)$ as $m\to\infty$, where $\{λ_m\}_{m\in\mathbb{Z}}$ and $\{λ_m^0\}_{m\in\mathbb{Z}}$ are sequences of eigenvalues of perturbed and unperturbed ($Q=0$) BVP, respectively. Further, we prove that the system of root vectors of the above BVP constitutes a Riesz basis in a certain weighted $L^2$-space, provided that the boundary conditions are strictly regular. The main results are applied to establish asymptotic behavior of eigenvalues and eigenvectors, and the Riesz basis property for the dynamic generator of the Timoshenko beam model. We also found a new case when eigenvalues have an explicit asymptotic, which to the best of our knowledge is new even in the case of constant parameters of the model.

math.SP

Stability of spectral characteristics and Bari basis property of boundary value problems for $2 \times 2$ Dirac type systems

The paper is concerned with the stability property under perturbation $Q\to\widetilde Q$ of different spectral characteristics of a BVP associated in $L^2([0,1];\Bbb C^2)$ with the following $2\times2$ Dirac type equation $$L_U(Q)y=-iB^{-1}y'+Q(x)y=λy,\quad B={\rm diag}(b_1,b_2),\quad b_1<0<b_2,\quad y={\rm col}(y_1,y_2),\quad(1)$$ with a potential matrix $Q\in L^p=L^p([0,1];\Bbb C^{2\times2})$ and subject to regular boundary conditions $Uy=\{U_1,U_2\}y=0$. Our approach to spectral stability relies on the existence of triangular transformation operators $K_Q^\pm$ for system (1) with $Q\in L^1$ established in our previous works. We prove the Lipshitz property of the mapping $Q\to K_Q^\pm$ from the balls in $L^p$ to the special Banach spaces $X_{\infty,p}^2,X_{1,p}^2$, naturally arising here, and obtain similar property for Fourier transforms of $K_Q^\pm$. These properties are of independent interest and play a crucial role in the proofs of all stability results discussed in the paper. For instance, as an immediate consequence we get the Lipshitz property of the mapping $Q\toΦ_Q$, where $Φ_Q$ is the fundamental matrix of the system (1). Assuming boundary conditions (BC) to be strictly regular, we show that the mapping $Q\toσ(L_U(Q))-σ(L_U(0))$ sends $L^p,p\in[1,2]$, either into $l^{p'}$ or into $l^p(\{(1+|n|)^{p-2}\})$; we also establish its Lipshitz property on compacts. We show similar result for the mapping $Q\to F_Q-F_0$ into $l^{p'}(\Bbb Z; C([0,1];\Bbb C^2))$, where $F_Q$ is a sequence of normalized eigenfunctions of $L_U(Q)$. Certain modifications of these results are proved for balls in $L^p,p\in[1,2]$. If $Q\in L^2$ we establish a criterion for the system of root vectors of $L_U(Q)$ to form a Bari basis in $L^2([0,1];\Bbb C^2)$. Under a simple additional assumption this system forms a Bari basis if and only if BC are self-adjoint.

math.SP

Completeness property of one-dimensional perturbations of normal and spectral operators generated by first order systems

The paper is concerned with completeness property of rank one perturbations of unperturbed operators generated by special boundary value problems (BVP) for the following $2 \times 2$ system \begin{equation} L y = -i B^{-1} y' + Q(x) y = λy , \quad B = \begin{pmatrix} b_1 & 0 \\ 0 & b_2 \end{pmatrix}, \quad y = \begin{pmatrix} y_1 \\ y_2 \end{pmatrix}, \end{equation} on a finite interval assuming that a potential matrix $Q$ is summable, and $b_1 b_2^{-1} \notin \mathbb{R}$ (essentially non-Dirac type case). We assume that unperturbed operator generated by a BVP belongs to one of the following three subclasses of the class of spectral operators: (a) normal operators; (b) operators similar either to a normal or almost normal; (c) operators that meet Riesz basis property with parentheses. We show that in each of the three cases there exists (in general, non-unique) operator generated by a quasi-periodic BVP and its certain rank-one perturbations (in the resolvent sense) generated by special BVPs which are complete while their adjoint are not. In connection with the case (b) we investigate Riesz basis property of quasi-periodic BVP under certain assumptions on a potential matrix $Q$. We also find a simple formula for the rank of the resolvent difference for operators corresponding to two BVPs for $n \times n$ system in terms of the coefficients of boundary linear forms.

math.SP

Scattering matrices and Dirichlet-to-Neumann maps

A general representation formula for the scattering matrix of a scattering system consisting of two self-adjoint operators in terms of an abstract operator valued Titchmarsh-Weyl $m$-function is proved. This result is applied to scattering problems for different self-adjoint realizations of Schrödinger operators on unbounded domains, Schrödinger operators with singular potentials supported on hypersurfaces, and orthogonal couplings of Schrödinger operators. In these applications the scattering matrix is expressed in an explicit form with the help of Dirichlet-to-Neumann maps.

math-ph

On the Riesz basis property of root vectors system for $2 \times 2$ Dirac type operators

The paper is concerned with the Riesz basis property of a boundary value problem associated in $L^2[0,1] \otimes \mathbb{C}^2$ with the following $2 \times 2$ Dirac type equation $$ L y = -i B^{-1} y' + Q(x) y = λy, \quad B = \begin{pmatrix} b_1 & 0 \\ 0 & b_2 \end{pmatrix}, \quad y = \begin{pmatrix} y_1 \\ y_2 \end{pmatrix}, \quad (1) $$ with a summable potential matrix $Q \in L^1[0,1] \otimes \mathbb{C}^{2 \times 2}$ and $b_1 < 0 < b_2$. If $b_2 = -b_1 =1$ this equation is equivalent to one dimensional Dirac equation. It is proved that the system of root functions of a linear boundary value problem constitutes a Riesz basis in $L^2[0,1] \otimes \mathbb{C}^2$ provided that the boundary conditions are strictly regular. By analogy with the case of ordinary differential equations, boundary conditions are called strictly regular if the eigenvalues of the corresponding unperturbed $(Q=0)$ operator are asymptotically simple and separated. As distinguished from the Dirac case there is no simple algebraic criterion of the strict regularity whenever $b_1 + b_2 \not = 0$. However under certain restrictions on coefficients of the boundary linear forms we present certain algebraic criteria of the strict regularity in the latter case. In particular, it is shown that regular separated boundary conditions are always strictly regular while periodic (antiperiodic) boundary conditions are strictly regular if and only if $b_1 + b_2 \not = 0.$ The proof of the main result is based on existence of triangular transformation operators for system (1). Their existence is also established here in the case of a summable $Q$. In the case of regular (but not strictly regular) boundary conditions we prove the Riesz basis property with parentheses. The main results are applied to establish the Riesz basis property of the dynamic generator of spatially non-homogenous damped Timoshenko beam model.

math.SP

On the completeness and Riesz basis property of root subspaces of boundary value problems for first order systems and applications

The paper is concerned with the completeness property of root functions of general boundary value problems for $n \times n$ first order systems of ordinary differential equations on a finite interval. In comparison with the recent paper [45] we substantially relax the assumptions on boundary conditions guarantying the completeness of root vectors, allowing them to be non-weakly regular and even degenerate. Emphasize that in this case the completeness property substantially depends on the values of a potential matrix at the endpoints of the interval. It is also shown that the system of root vectors of the general $n \times n$ Dirac type system subject to certain boundary conditions forms a Riesz basis with parentheses. We also show that arbitrary complete dissipative boundary value problem for Dirac type operator with a summable potential matrix admits the spectral synthesis in $L^2([0,1]; \mathbb{C}^n)$. Finally, we apply our results to investigate completeness and the Riesz basis property of the dynamic generator of spatially non-homogenous damped Timoshenko beam model.

math.SP

Perturbation determinants and trace formulas for singular perturbations

We use the boundary triplet approach to extend the classical concept of perturbation determinants to a more general setup. In particular, we examine the concept of perturbation determinants to pairs of proper extensions of closed symmetric operators. For an ordered pair of extensions we express the perturbation determinant in terms of the abstract Weyl function and the corresponding boundary operators. A crucial role in our approach plays so-called almost solvable extensions. We obtain trace formulas for pairs of self-adjoint, dissipative and other pairs of extensions and express the spectral shift function in terms of the abstract Weyl function and the characteristic function of almost solvable extensions. We emphasize that for pairs of dissipative extensions our results are new even for the case of additive perturbations. In this case we improve and complete some classical results of M.G. Krein for pairs of self-adjoint and dissipative operators. We apply the main results to ordinary differential operators and to elliptic operators as well.

math-ph

Spectral theory of Schrödinger operators with infinitely many point interactions and radial positive definite functions

A number of results on radial positive definite functions on ${\mathbb R^n}$ related to Schoenberg's integral representation theorem are obtained. They are applied to the study of spectral properties of self-adjoint realizations of two- and three-dimensional Schrödinger operators with countably many point interactions. In particular, we find conditions on the configuration of point interactions such that any self-adjoint realization has purely absolutely continuous non-negative spectrum. We also apply some results on Schrödinger operators to obtain new results on completely monotone functions.

math.FA

On the unitary equivalence of absolutely continuous parts of self-adjoint extensions

The classical Weyl-von Neumann theorem states that for any self-adjoint operator $A$ in a separable Hilbert space $\mathfrak H$ there exists a (non-unique) Hilbert-Schmidt operator $C = C^*$ such that the perturbed operator $A+C$ has purely point spectrum. We are interesting whether this result remains valid for non-additive perturbations by considering self-adjoint extensions of a given densely defined symmetric operator $A$ in $\mathfrak H$ and fixing an extension $A_0 = A_0^*$. We show that for a wide class of symmetric operators the absolutely continuous parts of extensions $\widetilde A = {\widetilde A}^*$ and $A_0$ are unitarily equivalent provided that their resolvent difference is a compact operator. Namely, we show that this is true whenever the Weyl function $M(\cdot)$ of a pair $\{A,A_0\}$ admits bounded limits $M(t) := \wlim_{y\to+0}M(t+iy)$ for a.e. $t \in \mathbb{R}$. This result is applied to direct sums of symmetric operators and Sturm-Liouville operators with operator potentials.

math-ph

Finite Rank Perturbations, Scattering Matrices and Inverse Problems

In this paper the scattering matrix of a scattering system consisting of two selfadjoint operators with finite dimensional resolvent difference is expressed in terms of a matrix Nevanlinna function. The problem is embedded into an extension theoretic framework and the theory of boundary triplets and associated Weyl functions for (in general nondensely defined) symmetric operators is applied. The representation results are extended to dissipative scattering systems and an explicit solution of an inverse scattering problem for the Lax-Phillips scattering matrix is presented.

math-ph

Spectral Theory of Elliptic Operators in Exterior Domains

We consider various closed (and self-adjoint) extensions of elliptic differential expressions of the type $\cA=\sum_{0\le |α|,|β|\le m}(-1)^αD^αa_{α, β}(x)D^β$, $a_{α, β}(\cdot)\in C^{\infty}({\overlineΩ})$, on smooth (bounded or unbounded) domains in $\bbR^n$ with compact boundary. Using the concept of boundary triples and operator-valued Weyl-Titchmarsh functions, we prove various trace ideal properties of powers of resolvent differences of these closed realizations of $\cA$ and derive estimates on eigenvalues of certain self-adjoint realizations in spectral gaps of the Dirichlet realization. Our results extend classical theorems due to Visik, Povzner, Birman, and Grubb.

math.SP

The similarity problem for $J$-nonnegative Sturm-Liouville operators

Sufficient conditions for the similarity of the operator $A := 1/r(x) (-d^2/dx^2 +q(x))$ with an indefinite weight $r(x)=(\sgn x)|r(x)|$ are obtained. These conditions are formulated in terms of Titchmarsh-Weyl $m$-coefficients. Sufficient conditions for the regularity of the critical points 0 and $\infty$ of $J$-nonnegative Sturm-Liouville operators are also obtained. This result is exploited to prove the regularity of 0 for various classes of Sturm-Liouville operators. This implies the similarity of the considered operators to self-adjoint ones. In particular, in the case $r(x)=\sgn x$ and $q\in L^1(R, (1+|x|)dx)$, we prove that $A$ is similar to a self-adjoint operator if and only if $A$ is $J$-nonnegative. The latter condition on $q$ is sharp, i.e., we construct $q\in \cap_{γ<1} L^1(R, (1+|x|)^γdx)$ such that $A$ is $J$-nonnegative with the singular critical point 0. Hence $A$ is not similar to a self-adjoint operator. For periodic and infinite-zone potentials, we show that $J$-positivity is sufficient for the similarity of $A$ to a self-adjoint operator. In the case $q\equiv 0$, we prove the regularity of the critical point 0 for a wide class of weights $r$. This yields new results for "forward-backward" diffusion equations.

math.SP

Trace formulae for dissipative and coupled scattering systems

For scattering systems consisting of a (family of) maximal dissipative extension(s) and a selfadjoint extension of a symmetric operator with finite deficiency indices, the spectral shift function is expressed in terms of an abstract Titchmarsh-Weyl function and a variant of the Birman-Krein formula is proved.

math-ph

Scattering matrices and Weyl functions

For a scattering system $\{A_Θ,A_0\}$ consisting of selfadjoint extensions $A_Θ$ and $A_0$ of a symmetric operator $A$ with finite deficiency indices, the scattering matrix $\{S_\gT(\gl)\}$ and a spectral shift function $ξ_Θ$ are calculated in terms of the Weyl function associated with the boundary triplet for $A^*$ and a simple proof of the Krein-Birman formula is given. The results are applied to singular Sturm-Liouville operators with scalar and matrix potentials, to Dirac operators and to Schrödinger operators with point interactions.

math-ph

On the spectral Theory of Operator Measures

In the first section we provide a solution to the M. G. Krein problem about an inner description of the space $L_2(Σ,H).$ In the second section we introduce the multiplicity function for an operator measure. Making use of the description of the space $L_2(Σ,H)$ we establish the correctness of the definition and give a criterion for a spectral measure to be a dilation of a given operator measure. In the third section we prove that the set of principal vectors of an operator measure is an everywhere dense $G_δ.$ This implies, in particular, that there are a lot of principal vectors in any cyclic subspace of a selfadjoint operator. In the 4th section we introduce Hellinger spectral types for an arbitrary operator measure and prove the existence of subspaces, realizing them. In the 5-th section we give an answer to an old question of Paulsen and provide an analytic description of completely bounded operator measures. The results of this section belong to the second author.

math.SP

On the number of square integrable solutions and self-adjointness of symmetric first order systems of differential equations

The main purpose of this paper is to investigate the formal deficiency indices ${\cal N}_{\pm}(I)$ of a symmetric first order system $$ Jf'+Bf=λ{\cal H} f $$ on an interval $I$, where $I=\mathbb{R}$ or $I=\mathbb{R}_\pm.$ Here $J,B,{cal H}$ are $n\times n$ matrix valued functions and the Hamiltonian ${\cal H}\ge 0$ may be singular even everywhere. We obtain two results for such a system to have minimal numbers ${\cal N}_\pm(\mathbb{R})=0$ (resp. ${\cal N}_\pm(\mathbb{R}_\pm)=n$) and a criterion for their maximality ${\cal N}_{\pm}(\mathbb{R}_+)=2n.$ Some conditions for a canonical system to have intermediate numbers ${\cal N}_\pm(\mathbb{R}_+)$ are presented, too. We also obtain a generalization of the well-known Titchmarsh-Sears theorem for second order Sturm-Liouville type equations. This contains results due to Lidskii and Krein as special cases. It is important to note that in general the above system does not give rise to an operator but rather to a symmetric linear relation in a Hilbert space. These relations are investigated in detail. As a byproduct we obtain very short proofs of (generalizations of) the main results of a paper by Kogan and Rofe-Beketov (Proc. Roy. Soc. Edinb. 74 (1974/75))as well as a criterion for the quasi-regularity of canonical systems. This covers the Kac-Krein theorem and some results from the quoted paper of Kogan and Rofe-Beketov.

math.FA