The ground state eigenvalue of Hill's equation with white noise potential

The distribution of the ground state eigenvalue λ₀(Q) of Hill's operator Q = −d²/dx² + q(x) on the circle of perimeter 1 is expressed in two different ways in case the potential q is standard white noise. Let WN be the associated white noise measure, and let CBM be the measure for circular Brownian motion p(x), 0 ≤ x < 1, formed from the standard Brownian motion b(x), 0 ≤ x ≤ 1, starting at b(0) = a, by conditioning so that b(1) = a, and distributing this common level over the line according to the measure da. The connection is based upon the Ricatti correspondence q(x) = λ + p′ (x) + p²(x). The two versions of the distribution are (1) in which $\overline{p}$ is the mean value ∫ pdx, and (2) the left‐hand side of (2) being the density for (1) and CBM₀ the conditional circular Brownian measure on $\overline{p}$ = 0. (1) and (2) are related by the divergence theorem in function space as suggested by the recognition of the Jacobian factor the outward‐pointing normal component of the vector field v(x) = ∂Δ(λ)/∂q(x), 0 ≤ x < 1, Δ being the Hill's discriminant for Q. The Ricatti correspondence prompts the idea that (1) and (2) are instances of the Cameron‐Martin formula, but it is not so: The latter has to do with the initial value problem for Ricatti, but it is the periodic problem that figures here, so the proof must be done by hand, by finite‐dimensional approximation. The adaptation of 1 and 2 to potentials of Ornstein‐Uhlenbeck type is reported without details. © 1999 John Wiley & Sons, Inc.     

On the band gap structure of Hill's equation

We revisit the old problem of finding the stability and instability intervals of a second-order elliptic equation on the real line with periodic coefficients (Hill's equation). It is well known that the stability intervals correspond to the spectrum of the Bloch family of operators defined on a single period. Here we propose a characterization of the instability intervals. We introduce a new family of non-self-adjoint operators, formally equivalent to the Bloch ones but with an imaginary Bloch parameter, that we call exponential. We prove that they admit a countable infinite number of eigenvalues which, when they are real, completely characterize the intervals of instability of Hill's equation.     

The spectrum of Hill's equation

Letq be an infinitely differentiable function of period 1. Then the spectrum of Hill's operatorQ=?d ²/dx ²+q(x) in the class of functions of period 2 is a discrete series - ∞<λ₀<λ₁≦λ₂<λ₃≦λ₄<...<λ_2i?1≦λ_2i ↑∞. Let the numer of simple eigenvalues be 2n+1<=∞. Borg [1] proved thatn=0 if and only ifq is constant. Hochstadt [21] proved thatn=1 if and only ifq=c+2p with a constantc and a Weierstrassian elliptic functionp. Lax [29] notes thatn=m if¹ q=4k ² K ² m(m+1)sn ²(2Kx,k). The present paper studies the casen<∞, continuing investigations of Borg [1], Buslaev and Faddeev [2], Dikii [3, 4], Flaschka [10], Gardneret al. [12], Gelfand [13], Gelfand and Levitan [14], Hochstadt [21], and Lax [28–30] in various directions. The content may be summed up in the statement thatq is an abelian function; in fact, from the present standpoint, the whole subject appears as a part of the classical function theory of the hyperelliptic irrationality \(\ell (\lambda ) = \sqrt { - (\lambda - \lambda _0 )(\lambda - \lambda _1 )...(\lambda - \lambda _{2n} )} .\) The casen=∞ requires the development of the theory of abelian and theta functions for infinite genus; this will be reported upon in another place. Some of the results have been obtained independently by Novikov [34], Dubrovin and Novikov [6] and A. R. Its and V. B. Matveev [22].     

On a random scaled porous media equation

It is shown that a random scaled porous media equation arising from a stochastic porous media equation with linear multiplicative noise through a random transformation is well-posed in L^∞. In the fast diffusion case we show existence in Lp.     

On the eigenvalue estimation for solution to Lyapunov equation

This paper deals with the problems of eigenvalue estimation for the solution to the perturbed matrix Lyapunov equation. We obtain some eigenvalue inequalities on condition that X is a positive semidefinite solution to the equation A^TXA − X = −Q, which can be used in control theory and linear system stability.     

On ground state of fractional $p$-Kirchhoff equation involving subcritical and critical exponential growth

In this paper, we concern the existence of nontrivial ground state solutions of fractional $p$-Kirchhoff equation $$\left\{\begin{array}{ll} m\left(\|u\|^p\right) [(-\Delta)_p^su+V(x)|u|^{p-2}u] =f(x,u) \quad\text{in}\, \mathbb{R}^N, \vspace{0.2 cm}\\ \|u\|=\left(\int_{\mathbb{R}^{2N}}\frac{|u(x)-u(y)|^p}{|x-y|^{N+ps}}dxdy +\int_{\mathbb{R}^N}V(x)|u|^pdx\right)^{\frac{1}{p}}, \end{array}\right.$$ where $m:[0,+\infty)\rightarrow [0,+\infty)$ is a continuous function, $(-\Delta)_p^s$ is the fractional $p$-Laplacian operator with $0相似文献   

On the calculation of the largest eigenvalue of an integral equation

Two methods are given for obtaining computable bounds for the largest eigenvalue of a linear integral equation with a continuous, symmetric, and nonnegative kernel. Several numerical examples are presented.Sponsored by the Mathematics Research Center, United States Army, Madison, Wisconsin, under Contract No.: DA-31-9 24-ARO-D-462.     

On the second eigenvalue and random walks in randomd-regular graphs

The main goal of this paper is to estimate the magnitude of the second largest eigenvalue in absolute value, λ₂, of (the adjacency matrix of) a randomd-regular graph,G. In order to do so, we study the probability that a random walk on a random graph returns to its originating vertex at thek-th step, for various values ofk. Our main theorem about eigenvalues is that $$E|\lambda _2 (G)|^m \leqslant \left( {2\sqrt {2d - 1} \left( {1 + \frac{{\log d}}{{\sqrt {2d} }} + 0\left( {\frac{1}{{\sqrt d }}} \right)} \right) + 0\left( {\frac{{d^{3/2} \log \log n}}{{\log n}}} \right)} \right)^m $$ for any \(m \leqslant 2\left\lfloor {log n\left\lfloor {\sqrt {2d - } 1/2} \right\rfloor /\log d} \right\rfloor \) , where E denotes the expected value over a certain probability space of 2d-regular graphs. It follows, for example, that for fixedd the second eigenvalue's magnitude is no more than \(2\sqrt {2d - 1} + 2\log d + C'\) with probability 1?n ^?C for constantsC andC′ for sufficiently largen.