Diophantine approximation, Bessel functions and radially symmetric periodic solutions of semilinear wave equations in a ball

The aim of this paper is to consider the radially-symmetric periodic-Dirichlet problem on for the equation

where is the classical Laplacian operator, and denotes the open ball of center and radius in When is a sufficiently large irrational with bounded partial quotients, we combine some number theory techniques with the asymptotic properties of the Bessel functions to show that is not an accumulation point of the spectrum of the linear part. This result is used to obtain existence conditions for the nonlinear problem.

     

Gauss sums and Kloosterman sums over residue rings of algebraic integers

Let denote the ring of integers of an algebraic number field of degree which is totally and tamely ramified at the prime . Write , where . We evaluate the twisted Kloosterman sum

where and denote trace and norm, and where is a Dirichlet character (mod ). This extends results of Salié for and of Yangbo Ye for prime dividing Our method is based upon our evaluation of the Gauss sum

which extends results of Mauclaire for .

     

Geometry of chain complexes and outer automorphisms under derived equivalence

The two main theorems proved here are as follows: If is a finite dimensional algebra over an algebraically closed field, the identity component of the algebraic group of outer automorphisms of is invariant under derived equivalence. This invariance is obtained as a consequence of the following generalization of a result of Voigt. Namely, given an appropriate geometrization of the family of finite -module complexes with fixed sequence of dimensions and an ``almost projective' complex , there exists a canonical vector space embedding

where is the pertinent product of general linear groups acting on , tangent spaces at are denoted by , and is identified with its image in the derived category .

     

Spherical classes and the Lambda algebra

Let be Singer's invariant-theoretic model of the dual of the lambda algebra with , where denotes the mod 2 Steenrod algebra. We prove that the inclusion of the Dickson algebra, , into is a chain-level representation of the Lannes-Zarati dual homomorphism

The Lannes-Zarati homomorphisms themselves, , correspond to an associated graded of the Hurewicz map

Based on this result, we discuss some algebraic versions of the classical conjecture on spherical classes, which states that Only Hopf invariant one and Kervaire invariant one classes are detected by the Hurewicz homomorphism. One of these algebraic conjectures predicts that every Dickson element, i.e. element in , of positive degree represents the homology class in for 2$">.

We also show that factors through , where denotes the differential of . Therefore, the problem of determining should be of interest.

     

Non-trivial quadratic approximations to zero of a family of cubic Pisot numbers

This paper gives exact rates of quadratic approximations to an infinite class of cubic Pisot numbers. We show that for any cubic Pisot number , with minimal polynomial , such that , and where has only one real root, then there exists a , explicitly given here, such that:

(1)

For all 0$">, all but finitely many integer quadratics satisfy

where is the height function.

(2)

For all 0$"> there exists a sequence of integer quadratics such that

Furthermore, for all in this class of cubic Pisot numbers. What is surprising about this result is how precise it is, giving exact upper and lower bounds for these approximations.

     

Limits of interpolatory processes

Given distinct real numbers and a positive approximation of the identity , which converges weakly to the Dirac delta measure as goes to zero, we investigate the polynomials which solve the interpolation problem

with prescribed data . More specifically, we are interested in the behavior of when the data is of the form for some prescribed function . One of our results asserts that if is sufficiently nice and has sufficiently well-behaved moments, then converges to a limit which can be completely characterized. As an application we identify the limits of certain fundamental interpolatory splines whose knot set is , where is an arbitrary finite subset of the integer lattice , as their degree goes to infinity.

     

Whitney's extension problem for multivariate -functions

We prove that the trace of the space to an arbitrary closed subset is characterized by the following ``finiteness' property. A function belongs to the trace space if and only if the restriction to an arbitrary subset consisting of at most can be extended to a function such that

The constant is sharp.

The proof is based on a Lipschitz selection result which is interesting in its own right.

     

Random variable dilation equation and multidimensional prescale functions

A random variable satisfying the random variable dilation equation , where is a discrete random variable independent of with values in a lattice and weights and is an expanding and -preserving matrix, if absolutely continuous with respect to Lebesgue measure, will have a density which will satisfy a dilation equation

We have obtained necessary and sufficient conditions for the existence of the density and a simple sufficient condition for 's existence in terms of the weights Wavelets in can be generated in several ways. One is through a multiresolution analysis of generated by a compactly supported prescale function . The prescale function will satisfy a dilation equation and its lattice translates will form a Riesz basis for the closed linear span of the translates. The sufficient condition for the existence of allows a tractable method for designing candidates for multidimensional prescale functions, which includes the case of multidimensional splines. We also show that this sufficient condition is necessary in the case when is a prescale function.

     

Newton polygons and local integrability of negative powers of smooth functions in the plane

Let be any smooth real-valued function with . For a sufficiently small neighborhood of the origin, we study the number

It is known that sometimes this number can be expressed in a natural way using the Newton polygon of . We provide necessary and sufficient conditions for this Newton polygon characterization to hold. The behavior of the integral at the supremal is also analyzed.

     

Projective metrics and mixing properties on towers

We study the decay of correlations for towers. Using Birkhoff's projective metrics, we obtain a rate of mixing of the form:

where goes to zero in a way related to the asymptotic mass of upper floors, is some Lipschitz norm and is some norm. The fact that the dependence on is given by an norm is useful to study asymptotic laws of successive entrance times.

     

Representation formulae and inequalities for solutions of a class of second order partial differential equations

Let be a possibly degenerate second order differential operator and let be its fundamental solution at ; here is a suitable distance. In this paper we study necessary and sufficient conditions for the weak solutions of on to satisfy the representation formula

We prove that (R) holds provided is superlinear, without any assumption on the behavior of at infinity. On the other hand, if satisfies the condition

then (R) holds with no growth assumptions on .

     

Uniqueness of solution to a free boundary problem from combustion

We investigate the uniqueness and agreement between different kinds of solutions for a free boundary problem in heat propagation that in classical terms is formulated as follows: to find a continuous function defined in a domain and such that

0\}. \end{displaymath}">

We also assume that the interior boundary of the positivity set, \nobreak 0\}$">, so-called free boundary, is a regular hypersurface on which the following conditions are satisfied:

Here denotes outward unit spatial normal to the free boundary. In addition, initial data are specified, as well as either Dirichlet or Neumann data on the parabolic boundary of . This problem arises in combustion theory as a limit situation in the propagation of premixed flames (high activation energy limit).

The problem admits classical solutions only for good data and for small times. Several generalized concepts of solution have been proposed, among them the concepts of limit solution and viscosity solution. We investigate conditions under which the three concepts agree and produce a unique solution.

     

Sums of unit fractions

Erdos and Straus conjectured that for any positive integer the equation has a solution in positive integers , and . Let k \geq 3$"> and

We show that parametric solutions can be used to find upper bounds on where the number of parameters increases exponentially with . This enables us to prove

0. \end{displaymath}">

This improves upon earlier work by Viola (1973) and Shen (1986), and is an ``exponential generalization' of the work of Vaughan (1970), who considered the case .

     

-estimates for the linearized Monge-Ampère equation

Let be a strictly convex domain and let be a convex function such that    det in . The linearized Monge-Ampère equation is

where det is the matrix of cofactors of . We prove that there exist and depending only on , and such that

for all solutions to the equation .

     

A note on Meyers' Theorem in

Lower semicontinuity properties of multiple integrals

are studied when may grow linearly with respect to the highest-order derivative, and admissible sequences converge strongly in It is shown that under certain continuity assumptions on convexity, -quasiconvexity or -polyconvexity of

ensures lower semicontinuity. The case where is -quasiconvex remains open except in some very particular cases, such as when

     

Geometry of phase space and solutions of semilinear elliptic equations in a ball

We consider the problem

where denotes the unit ball in , , and . Merle and Peletier showed that for there is a unique value such that a radial singular solution exists. This value is the only one at which an unbounded sequence of classical solutions of (1) may accumulate. Here we prove that if additionally

then for close to , a large number of classical solutions of (1) exist. In particular infinitely many solutions are present if . We establish a similar assertion for the problem

where , , and satisfies the same condition as above.

     

On the representation of integers as linear combinations of consecutive values of a polynomial

Let be a cyclotomic field with ring of integers and let be a polynomial whose values on belong to . If the ideal of generated by the values of on is itself, then every algebraic integer of may be written in the following form:

for some integer , where the 's are roots of unity of . Moreover, there are two effective constants and such that the least integer (for a fixed ) is less than , where

     

Solutions of fully nonlinear elliptic equations with patches of zero gradient: Existence, regularity and convexity of level curves

In this paper, we first construct ``viscosity' solutions (in the Crandall-Lions sense) of fully nonlinear elliptic equations of the form

In fact, viscosity solutions are surprisingly weak. Since candidates for solutions are just continuous, we only require that the ``test' polynomials (those tangent from above or below to the graph of at a point ) satisfy the correct inequality only if . That is, we simply disregard those test polynomials for which .

Nevertheless, this is enough, by an appropriate use of the Alexandroff-Bakelman technique, to prove existence, regularity and, in two dimensions, for , (0$">) and constant boundary conditions on a convex domain, to prove that there is only one convex patch.

     

The steepest point of the boundary layers of singularly perturbed semilinear elliptic problems

We consider the nonlinear singularly perturbed problem

0 \quad \mbox{in} \enskip \Omega, u = 0 \quad \mbox{on} \enskip \partial\Omega, \end{displaymath}">

where () is an appropriately smooth bounded domain and 0$"> is a small parameter. It is known that under some conditions on , the solution corresponding to develops boundary layers when . We determine the steepest point of the boundary layers on the boundary by establishing an asymptotic formula for the slope of the boundary layers with exact second term.

     

Strongly indefinite functionals and multiple solutions of elliptic systems

We study existence and multiplicity of solutions of the elliptic system

where , is a smooth bounded domain and . We assume that the nonlinear term

where , , and . So some supercritical systems are included. Nontrivial solutions are obtained. When is even in , we show that the system possesses a sequence of solutions associated with a sequence of positive energies (resp. negative energies) going toward infinity (resp. zero) if 2$"> (resp. ). All results are proved using variational methods. Some new critical point theorems for strongly indefinite functionals are proved.