How smooth is the smoothest function in a given re nable space ?
نویسندگان
چکیده
A closed subspace V of L 2 := L 2 (IR d) is called PSI (principal shift-invariant) if it is the smallest space that contains all the shifts (i.e., integer translates) of some function 2 L 2. Ideally, each function f in such PSI V can be written uniquely as a convergent series f = X 2ZZ d c()(?) with kck`2 kfk L 2. In this case one says that the shifts of form a Riesz basis or that they are L 2-stable; this is, in particular, the case when these shifts form an orthonormal set. We are interested here in PSI spaces which are reenable in the sense that, for some integer N > 1, the space is a subspace of V. The role of reenable PSI spaces in the construction of wavelets from multiresolution analysis, as well as in the study of subdivision algorithms is well-known, well-understood and well-documented (cf. e.g., D2] and CDM]). The two properties of a reenable PSI space that we compare here are: (s) the smoothness of the \smoothest" non-zero function g 2 V. (ao) the approximation orders provided by V. This latter notion refers to the decay of the error when approximating smooth functions from dilations of V ; roughly speaking, V provides approximation order k if dist(f; V j) = O(N ?jk) for every suuciently smooth function f. Here, V j := V (N j). One of the early discoveries in this area was the non-trivial observation that for a reenable PSI space V , (s) and (ao) are connected. For example, a result in M] shows that if decays rapidly and its shifts are L 2-stable, then V provides approximation order k as soon as lies in the Sobolev space W k?1 2. A closely related result appears in CDM]. More recently, the following is proved in R]: Result 1. Let V be an N-reenable PSI space. Then the following conditions are equivalent: (a) V provides approximation order k.
منابع مشابه
How Smooth Is the Smoothest Function in a Given Refinable Space
is a subspace of V. The role of refinable PSI spaces in the construction of wavelets from multiresolution analysis, as well as in the study of subdivision algorithms is wellknown, well-understood, and well-documented (cf. e.g., [6, 4]). The two properties of a refinable PSI space that we compare here are: (s) the smoothness of the "smoothest" nonzero function g ∈ V. (ao) the approximation order...
متن کاملStability and Independence of the Shifts of Finitely Many Refinable Functions
Typical constructions of wavelets depend on the stability of the shifts of an underlying re nable function. Unfortunately, several desirable properties are not available with compactly supported orthogonal wavelets, e.g. symmetry and piecewise polynomial structure. Presently, multiwavelets seem to o er a satisfactory alternative. The study of multiwavelets involves the consideration of the prop...
متن کاملHow Many Holes Can Locally Linearly Independent Reenable Function Vectors Have?
In this paper we consider the support properties of locally linearly independent re nable function vectors r T We propose an algorithm for computing the global support of the components of Further for T we investigate the supports especially the possibility of holes of re nable function vectors if local linear independence is assumed Finally we give some necessary conditions for local linear in...
متن کاملApproximation Properties of Multi Scaling Functions A Fourier Approach
In this paper we consider approximation properties of a nite set of functions r which are not necessarily compactly supported but have a suitable decay rate Assuming that the function vector r is re n able we sketch a new way how to derive necessary and su cient conditions for the re nement mask in Fourier domain Introduction For applications of multi wavelets in nite element methods the proble...
متن کاملMangasarian-Fromovitz and Zangwill Conditions For Non-Smooth Infinite Optimization problems in Banach Spaces
In this paper we study optimization problems with infinite many inequality constraints on a Banach space where the objective function and the binding constraints are Lipschitz near the optimal solution. Necessary optimality conditions and constraint qualifications in terms of Michel-Penot subdifferential are given.
متن کامل