Abstract
Let \(X({\mathbb {R}})\) be a separable rearrangement-invariant space and w be a suitable Muckenhoupt weight. We show that for any semi-almost periodic Fourier multiplier a on \(X({\mathbb {R}},w)=\{f:fw\in X({\mathbb {R}})\}\) there exist uniquely determined almost periodic Fourier multipliers \(a_l,a_r\) on \(X({\mathbb {R}},w)\), such that
for some monotonically increasing function u with \(u(-\infty )=0\), \(u(+\infty )=1\) and some continuous and vanishing at infinity Fourier multiplier \(a_0\) on \(X({\mathbb {R}},w)\). This result extends previous results by Sarason (Duke Math J 44:357–364, 1977) for \(L^2({\mathbb {R}})\) and by Karlovich and Loreto Hernández (Integral Equ Oper Theor 62:85–128, 2008) for weighted Lebesgue spaces \(L^p({\mathbb {R}},w)\) with weights in a suitable subclass of the Muckenhoupt class \(A_p({\mathbb {R}})\).
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
Let \(C({\overline{{\mathbb {R}}}})\) be the \(C^*\)-algebra of all continuous functions on the two-point compactification of the real line \({\overline{{\mathbb {R}}}}=[-\infty ,+\infty ]\) and
where \({{\dot{{\mathbb {R}}}}}={\mathbb {R}}\cup \{\infty \}\) is the one-point compactification of the real line. Let APP denote the set of all almost periodic polynomials, that is, finite sums of the form \(\sum _{\lambda \in \varLambda }c_\lambda e_\lambda \), where
\(c_\lambda \in {\mathbb {C}}\) and \(\varLambda \subset {\mathbb {R}}\) is a finite subset of \({\mathbb {R}}\). The smallest closed subalgebra of \(L^\infty ({\mathbb {R}})\) that contains APP is denoted by AP and called the algebra of (uniformly) almost periodic functions. Sarason [36] introduced the algebra of semi-almost periodic functions as the smallest closed subalgebra of \(L^\infty ({\mathbb {R}})\) that contains AP and \(C({\overline{{\mathbb {R}}}})\):
It is not difficult to see that AP and SAP are \(C^*\)-subalgebras of \(L^\infty ({\mathbb {R}})\).
Theorem 1.1
(Sarason [36], see also [10, Theorem 1.21]) Let \(u\in C({\overline{{\mathbb {R}}}})\) be any function for which \(u(-\infty )=0\) and \(u(+\infty )=1\). If \(a\in SAP\), then there exist \(a_l,a_r\in AP\) and \(a_0\in C({{\dot{{\mathbb {R}}}}})\) such that \(a_0(\infty )=0\) and
The functions \(a_l,a_r\) are uniquely determined by a and independent of the particular choice of u. The maps \(a\mapsto a_l\) and \(a\mapsto a_r\) are \(C^*\)-algebra homomorphisms of SAP onto AP.
The uniquely determined function \(a_l\) (resp. \(a_r\)) is called the left (resp. right) almost periodic representative of the semi-almost periodic function a.
Let \(\mathcal{F}:L^2({\mathbb {R}})\rightarrow L^2({\mathbb {R}})\) denote the Fourier transform:
and let \(\mathcal{F}^{-1}:L^2({\mathbb {R}})\rightarrow L^2({\mathbb {R}})\) be the inverse of \(\mathcal{F}\),
It is well known that the Fourier convolution operator
is bounded on the space \(L^2({\mathbb {R}})\) for every \(a\in L^\infty ({\mathbb {R}})\).
Let \(X({\mathbb {R}})\) be a separable Banach function space (see Sect. 2.1 for the definition and some properties of Banach function spaces). Then \(L^2({\mathbb {R}})\cap X({\mathbb {R}})\) is dense in \(X({\mathbb {R}})\) (see, e.g., [15, Lemma 2.1]). A function \(a\in L^\infty ({\mathbb {R}})\) is called a Fourier multiplier on \(X({\mathbb {R}})\) if the convolution operator \(W^0(a)\) defined by (1.2) maps the set \(L^2({\mathbb {R}})\cap X({\mathbb {R}})\) into the space \(X({\mathbb {R}})\) and extends to a bounded linear operator on \(X({\mathbb {R}})\). The function a is called the symbol of the Fourier convolution operator \(W^0(a)\). The set \({\mathcal {M}}_{X({\mathbb {R}})}\) of all Fourier multipliers on \(X({\mathbb {R}})\) is a unital normed algebra under pointwise operations and the norm:
where \({\mathcal {B}}(X({\mathbb {R}}))\) denotes the Banach algebra of all bounded linear operators on the space \(X({\mathbb {R}})\).
Note that the Lebesgue spaces \(L^p({\mathbb {R}})\), \(1\le p\le \infty \), constitute the simplest example of Banach function spaces. Motivated by the work of Duduchava and Saginashvili [14], Karlovich and Spitkovsky [29] (see also [10, Section 19.1]) introduced the algebra \(SAP_{L^p({\mathbb {R}})}\) of semi-almost periodic Fourier multipliers on the Lebesgue spaces \(L^p({\mathbb {R}})\), \(1<p<\infty \), and proved an analogue of Sarason’s Theorem 1.1 for \(SAP_{L^p({\mathbb {R}})}\) (see [29, Lemma 3.1(iv)] and [10, Proposition 19.3]).
We should mention that, after Sarason’s pioneering paper [36], various classes of Toeplitz and convolution type operators involving semi-almost periodic functions were studied on various function spaces, for instance, by Saginashvili [35], Grudsky [19]; Böttcher et al. [3,4,5,6, 8,9,10]; Nolasco and Castro [32, 33]; Bogveradze and Castro [2]; the second author and Spitkovsky [25].
Let \({\mathfrak {M}}({\mathbb {R}})\) denote the set of all measurable complex-valued Lebesgue measurable functions on \({\mathbb {R}}\). As usual, we identify two functions on \({\mathbb {R}}\) which are equal almost everywhere. A measurable function \(w:{\mathbb {R}}\rightarrow [0,\infty ]\) is called a weight if the set \(w^{-1}(\{0,\infty \})\) has measure zero. For \(1<p<\infty \), the Muckenhoupt class \(A_p({\mathbb {R}})\) is defined as the class of all weights \(w:{\mathbb {R}}\rightarrow [0,\infty ]\) such that \(w\in L_{\mathrm{loc}}^p({\mathbb {R}})\), \(w^{-1}\in L_{\mathrm{loc}}^{p'}({\mathbb {R}})\) and
where \(1/p+1/p'=1\) and the supremum is taken over all intervals \(I\subset {\mathbb {R}}\) of finite length |I|. Since \(w\in L_{\mathrm{loc}}^p({\mathbb {R}})\) and \(w^{-1}\in L_{\mathrm{loc}}^{p'}({\mathbb {R}})\), the weighted Lebesgue space
is a separable Banach function space (see, e.g., [26, Lemma 2.4]) with the norm:
Note that if \(w\in A_p({\mathbb {R}})\), then it may happen that the function \(e_\lambda \) does not belong to \({\mathcal {M}}_{L^p({\mathbb {R}},w)}\) for some \(\lambda \in {\mathbb {R}}\). Hence,order to generalize Theorem 1.1 to the setting of weighted Lebesgue spaces \(L^p({\mathbb {R}},w)\), one has to restrict the study to a narrower class of weights. For \(1<p<\infty \), let
For a weight \(w\in A_p^0({\mathbb {R}})\), Karlovich and Loreto Hernández defined the algebra \(SAP_{L^p({\mathbb {R}},w)}\) of semi-almost periodic Fourier multipliers on the weighted Lebesgue space \(L^p({\mathbb {R}},w)\) and proved an analogue of Theorem 1.1 in this setting (see [27, Theorem 3.1]). The aim of this paper is to extend this result to the setting of separable rearrangement-invariant Banach function spaces with suitable Muckenhoupt weights.
It is well known that the Lebesgue spaces \(L^p({\mathbb {R}})\), \(1\le p\le \infty \), fall in the class of rearrangement-invariant Banach function spaces. Other classical examples of rearrangement-invariant Banach function spaces are Orlicz spaces \(L^\varPhi ({\mathbb {R}})\) and Lorentz spaces \(L^{p,q}({\mathbb {R}})\), \(1\le p,q\le \infty \). For a rearrangement-invariant Banach function space \(X({\mathbb {R}})\), its Boyd indices \(\alpha _X,\beta _X\) are important interpolation characteristics. In particular, \(\alpha _{L^p}=\beta _{L^p}=1/p\) for \(1\le p\le \infty \). In general, \(0\le \alpha _X\le \beta _X\le 1\) and it may happen that \(\alpha _X<\beta _X\). We postpone formal definitions of rearrangement-invariant Banach function spaces and their Boyd indices until Sects. 2.2–2.3 and refer to [1, Chap. 3] and [30, Chap. 2] for the detailed study of these concepts.
Let \(X({\mathbb {R}})\) be a separable rearrangement-invariant Banach function space with the Boyd indices \(\alpha _X,\beta _X\) satisfying \(0<\alpha _X,\beta _X<1\). Suppose that a weight w belongs to \(A_{1/\alpha _X}({\mathbb {R}})\cap A_{1/\beta _X}({\mathbb {R}})\). Then
is a separable Banach function space (see Lemma 2.3(b) below). Suppose that \(a:{\mathbb {R}}\rightarrow {\mathbb {C}}\) is a function of finite total variation V(a) given by
where the supremum is taken over all partitions of \({\mathbb {R}}\) of the form
with \(n\in {\mathbb {N}}\). The set \(V({\mathbb {R}})\) of all functions of finite total variation on \({\mathbb {R}}\) with the norm
is a unital non-separable Banach algebra. It follows from [21, Corollary 2.2] that there exists a constant \(c_{X({\mathbb {R}},w)}\in (0,\infty )\) such that for all \(a\in V({\mathbb {R}})\),
This inequality is usually called a Stechkin-type inequality (see, e.g., [13, Theorem 2.11] and [10, Theorem 17.1] for the case of Lebesgue spaces and Lebesgue spaces with Muckenhoupt weights, respectively). Let \(C_{X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\) and \(C_{X({\mathbb {R}},w)}({\overline{{\mathbb {R}}}})\) denote the closures of \(C({{\dot{{\mathbb {R}}}}})\cap V({\mathbb {R}})\) and \(C({\overline{{\mathbb {R}}}})\cap V({\mathbb {R}})\) with respect to the norm of \({\mathcal {M}}_{X({\mathbb {R}},w)}\), respectively.
If \(w\in A_{1/\alpha _X}^0({\mathbb {R}})\cap A_{1/\beta _X}^0({\mathbb {R}})\), then \(APP\subset {\mathcal {M}}_{X({\mathbb {R}},w)}\) (see Corollary 5.2 below). Because of this observation, we will refer to \(A_{1/\alpha _X}^0({\mathbb {R}})\cap A_{1/\beta _X}^0({\mathbb {R}})\) as the class of suitable Muckenhoupt weights. By \(AP_{X({\mathbb {R}},w)}\) we denote the closure of APP with respect to the norm of \({\mathcal {M}}_{X({\mathbb {R}},w)}\). Finally, let \(SAP_{X({\mathbb {R}},w)}\) be the smallest closed subalgbera of \({\mathcal {M}}_{X({\mathbb {R}},w)}\) that contains the algebras \(AP_{X({\mathbb {R}},w)}\) and \(C_{X({\mathbb {R}},w)}({\overline{{\mathbb {R}}}})\):
In this paper we present a self-contained proof of the following result.
Theorem 1.2
(Main result) Let \(X({\mathbb {R}})\) be a separable rearrangement-invariant Banach function space with the Boyd indices satisfying \(0<\alpha _X\), \(\beta _X<1\). Suppose that \(w\in A_{1/\alpha _X}^0({\mathbb {R}})\cap A_{1/\beta _X}^0({\mathbb {R}})\). Let \(u\in C({\overline{{\mathbb {R}}}})\) be any real-valued monotonically increasing function such that \(u(-\infty )=0\) and \(u(+\infty )=1\). Then for every function \(a\in SAP_{X({\mathbb {R}},w)}\) there exist functions \(a_l,a_r\in AP_{X({\mathbb {R}},w)}\) and a function \(a_0\in C_{X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\) such that \(a_0(\infty )=0\) and (1.1) holds. The functions \(a_l\), \(a_r\) are uniquely determined by the function a and are independent of the particular choice of the function u. The maps \(a\mapsto a_l\) and \(a\mapsto a_r\) are continuous Banach algebra homomorphisms of \(SAP_{X({\mathbb {R}},w)}\) onto \(AP_{X({\mathbb {R}},w)}\) of norm 1.
The paper is organized as follows. In Sect. 2, we collect definitions and properties of rearrangement-invariant Banach functions spaces and their Boyd indices \(\alpha _X,\beta _X\). Further, we discuss properties of weighted rearrangement-invariant spaces \(X({\mathbb {R}},w)\) and state several results about general Fourier multipliers on \(X({\mathbb {R}},w)\) for weights w belonging to the intersection of the Muckenhoupt classes \(A_{1/\alpha _X}({\mathbb {R}})\cap A_{1/\beta _X}({\mathbb {R}})\).
In Sect. 3, we show that, under the assumption \(w\in A_{1/\alpha _X}({\mathbb {R}})\cap A_{1/\beta _X}({\mathbb {R}})\), the set of continuous Fourier multipliers vanishing at infinity on the space \(X({\mathbb {R}},w)\) coincides with the closure of the set of all smooth compactly supported functions with respect to the norm of \({\mathcal {M}}_{X({\mathbb {R}},w)}\).
Relying on the results of the previous section, in Sect. 4, we show that \(C_{X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})=C_{X({\mathbb {R}},w)}({\overline{{\mathbb {R}}}})\cap C({{\dot{{\mathbb {R}}}}})\) and that the algebra \(C_{X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\) is contained in the algebra \(SO_{X({\mathbb {R}},w)}\) of slowly oscillating Fourier multipliers (see [21]).
In Sect. 5, we show that if \(w\in A_{1/\alpha _X}^0({\mathbb {R}})\cap A_{1/\beta _X}^0({\mathbb {R}})\), then the set of almost periodic polynomials APP is contained in \({\mathcal {M}}_{X({\mathbb {R}},w)}\). We give an example of a nontrivial weight in \(A_{1/\alpha _X}^0({\mathbb {R}})\cap A_{1/\beta _X}^0({\mathbb {R}})\) (based on an example from [27]). Further, we show that the product of an almost periodic Fourier multiplier and a continuous Fourier multiplier vanishing at infinity is a continuous Fourier multiplier vanishing at infinity.
Section 6 is devoted to the proof of the main result. We show that the set \({\mathcal {A}}_u\) of functions of the form (1.1) with \(a_l,a_r\in AP_{X({\mathbb {R}},w)}\) and \(a_0\in C_{X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\) such that \(a_0(\infty )=0\) forms an algebra, and that the mappings \(a\mapsto a_l\) and \(a\mapsto a_r\) are algebraic homomorphisms of \({\mathcal {A}}_u\) onto \(AP_{X({\mathbb {R}},w)}\). We prove that
which implies that the algebra \({\mathcal {A}}_u\) is closed. Since the closure of \({\mathcal {A}}_u\) with respect to the norm of \({\mathcal {M}}_{X({\mathbb {R}},w)}\) coincides with \(SAP_{X({\mathbb {R}},w)}\), we conclude that \({\mathcal {A}}_u\) is equal to \(SAP_{X({\mathbb {R}},w)}\). Moreover, inequalities (1.5) mean that \(a\mapsto a_l\) and \(a\mapsto a_r\) are Banach algebra homomorphisms of \(SAP_{X(|R,w)}\) onto \(AP_{X({\mathbb {R}},w)}\) of norm 1.
2 Preliminaries
2.1 Banach function spaces
Let \({\mathbb {R}}_+:=(0,\infty )\) and \({\mathbb {S}}\in \{{\mathbb {R}}_+,{\mathbb {R}}\}\). The set of all Lebesgue measurable complex-valued functions on \({\mathbb {S}}\) is denoted by \({\mathfrak {M}}({\mathbb {S}})\). Let \({\mathfrak {M}}^+({\mathbb {S}})\) be the subset of functions in \({\mathfrak {M}}({\mathbb {S}})\) whose values lie in \([0,\infty ]\). The Lebesgue measure of a measurable set \(E\subset {\mathbb {S}}\) is denoted by |E| and its characteristic function is denoted by \(\chi _E\). Following [1, Chap. 1, Definition 1.1], a mapping \(\rho :{\mathfrak {M}}^+({\mathbb {S}})\rightarrow [0,\infty ]\) is called a Banach function norm if, for all functions \(f,g, f_n \ (n\in {\mathbb {N}})\) in \({\mathfrak {M}}^+({\mathbb {S}})\), for all constants \(a\ge 0\), and for all measurable subsets E of \({\mathbb {S}}\), the following properties hold:
with \(C_E \in (0,\infty )\) which may depend on E and \(\rho \) but is independent of f. When functions differing only on a set of measure zero are identified, the set \(X({\mathbb {S}})\) of all functions \(f\in {\mathfrak {M}}({\mathbb {S}})\) for which \(\rho (|f|)<\infty \) is called a Banach function space. For each \(f\in X({\mathbb {S}})\), the norm of f is defined by
Under the natural linear space operations and under this norm, the set \(X({\mathbb {S}})\) becomes a Banach space (see [1, Chap. 1, Theorems 1.4 and 1.6]). If \(\rho \) is a Banach function norm, its associate norm \(\rho '\) is defined on \({\mathfrak {M}}^+({\mathbb {S}})\) by
It is a Banach function norm itself [1, Chap. 1, Theorem 2.2]. The Banach function space \(X'({\mathbb {R}})\) determined by the Banach function norm \(\rho '\) is called the associate space (Köthe dual) of \(X({\mathbb {S}})\). The associate space \(X'({\mathbb {S}})\) is naturally identified with a subspace of the (Banach) dual space \([X({\mathbb {S}})]^*\).
2.2 Rearrangement-invariant Banach function spaces
Suppose that \({\mathbb {S}}\in \{{\mathbb {R}},{\mathbb {R}}_+\}\). Let \({\mathfrak {M}}_0({\mathbb {S}})\) and \({\mathfrak {M}}_0^+({\mathbb {S}})\) be the classes of a.e. finite functions in \({\mathfrak {M}}({\mathbb {S}})\) and \({\mathfrak {M}}^+({\mathbb {S}})\), respectively. The distribution function \(\mu _f\) of a function \(f\in {\mathfrak {M}}_0({\mathbb {S}})\) is given by
Two functions \(f,g\in {\mathfrak {M}}_0({\mathbb {S}})\) are said to be equimeasurable if \(\mu _f(\lambda )=\mu _g(\lambda )\) for all \(\lambda \ge 0\). The non-increasing rearrangement of \(f\in {\mathfrak {M}}_0({\mathbb {S}})\) is the function defined by
We here use the standard convention that \(\inf \emptyset =+\infty \).
A Banach function norm \(\rho :{\mathfrak {M}}^+({\mathbb {S}}) \rightarrow [0,\infty ]\) is called rearrangement-invariant if for every pair of equimeasurable functions \(f,g \in {\mathfrak {M}}^+_0({\mathbb {S}})\) the equality \(\rho (f)=\rho (g)\) holds. In that case, the Banach function space \(X({\mathbb {S}})\) generated by \(\rho \) is said to be a rearrangement-invariant Banach function space (or simply rearrangement-invariant space). Lebesgue, Orlicz, and Lorentz spaces are classical examples of rearrangement-invariant Banach function spaces (see, e.g., [1] and the references therein). By [1, Chap. 2, Proposition 4.2], if a Banach function space \(X({\mathbb {S}})\) is rearrangement-invariant, then its associate space \(X'({\mathbb {S}})\) is rearrangement-invariant, too.
2.3 Boyd indices
Suppose \(X({\mathbb {R}})\) is a rearrangement-invariant Banach function space generated by a rearrangement-invariant Banach function norm \(\rho \). In this case, the Luxemburg representation theorem [1, Chap. 2, Theorem 4.10] provides a unique rearrangement-invariant Banach function norm \({\overline{\rho }}\) over the half-line \({\mathbb {R}}_+\) equipped with the Lebesgue measure, defined by
and such that \(\rho (f)={\overline{\rho }}(f^*)\) for all \(f\in {\mathfrak {M}}_0^+({\mathbb {R}})\). The rearrangement-invariant Banach function space generated by \({\overline{\rho }}\) is denoted by \({\overline{X}}({\mathbb {R}}_+)\).
For each \(t>0\), let \(E_t\) denote the dilation operator defined on \({\mathfrak {M}}({\mathbb {R}}_+)\) by
With \(X({\mathbb {R}})\) and \({\overline{X}}({\mathbb {R}}_+)\) as above, let \(h_X(t)\) denote the operator norm of \(E_{1/t}\) as an operator on \({\overline{X}}({\mathbb {R}}_+)\). By [1, Chap. 3, Proposition 5.11], for each \(t>0\), the operator \(E_t\) is bounded on \({\overline{X}}({\mathbb {R}}_+)\) and the function \(h_X\) is increasing and submultiplicative on \((0,\infty )\). The Boyd indices of \(X({\mathbb {R}})\) are the numbers \(\alpha _X\) and \(\beta _X\) defined by
By [1, Chap. 3, Proposition 5.13], \(0\le \alpha _X\le \beta _X\le 1\). The Boyd indices are said to be nontrivial if \(\alpha _X,\beta _X\in (0,1)\). The Boyd indices of the Lebesgue space \(L^p({\mathbb {R}})\), \(1\le p\le \infty \), are both equal to 1/p. Note that the Boyd indices of a rearrangement-invariant space may be different [1, Chap. 3, Exercises 6, 13].
The next theorem follows from the Boyd interpolation theorem [11, Theorem 1] for quasi-linear operators of weak types (p, p) and (q, q). Its proof can also be found in [1, Chap. 3, Theorem 5.16] and [30, Theorem 2.b.11].
Theorem 2.1
Let \(1\le q<p\le \infty \) and \(X({\mathbb {R}})\) be a rearrangement-invariant Banach function space with the Boyd indices \(\alpha _X,\beta _X\) satisfying \(1/p<\alpha _X\), \(\beta _X<1/q\). Then there exists a constant \(C_{p,q}\in (0,\infty )\) such that if a linear operator \(T:{\mathfrak {M}}({\mathbb {R}})\rightarrow {\mathfrak {M}}({\mathbb {R}})\) is bounded on the Lebesgue spaces \(L^p({\mathbb {R}})\) and \(L^q({\mathbb {R}})\), then it is also bounded on the rearrangement-invariant Banach function space \(X({\mathbb {R}})\) and
Notice that estimate (2.1) is not stated explicitly in [1, 11, 30]. However, it can be extracted from the proof of the Boyd interpolation theorem.
2.4 Weighted Banach function spaces
Let \(X({\mathbb {R}})\) be a Banach function space generated by a Banach function norm \(\rho \). We say that \(f\in X_{\mathrm{loc}}({\mathbb {R}})\) if \(f\chi _E\in X({\mathbb {R}})\) for any measurable set \(E\subset {\mathbb {R}}\) of finite measure.
Lemma 2.2
[26, Lemma 2.4] Let \(X({\mathbb {R}})\) be a Banach function space generated by a Banach function norm \(\rho \), let \(X'({\mathbb {R}})\) be its associate space, and let \(w:{\mathbb {R}}\rightarrow [0,\infty ]\) be a weight. Suppose that \(w\in X_{\mathrm{loc}}({\mathbb {R}})\) and \(1/w\in X_{\mathrm{loc}}'({\mathbb {R}})\). Then
is a Banach function norm and
is a Banach function space generated by the Banach function norm \(\rho _w\). The space \(X'({\mathbb {R}},w^{-1})\) is the associate space of \(X({\mathbb {R}},w)\).
2.5 Density of nice functions in separable rearrangement-invariant Banach function spaces with Muckenhoupt weights
Recall that the (noncentered) Hardy–Littlewood maximal function Mf of a function \(f\in L_{\mathrm{loc}}^1({\mathbb {R}})\) is defined by
where the supremum is taken over all intervals \(I\subset {\mathbb {R}}\) of finite length containing the point x.
Let \({\mathcal {S}}({\mathbb {R}})\) be the Schwartz space of rapidly decreasing smooth functions and let us denote by \({\mathcal {S}}_0({\mathbb {R}})\) the set of all functions \(f\in {\mathcal {S}}({\mathbb {R}})\) such that their Fourier transforms \({\mathcal {F}}f\) have compact supports.
Lemma 2.3
Let \(X({\mathbb {R}})\) be a separable rearrangement-invariant Banach function space and \(X'({\mathbb {R}})\) be its associate space. Suppose that the Boyd indices of \(X({\mathbb {R}})\) satisfy \(0<\alpha _X\), \(\beta _X<1\) and \(w\in A_{1/\alpha _X}({\mathbb {R}})\cap A_{1/\beta _X}({\mathbb {R}})\). Then
-
(a)
\(w\in X_{\mathrm{loc}}({\mathbb {R}})\) and \(1/w\in X_{\mathrm{loc}}'({\mathbb {R}})\);
-
(b)
the Banach function space space \(X({\mathbb {R}},w)\) is separable;
-
(c)
the Hardy-Littlewood maximal operator M is bounded on the Banach function space \(X({\mathbb {R}},w)\) and on its associate space \(X'({\mathbb {R}},w^{-1})\);
-
(d)
the set \({\mathcal {S}}_0({\mathbb {R}})\) is dense in the Banach function space \(X({\mathbb {R}},w)\).
Proof
Parts (a) and (c) are proved in [21, Section 4.3]. Part (b) follows from part (a), Lemma 2.2 and [26, Lemmas 2.7 and 2.11]. Part (d) is a consequence of parts (b), (c) and [16, Theorem 4]. \(\square \)
2.6 The Banach algebra \({\mathcal {M}}_{X({\mathbb {R}},w)}\) of Fourier multipliers
The following result plays an important role in this paper.
Theorem 2.4
Let \(X({\mathbb {R}})\) be a separable rearrangement-invariant Banach function space with the Boyd indices satisfying \(0<\alpha _X\), \(\beta _X<1\). Suppose that a weight w belongs to \(A_{1/\alpha _X}({\mathbb {R}})\cap A_{1/\beta _X}({\mathbb {R}})\). If \(a\in {\mathcal {M}}_{X({\mathbb {R}},w)}\), then
The constant 1 on the right-hand side of (2.2) is best possible.
This theorem follows from Lemma 2.3(b) and [15, Theorem 2.4] (which was deduced from [24, Corollary 4.2]).
Inequality (2.2) was established earlier in [22, Theorem 1] with some constant on the right-hand side that depends on the space \(X({\mathbb {R}},w)\).
Since (2.2) is available, an easy adaptation of the proof of [18, Proposition 2.5.13] leads to the following (we refer to the proof of [22, Corollary 1] for details).
Corollary 2.5
Let \(X({\mathbb {R}})\) be a separable rearrangement-invariant Banach function space with the Boyd indices satisfying \(0<\alpha _X\), \(\beta _X<1\). Suppose that a weight w belongs to \(A_{1/\alpha _X}({\mathbb {R}})\cap A_{1/\beta _X}({\mathbb {R}})\). Then the set of the Fourier multipliers \({\mathcal {M}}_{X({\mathbb {R}},w)}\) is a Banach algebra under pointwise operations and the norm \(\Vert \cdot \Vert _{{\mathcal {M}}_{X({\mathbb {R}},w)}}\).
As usual, we denote by \(C_c^\infty ({\mathbb {R}})\) the set of all infinitely differentiable functions with compact support.
Theorem 2.6
Suppose that a non-negative even function \(\varphi \in C_c^\infty ({\mathbb {R}})\) satisfies the condition
and the function \(\varphi _\delta \) is defined for \(\delta >0\) by
Let \(X({\mathbb {R}})\) be a separable rearrangement-invariant Banach function space with the Boyd indices satisfying \(0<\alpha _X\), \(\beta _X<1\). Suppose that a weight w belongs to \(A_{1/\alpha _X}({\mathbb {R}})\cap A_{1/\beta _X}({\mathbb {R}})\). If \(a\in {\mathcal {M}}_{X({\mathbb {R}},w)}\), then for every \(\delta >0\),
Proof
The proof is analogous to the proof of [23, Theorem 2.6]. It follows from Lemma 2.3(c) and [26, Theorems 3.8(a) and 3.9(c)] that if the weight w belongs to \(A_{1/\alpha _X}({\mathbb {R}})\cap A_{1/\beta _X}({\mathbb {R}})\), then
Therefore, by [24, Lemma 1.3], the Banach function space \(X({\mathbb {R}},w)\) satisfies the hypotheses of [24, Theorem 1.3]. It is shown in its proof (see [24, Section 4.2]) that for every \(\delta >0\) and every \(f\in {\mathcal {S}}({\mathbb {R}})\cap X({\mathbb {R}},w)\),
where
Then, for every \(\delta >0\),
By Lemma 2.3(b), the Banach function space \(X({\mathbb {R}},w)\) is separable. Then it follows from [1, Chap. 1, Corollary 5.6] and [24, Theorems 2.3 and 6.1] that for every \(\delta >0\), the left-hand side of inequality (2.6) coincides with the multiplier norm \(\Vert a*\varphi _\delta \Vert _{{\mathcal {M}}_{X({\mathbb {R}},w)}}\), which completes the proof of inequality (2.5). \(\square \)
3 Continuous Fourier multipliers vanishing at infinity
3.1 The case of Lebesgue spaces with Muckenhoupt weights
The closure of a subset \(\mathfrak{S}\) of a Banach space \(\mathcal{E}\) in the norm of \(\mathcal{E}\) will be denoted by \({\text {clos}}_{\mathcal {E}}({\mathfrak {S}})\).
Let \(C_0({\mathbb {R}})\) be the set of all functions \(f\in C({{\dot{{\mathbb {R}}}}})\) such that \(f(\infty )=0\).
Lemma 3.1
Let \(1<p<\infty \) and \(w\in A_p({\mathbb {R}})\). Then
Proof
The idea of the proof is borrowed from [20, Theorem 1.16] (see also [23, Theorem 3.1]). If \(w\in A_p({\mathbb {R}})\), then \(w^{1+\delta _2}\in A_{p(1+\delta _1)}({\mathbb {R}})\) whenever \(|\delta _1|\) and \(|\delta _2|\) are sufficiently small (see, e.g., [7, Theorem 2.31]). If \(p\ge 2\), then one can find sufficiently small \(\delta _1,\delta _2>0\) and a number \(\theta \in (0,1)\) such that
If \(1<p<2\), then one can find a sufficiently small number \(\delta _2>0\), a number \(\delta _1<0\) with sufficiently small \(|\delta _1|\), and a number \(\theta \in (0,1)\) such that all conditions in (3.1) are fulfilled.
Let us use the following abbreviations:
For \(n\in {\mathbb {N}}\), let
Then \(\psi _n\) has compact support and \(\Vert \psi _n\Vert _{V({\mathbb {R}})}=3\). By the Stechkin-type inequality (1.4),
where \(c_\theta \) is three times \(c_{L^{p(1+\delta _1)}({\mathbb {R}},w^{1+\delta _2})}\), and the latter constant is the constant from (1.4).
Let \(a\in \ C_0({\mathbb {R}})\cap V({\mathbb {R}})\). Fix \(\varepsilon >0\). For \(n\in {\mathbb {N}}\), take \(b_n:=a\psi _n\). Then
and \(b_n\in C_0({\mathbb {R}})\) has compact support. Taking into account the Stechkin-type inequality (1.4), we get
and
It follows from (3.1) and the Stein–Weiss interpolation theorem (see, e.g., [1, Chap. 3, Theorem 3.6]) that
Combining (3.3), (3.4) and (3.6), we see that there exists \(n_0\in {\mathbb {N}}\) such that
Let \(\varphi \in C_c^\infty ({\mathbb {R}})\) be a non-negative even function satisfying (2.3) and for \(\delta >0\) let the function \(\varphi _\delta \) be defined by (2.4). By Theorem 2.6 and inequality (3.5), for every \(\delta >0\),
It follows from [12, Propositions 4.18, 4.20–4.21] that \(b_{n_0}*\varphi _\delta \in C_c^\infty ({\mathbb {R}})\) and
In view of (3.1) and the Stein-Weiss interpolation theorem (see, e.g., [1, Chap. 3, Theorem 3.6]), we see that
Combining (3.8)–(3.10), we conclude that there exists \(\delta _0>0\) such that
Hence, it follows from (3.7) and (3.11) that for every function a in the intersection \(C_0({\mathbb {R}})\cap V({\mathbb {R}})\) and every \(\varepsilon >0\) there exists a function \(b_{n_0}*\varphi _{\delta _0}\in C_c^\infty ({\mathbb {R}})\) such that \(\Vert a-b_{n_0}*\varphi _{\delta _0}\Vert _{{\mathcal {M}}_p}<\varepsilon \). Therefore, \(a\in {\text {clos}}_{{\mathcal {M}}_p}\big (C_c^\infty ({\mathbb {R}})\big )\). \(\square \)
3.2 The case of rearrangement-invariant spaces with Muckenhoupt weights
The following lemma is an extension of the previous result to the case of rearrangement-invariant Banach function spaces with Muckenhoupt weights.
Lemma 3.2
Let \(X({\mathbb {R}})\) be a separable rearrangement-invariant Banach function space with the Boyd indices satisfying \(0<\alpha _X\), \(\beta _X<1\). Suppose that a weight w belongs to \(A_{1/\alpha _X}({\mathbb {R}})\cap A_{1/\beta _X}({\mathbb {R}})\). Then
Proof
Since \(\alpha _X,\beta _X\in (0,1)\) and \(w\in A_{1/\alpha _X}({\mathbb {R}})\cap A_{1/\beta _X}({\mathbb {R}})\), it follows from [7, Theorem 2.31] that there exist p and q such that
Let \(C_{p,q}\in (0,\infty )\) be the constant from estimate (2.1). Fix \(\varepsilon >0\) and take a function \(a\in C_0({\mathbb {R}})\cap V({\mathbb {R}})\). As in the proof of inequality (3.7) (see the proof of Lemma 3.1), it can be shown that there exists \(n_0\in {\mathbb {N}}\) such that
where \(b_n=a\psi _n\) and \(\psi _n\) is given by (3.2) for every \(n\in {\mathbb {N}}\). It follows from (3.12), (3.13) and Theorem 2.1 that
As in the proof of inequality (3.11) (see the proof of Lemma 3.1), it can be shown that there exists \(\delta _0>0\) such that
where \(\varphi \in C_c^\infty ({\mathbb {R}})\) is a non-negative even function satisfying (2.3) and the functions \(\varphi _\delta \) are defined for all \(\delta >0\) by (2.4). Arguing as in the proof of (3.14), we deduce from (3.12), (3.15) and Theorem 2.1 that
It follows from (3.14) and (3.16) that for every function a in the intersection \(C_0({\mathbb {R}})\cap V({\mathbb {R}})\) and every \(\varepsilon >0\) there exists a function \(b_{n_0}*\varphi _{\delta _0}\in C_c^\infty ({\mathbb {R}})\) such that \(\Vert a-b_{n_0}*\varphi _{\delta _0}\Vert _{{\mathcal {M}}_{X({\mathbb {R}},w)}}<\varepsilon \). Therefore, \(a\in {\text {clos}}_{{\mathcal {M}}_{X(R,w)}}\big (C_c^\infty ({\mathbb {R}})\big )\). \(\square \)
Now we are in a position to prove the main result of this section.
Theorem 3.3
Let \(X({\mathbb {R}})\) be a separable rearrangement-invariant Banach function space with the Boyd indices satisfying \(0<\alpha _X\), \(\beta _X<1\). Suppose that a weight w belongs to \(A_{1/\alpha _X}({\mathbb {R}})\cap A_{1/\beta _X}({\mathbb {R}})\). Consider the set
Then
Proof
Let \(a\in C_{X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\) be such that \(a(\infty )=0\). Fix \(\varepsilon >0\). By the definition of the algebra \(C_{X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\), there exists a function \(b\in C({{\dot{{\mathbb {R}}}}})\cap V({\mathbb {R}})\) such that
It follows from this observation and the continuous embedding of \({\mathcal {M}}_{X({\mathbb {R}},w)}\) into \(L^\infty ({\mathbb {R}})\) (see Theorem 2.4) that
Take \(c=b-b(\infty )\in C_0({\mathbb {R}})\cap V({\mathbb {R}})\). By Lemma 3.2, there exists a function \(d\in C_c^\infty ({\mathbb {R}})\subset {\mathcal {M}}_{X({\mathbb {R}},w)}\) such that
Combining inequalities (3.19)–(3.21), we see that
Hence
Let us prove the reverse embedding. Take \(a\in {\text {clos}}_{{\mathcal {M}}_{X({\mathbb {R}},w)}}\big (C_c^\infty ({\mathbb {R}})\big )\). Then there exists a sequence \(\{a_n\}_{n\in {\mathbb {N}}}\subset C_c^\infty ({\mathbb {R}})\) such that
Since \(C_c^\infty ({\mathbb {R}})\subset C({{\dot{{\mathbb {R}}}}})\cap V({\mathbb {R}})\), the above equality and the continuous embedding of the algebra \({\mathcal {M}}_{X({\mathbb {R}},w)}\) into the algebra \(L^\infty ({\mathbb {R}})\) (see Theorem 2.4) imply that \(a\in C_{X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\) and
Thus
Combining (3.22) and (3.23), we arrive at (3.18). \(\square \)
4 Continuous and slowly oscillating Fourier multipliers
4.1 Continuous Fourier multipliers on one and two-point compactifications of the real line
For a function \(f\in C({\overline{{\mathbb {R}}}})\), let
It is easy to see that
Therefore \(J_f\in C({\overline{{\mathbb {R}}}})\cap V({\mathbb {R}})\) and \(f-J_f\in C_0({\mathbb {R}})\).
The next lemma extends [29, Lemma 3.1(i)] from the setting of Lebesgue spaces to the setting of rearrangement-invariant Banach function spaces with Muckenhoupt weights.
Lemma 4.1
Let \(X({\mathbb {R}})\) be a separable rearrangement-invariant Banach function space with the Boyd indices satisfying \(0<\alpha _X\), \(\beta _X<1\). Suppose that \(w\in A_{1/\alpha _X}({\mathbb {R}})\cap A_{1/\beta _X}({\mathbb {R}})\). Then
Proof
The proof is analogous to the proof of [29, Lemma 3.1(i)] (see also [23, Lemma 3.2]). It is obvious that \(C_{X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\subset C_{X({\mathbb {R}},w)}({\overline{{\mathbb {R}}}})\). On the other hand, it follows from Theorem 2.4 that \(C_{X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\subset C({{\dot{{\mathbb {R}}}}})\). Therefore,
To prove the opposite embedding, let us consider an arbitrary function \(a\in C_{X({\mathbb {R}},w)}({\overline{{\mathbb {R}}}})\) such that \(a(+\infty )=a(-\infty )\). Let \(\{a_n\}_{n\in {\mathbb {N}}}\subset C({\overline{{\mathbb {R}}}})\cap V({\mathbb {R}})\) be a sequence such that \(\Vert a_n-a\Vert _{{\mathcal {M}}_{X({\mathbb {R}},w)}}\rightarrow 0\) as \(n\rightarrow \infty \). According to Theorem 2.4, the sequence \(\{a_n\}_{n\in {\mathbb {N}}}\) converges to a uniformly on \({\mathbb {R}}\). Hence, in particular, \(a_n(\pm \infty )\rightarrow a(\infty )\) as \(n\rightarrow \infty \). Let the functions \(b_n:=J_{a_n-a(\infty )}\) be defined by (4.1) with \(a_n-a(\infty )\) in place of f. By the Stechkin-type inequality (1.4) and equality (4.2), we have
Therefore, \(\Vert b_n\Vert _{{\mathcal {M}}_{X({\mathbb {R}},w)}}\rightarrow 0\) as \(n\rightarrow \infty \) and thus,
Since \(a_n-b_n\in C({{\dot{{\mathbb {R}}}}})\cap V({\mathbb {R}})\), the latter equality implies that \(a\in C_{X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\). Thus
Combining embeddings (4.4)–(4.5), we arrive at equality (4.3). \(\square \)
4.2 Embedding of the algebra \(C_{X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\) into the algebra \(SO_{X({\mathbb {R}},w)}\) of slowly oscillating Fourier multipliers
Let \(C_b({\mathbb {R}}):=C({\mathbb {R}})\cap L^\infty ({\mathbb {R}})\). For a bounded measurable function \(f:{\mathbb {R}}\rightarrow {\mathbb {C}}\) and a set \(J\subset {\mathbb {R}}\), let
Let SO be the \(C^*\)-algebra of all slowly oscillating functions at \(\infty \) defined by
Consider the differential operator \((Df)(x)=xf'(x)\) and its iterations defined by \(D^0f=f\) and \(D^jf=D(D^{j-1}f)\) for \(j\in {\mathbb {N}}\). Let
where \(C^3({\mathbb {R}})\) denotes the set of all three times continuously differentiable functions. It is easy to see that \(SO^3\) is a commutative Banach algebra under pointwise operations and the norm
It follows from [21, Corollary 2.6] that if \(X({\mathbb {R}})\) is a separable rearrangement-invariant Banach function space with the Boyd indices \(\alpha _X,\beta _X\) such that \(0<\alpha _X\), \(\beta _X<1\) and \(w\in A_{1/\alpha _X}({\mathbb {R}})\cap A_{1/\beta _X}({\mathbb {R}})\), then there exists a constant \(c_{X({\mathbb {R}},w)}\in (0,\infty )\) such that for all \(a\in SO^3\),
The continuous embedding \(SO^3\subset {\mathcal {M}}_{X({\mathbb {R}},w)}\) allows us to define the algebra \(SO_{X({\mathbb {R}},w)}\) of slowly oscillating Fourier multipliers as the closure of \(SO^3\) with respect to the multiplier norm:
The following result is analogous to [28, Lemma 3.6].
Lemma 4.2
Let \(X({\mathbb {R}})\) be a separable rearrangement-invariant Banach function space with the Boyd indices satisfying \(0<\alpha _X\), \(\beta _X<1\). Suppose that a weight w belongs to \(A_{1/\alpha _X}({\mathbb {R}})\cap A_{1/\beta _X}({\mathbb {R}})\). Then \(C_{X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\subset SO_{X({\mathbb {R}},w)}\).
Proof
Let \(a\in C_{X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\). Fix \(\varepsilon >0\). Then there exists \(b\in C({{\dot{{\mathbb {R}}}}})\cap V({\mathbb {R}})\) such that
Then \(b-b(\infty )\in C_0({\mathbb {R}})\cap V({\mathbb {R}})\). By Lemma 3.2,
Then there exists \(c\in C_c^\infty ({\mathbb {R}})\) such that
It follows from inequalities (4.6) and (4.7) that
Since \(c+b(\infty )\in C_c^\infty ({\mathbb {R}})\dot{+}{\mathbb {C}}\subset SO^3\), we get \(a\in {\text {clos}}_{{\mathcal {M}}_{X({\mathbb {R}},w)}}(SO^3)=SO_{X({\mathbb {R}},w)}\). \(\square \)
5 Almost periodic Fourier multipliers and their products with continuous Fourier multipliers vanishing at infinity
5.1 The algebra \(AP_{X({\mathbb {R}},w)}\) of almost periodic Fourier multipliers
For \(\lambda \in {\mathbb {R}}\), let \(T_\lambda \) denote the translation operator defined by
Lemma 5.1
Let \(X({\mathbb {R}})\) be a rearrangement-invariant Banach function space and \(w:{\mathbb {R}}\rightarrow [0,\infty ]\) be a weight such that \(w\in X_{\mathrm{loc}}({\mathbb {R}})\) and \(1/w\in X_{\mathrm{loc}}'({\mathbb {R}})\). Suppose that \(\lambda \in {\mathbb {R}}\). Then the translation operator \(T_\lambda \) is bounded on the Banach function space \(X({\mathbb {R}},w)\) if and only if the function
belongs to the space \(L^\infty ({\mathbb {R}})\). In that case \(\Vert T_\lambda \Vert _{{\mathcal {B}}(X({\mathbb {R}},w))}=\Vert v_\lambda \Vert _{L^\infty ({\mathbb {R}})}\).
Proof
The operator \(T_\lambda \) is bounded on the space \(X({\mathbb {R}},w)\) if and only if the operator \(wT_\lambda w^{-1}I=T_\lambda (v_\lambda I)\) is bounded on the space \(X({\mathbb {R}})\). Moreover, their norms coincide. It is easy to see that for every \(f\in X({\mathbb {R}})\), the function \(T_\lambda f\) is equimeasurable with f, whence \(\Vert T_\lambda f\Vert _{X({\mathbb {R}})}=\Vert f\Vert _{X({\mathbb {R}})}\). Therefore,
By [31, Theorem 1], the multiplication operator \(v_\lambda I\) is bounded on the space \(X({\mathbb {R}})\) if and only if \(v_\lambda \in L^\infty ({\mathbb {R}})\) and \(\Vert v_\lambda I\Vert _{{\mathcal {B}}(X({\mathbb {R}}))}=\Vert v_\lambda \Vert _{L^\infty ({\mathbb {R}})}\). Thus, \(\Vert T_\lambda \Vert _{{\mathcal {B}}(X({\mathbb {R}},w))}=\Vert v_\lambda \Vert _{L^\infty ({\mathbb {R}})}\). \(\square \)
As a consequence of the previous result, we show that for all \(\lambda \in {\mathbb {R}}\), the exponential functions \(e_\lambda (x)=e^{i\lambda x}\), \(x\in {\mathbb {R}}\), are Fourier multipliers on separable rearrangement-invariant Banach function spaces with weights in the sublclass \(A_{1/\alpha _X}^0({\mathbb {R}})\cap A_{1/\beta _X}^0({\mathbb {R}})\) of the class of Muckenhoupt weights.
Corollary 5.2
Let \(X({\mathbb {R}})\) be a separable rearrangement-invariant Banach function space with the Boyd indices satisfying \(0<\alpha _X\le \beta _X<1\). Suppose that a weight w belongs to \(A_{1/\alpha _X}^0({\mathbb {R}})\cap A_{1/\beta _X}^0({\mathbb {R}})\). Then for every \(\lambda \in {\mathbb {R}}\), the function \(e_\lambda \) belongs to \({\mathcal {M}}_{X({\mathbb {R}},w)}\) and \(\Vert e_\lambda \Vert _{{\mathcal {M}}_{X({\mathbb {R}},w)}}=\Vert v_\lambda \Vert _{L^\infty ({\mathbb {R}})}\).
Proof
It follows from the definition of the classes \(A_{1/\alpha _X}^0({\mathbb {R}})\) and \(A_{1/\beta _X}^0({\mathbb {R}})\) that the function \(v_\lambda (x)=\frac{w(x+\lambda )}{w(x)}\), \(x\in {\mathbb {R}}\), is bounded for every \(\lambda \in {\mathbb {R}}\). By Lemma 2.3(a), \(w\in X_{\mathrm{loc}}({\mathbb {R}})\) and \(1/w\in X_{\mathrm{loc}}'({\mathbb {R}})\). Then, by Lemma 5.1, the operator \(T_\lambda \) is bounded on the Banach function space \(X({\mathbb {R}},w)\) and
It remains to observe that \(T_\lambda =W^0(e_\lambda )\). Thus \(e_\lambda \in {\mathcal {M}}_{X({\mathbb {R}},w)}\) and
which completes the proof. \(\square \)
Corollary 5.2 implies that if \(X({\mathbb {R}})\) is a separable rearrangement-invariant Banach function spaces and \(w\in A_{1/\alpha _X}^0({\mathbb {R}})\cap A_{1/\beta _X}^0({\mathbb {R}})\), then \(APP\subset {\mathcal {M}}_{X({\mathbb {R}},w)}\). We define the algebra \(AP_{X({\mathbb {R}},w)}\) of almost periodic Fourier multipliers by
It is natural to refer to the weights in \(A_{1/\alpha _X}^0\cap A_{1/\beta _X}^0\) as suitable Muckenhoupt weights. The class of suitable Muckenhoput weights contains many nontrivial weights as the following example shows.
For \(\delta ,\nu ,\eta \in {\mathbb {R}}\), consider the weight
Let \(r\in (1,\infty )\). It was shown in [27, Example 4.2] that if
then \(w\in A_r^0({\mathbb {R}})\). Hence if \(0<\alpha _X\le \beta _X<1\) and
then \(w\in A_{1/\alpha _X}^0({\mathbb {R}})\cap A_{1/\beta _X}^0({\mathbb {R}})\).
5.2 Products of almost periodic Fourier multipliers and continuous Fourier multipliers vanishing at infinity
The next lemma generalizes [29, Lemma 3.1(iii)] from the setting of Lebesgue spaces to the setting of rearrangement-invariant Banach function spaces with suitable Muckenhoupt weights.
Lemma 5.3
Let \(X({\mathbb {R}})\) be a separable rearrangement-invariant Banach function space with the Boyd indices satisfying \(0<\alpha _X\), \(\beta _X<1\). Suppose that w belongs to \(A_{1/\alpha _X}^0({\mathbb {R}})\cap A_{1/\beta _X}^0({\mathbb {R}})\) and \(C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\) is defined by (3.17). If \(a\in AP_{X({\mathbb {R}},w)}\) and \(\psi \in C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\), then \(a\psi \in C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\).
Proof
By Theorem 3.3, there exists a sequence \(\{\psi _n\}_{n\in {\mathbb {N}}}\subset C_c^\infty ({\mathbb {R}})\) such that
By the definition of the algebra \(AP_{X({\mathbb {R}},w)}\), there exists a sequence \(a_n\in APP\) such that
Then \(a_n\psi _n\in C_c^\infty ({\mathbb {R}})\subset C({{\dot{{\mathbb {R}}}}})\cap V({\mathbb {R}})\) for every \(n\in {\mathbb {N}}\). Moreover, (5.1)–(5.2) imply that
Hence \(a\psi \in C_{X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\). In view of the continuous embedding of \({\mathcal {M}}_{X({\mathbb {R}},w)}\) into \(L^\infty ({\mathbb {R}})\) (see Theorem 2.4) and the above equality, we obtain
Thus \((a\psi )(\infty )=0\) and \(a\psi \in C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\). \(\square \)
6 Proof of the main result
6.1 The algebra \({\mathcal {A}}_u\)
For a real-valued monotonically increasing function \(u\in C({\overline{{\mathbb {R}}}})\) such that
consider the set
Lemma 6.1
Let \(X({\mathbb {R}})\) be a separable rearrangement-invariant Banach function space with the Boyd indices satisfying \(0<\alpha _X\), \(\beta _X<1\). Suppose that \(w\in A_{1/\alpha _X}^0({\mathbb {R}})\cap A_{1/\beta _X}^0({\mathbb {R}})\). If \(u\in C({\overline{{\mathbb {R}}}})\) is a real-valued monotonically increasing function such that \(u(-\infty )=0\) and \(u(+\infty )=1\), then the set \({\mathcal {A}}_u\) is an algebra and the mappings \(a\mapsto a_l\) and \(a\mapsto a_r\) are algebraic homomorphisms of \({\mathcal {A}}_u\) onto \(AP_{X({\mathbb {R}},w)}\).
Proof
If \(a,b\in {\mathcal {A}}_u\), then
with some \(a_l,a_r,b_l,b_r\in AP_{X({\mathbb {R}},w)}\) and \(a_0,b_0\in C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\). Therefore
and
where
Since \(1-u,u\in C({\overline{{\mathbb {R}}}})\cap V({\mathbb {R}})\subset C_{X({\mathbb {R}},w)}({\overline{{\mathbb {R}}}})\) and \(a_0,b_0\in C_{0,X({\mathbb {R}},w)}({\mathbb {R}})\), it follows from Lemma 4.1 that
Then, by Lemma 5.3,
Since \(u-u^2\in C({\overline{{\mathbb {R}}}})\cap V({\mathbb {R}})\subset C_{X({\mathbb {R}},w)}({\overline{{\mathbb {R}}}})\) and \(u(\pm \infty )-u^2(\pm \infty )=0\), by Lemma 4.1, \(u-u^2\in C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\). Then, in view of Lemma 5.3, we also conclude that
It follows from (6.4) to (6.6) that \(c_0\in C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\). In view of this observation and equalities (6.2)–(6.3), we see that \(a+b,ab\in {\mathcal {A}}_u\). Therefore, \({\mathcal {A}}_u\) is an algebra. It is clear that the mappings \(a\mapsto a_l\) and \(a\mapsto a_r\) are algebraic homomorphisms of \({\mathcal {A}}_u\) onto \(AP_{X({\mathbb {R}},w)}\). \(\square \)
6.2 The multiplier norm of \(a=(1-u)a_r +ua_r+a_0\in {\mathcal {A}}_u\) dominates the multiplier norms of \(a_r\) and \(a_l\)
In this section we will prepare the proof of the fact that the algebraic homomorphisms \({\mathcal {A}}_u\rightarrow AP_{X({\mathbb {R}},w)}\) given by \(a\mapsto a_l\) and \(a\mapsto a_r\) are actually Banach algebra homomorphisms of norm 1. To this end, we will show that for \(a\in {\mathcal {A}}_u\),
For \(a\in L^\infty ({\mathbb {R}})\) and \(h\in {\mathbb {R}}\), we define
The following consequence of Kronecker’s theorem (see, e.g., [10, Theorem 1.12]) plays a crucial role in the proof of inequalities (6.7).
Lemma 6.2
If \(a_1,\ldots ,a_k\in APP\) is a finite collection of almost periodic polynomials, then there exists a sequence \(\{h_n\}_{n\in {\mathbb {N}}}\) of real numbers such that \(h_n\rightarrow +\infty \) as \(n\rightarrow \infty \) and
for each \(m\in \{1,\ldots ,k\}\).
For the sign “+”, the proof of the above lemma is given in [10, Lemma 10.2], for the sign “–”, the proof is analogous.
We start the proof of inequalities (6.7) for \(a=(1-v)a_l+va_r+a_0\) with a nice function v in place of u and nice functions \(a_l,a_r\) and \(a_0\).
Lemma 6.3
Let \(X({\mathbb {R}})\) be a separable rearrangement-invariant Banach function space with the Boyd indices satisfying \(0<\alpha _X\), \(\beta _X<1\). Suppose that \(w\in A_{1/\alpha _X}^0({\mathbb {R}})\cap A_{1/\beta _X}^0({\mathbb {R}})\). Let \(v\in C({\overline{{\mathbb {R}}}})\) be any real-valued monotonically increasing function such that there exists a point \(x_0>0\) such that \(v(x)=0\) for \(x<-x_0\) and \(v(x)=1\) for \(x>x_0\). If \(a_l,a_r\in APP\), \(a_0\in C_c^\infty ({\mathbb {R}})\), and
then inequalities (6.7) hold.
Proof
The idea of the proof is borrowed from the proof of [27, Theorem 3.1]. By Lemma 6.2, there is a sequence \(\{h_n\}_{n\in {\mathbb {N}}}\) of real numbers such that \(h_n\rightarrow +\infty \) as \(n\rightarrow \infty \) and
Let us show that
on the space \(X({\mathbb {R}},w)\). As
we have to prove that for every \(f\in X({\mathbb {R}},w)\),
Since the operators \(W^0(a^{h_n}-a_r)\) and \(W^0(a^{-h_n}-a_l)\) are uniformly bounded in \(n\in {\mathbb {N}}\) and the set \({\mathcal {S}}_0({\mathbb {R}})\) is dense in the space \(X({\mathbb {R}},w)\) in view of Lemma 2.3, applying [34, Lemma 1.4.1], we conclude that it is enough to prove equalities (6.12)–(6.13) for all \(f\in {\mathcal {S}}_0({\mathbb {R}})\).
Fix \(f\in {\mathcal {S}}_0({\mathbb {R}})\). Then, by a smooth version of Urysohn’s lemma (see, e.g., [17, Proposition 6.5]), there is a function \(\psi \in C_c^\infty ({\mathbb {R}})\) such that \(0\le \psi \le 1\), \({\text {supp}}{\mathcal {F}}f\subset {\text {supp}}\psi \) and \(\psi |_{{\text {supp}}{\mathcal {F}}f}=1\). Therefore, for all \(n\in {\mathbb {N}}\),
and
Since \(v(x)=1\) for \(x>x_0\) and \(v(x)=0\) for \(x<-x_0\) and \(a_0\in C_c^\infty ({\mathbb {R}})\), there exists \(N\in {\mathbb {N}}\) such that for all \(x\in {\text {supp}}\psi \) and \(n>N\),
Hence, for all \(n>N\) and \(x\in {\mathbb {R}}\),
It is clear that the functions on the right-hand sides of (6.16)–(6.17) belong to \(C_c^\infty ({\mathbb {R}})\). Therefore, by the Stechkin-type inequality (1.4), for all \(n>N\),
and, analogously,
Combining (6.14)–(6.15) and (6.18)–(6.19) with (6.9)–(6.10), we see that equalities (6.12)–(6.13) hold for every \(f\in {\mathcal {S}}_0({\mathbb {R}})\). Therefore, (6.11) are fulfilled for every \(f\in X({\mathbb {R}},w)\). Hence, by the Banach-Steinhaus theorem (see, e.g., [34, Theorem 1.4.2]),
and, analogously,
which completes the proof of (6.7). \(\square \)
Now we extend the previous result for functions a of the form (6.8) with general \(a_l,a_r\in AP_{X({\mathbb {R}},w)}\) and \(a_0\in C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\), keeping the nice function v as above.
Lemma 6.4
Let \(X({\mathbb {R}})\) be a separable rearrangement-invariant Banach function space with the Boyd indices satisfying \(0<\alpha _X\), \(\beta _X<1\). Suppose that \(w\in A_{1/\alpha _X}^0({\mathbb {R}})\cap A_{1/\beta _X}^0({\mathbb {R}})\). Let \(v\in C({\overline{{\mathbb {R}}}})\) be any real-valued monotonically increasing function such that there exists a point \(x_0>0\) such that \(v(x)=0\) for \(x<-x_0\) and \(v(x)=1\) for \(x>x_0\). If \(a_l,a_r\in AP_{X({\mathbb {R}},w)}\), \(a_0\in C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\), where \(C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\) is defined by (3.17), and a is given by equality (6.8), then inequalities (6.7) hold.
Proof
By the definition of \(AP_{X({\mathbb {R}},w)}\), there are sequences \(\{a_l^{(n)}\}_{n\in {\mathbb {N}}},\{a_r^{(n)}\}_{n\in {\mathbb {N}}}\) in APP such that
On the other hand, by Theorem 3.3, there is a sequence \(\{a_0^{(n)}\}_{n\in {\mathbb {N}}}\) in \(C_c^\infty ({\mathbb {R}})\) such that
For \(n\in {\mathbb {N}}\), consider the functions
It follows from equalities (6.20)–(6.22) and Lemma 6.3 that
which completes the proof of inequalities (6.7). \(\square \)
Now we observe that the algebra \({\mathcal {A}}_u\) does not depend on the particular choice of a real-valued monotonically increasing function \(u\in C({\overline{{\mathbb {R}}}})\) satisfying conditions (6.1).
Lemma 6.5
Let \(X({\mathbb {R}})\) be a separable rearrangement-invariant Banach function space with the Boyd indices satisfying \(0<\alpha _X\), \(\beta _X<1\). Suppose that \(w\in A_{1/\alpha _X}^0({\mathbb {R}})\cap A_{1/\beta _X}^0({\mathbb {R}})\). Let \(u,v\in C({\overline{{\mathbb {R}}}})\) be two real-valued monotonically increasing functions such that
Then \({\mathcal {A}}_u={\mathcal {A}}_v\).
Proof
If \(a\in {\mathcal {A}}_u\), then \(a=(1-u)a_l+ua_r+a_0\) for some \(a_l,a_r\in AP_{X({\mathbb {R}},w)}\) and \(a_0\in C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\). On the other hand, \(a=(1-v)a_l+va_r+b_0\) with
Since the functions u, v are monotonically increasing, we have \(u,v\in V({\mathbb {R}})\). Hence \(u-v\in V({\mathbb {R}})\cap C({\overline{{\mathbb {R}}}})\) and
Thus \(u-v\in C({{\dot{{\mathbb {R}}}}})\cap V({\mathbb {R}})\subset C_{X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\) and \((u-v)(\infty )=0\). Since the function \(a_r-a_l\) belongs to \(AP_{X({\mathbb {R}},w)}\), it follows from Lemma 5.3 that
Then \(b_0\in C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\) and \(a\in {\mathcal {A}}_v\). Therefore \({\mathcal {A}}_u\subset {\mathcal {A}}_v\). It can be shown analogously that \({\mathcal {A}}_v\subset {\mathcal {A}}_u\). Thus \({\mathcal {A}}_u={\mathcal {A}}_v\). \(\square \)
Combining Lemmas 6.4–6.5, we arrive at the main result of this subsection.
Theorem 6.6
Let \(X({\mathbb {R}})\) be a separable rearrangement-invariant Banach function space with the Boyd indices satisfying \(0<\alpha _X\), \(\beta _X<1\). Suppose that \(w\in A_{1/\alpha _X}^0({\mathbb {R}})\cap A_{1/\beta _X}^0({\mathbb {R}})\). Let \(u\in C({\overline{{\mathbb {R}}}})\) be a real-valued monotonically increasing function such that \(u(-\infty )=0\) and \(u(+\infty )=1\). If \(a\in {\mathcal {A}}_u\), that is,
where \(C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\) is defined by (3.17), then inequalities (6.7) hold.
6.3 Proof of Theorem 1.2
The idea of the proof is borrowed from the proof of [10, Theorem 1.21]. If \(a\in AP_{X({\mathbb {R}},w)}\), then \(a=(1-u)a+ua+0\), whence \(a\in {\mathcal {A}}_u\). If \(f\in C_{X({\mathbb {R}},w)}({\overline{{\mathbb {R}}}})\), then the function \(f_0=f-(1-u)f(-\infty )-uf(+\infty )\) belongs to \(C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\). Therefore \(f=(1-u)f(-\infty )+uf(+\infty )+f_0\in {\mathcal {A}}_u\). These observations imply that
On the other hand, it is obvious that
Combining (6.23)–(6.24), we arrive at the equality
By Theorem 6.6, for every \(a=(1-u)a_r+ua_r+a_0\in {\mathcal {A}}_u\) with \(a_l,a_r\in AP_{X({\mathbb {R}},w)}\) and \(a_0\in C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\), one has
Consequently, if \(\big \{(1-u)a_l^{(n)}+ua_r^{(n)}+a_0^{(n)}\big \}_{n\in {\mathbb {N}}}\) is a Cauchy sequence in \({\mathcal {A}}_u\), where \(\big \{a_l^{(n)}\big \}_{n\in {\mathbb {N}}}, \big \{a_r^{(n)}\big \}_{n\in {\mathbb {N}}}\) are sequences in \(AP_{X({\mathbb {R}},w)}\) and \(\big \{a_0^{(n)}\big \}_{n\in {\mathbb {N}}}\) is a sequence in \(C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\), then \(\big \{a_l^{(n)}\big \}_{n\in {\mathbb {N}}}\) and \(\big \{a_r^{(n)}\big \}_{n\in {\mathbb {N}}}\) are Cauchy sequences in \(AP_{X({\mathbb {R}},w)}\). Consequently, \(\big \{a_0^{(n)}\big \}_{n\in {\mathbb {N}}}\) is a Cauchy sequence in \(C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\). Since \(AP_{X({\mathbb {R}},w)}\) is closed by its definition and \(C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\) is closed in view of Theorem 3.3, we conclude that the limits
belong to \(AP_{X({\mathbb {R}},w)}\) and that the limit
belongs to \(C_{0,X({\mathbb {R}},w)}({{\dot{{\mathbb {R}}}}})\). Therefore, the limit
belongs to \({\mathcal {A}}_u\). Thus
It follows from (6.25) and (6.27) that \({\mathcal {A}}_u=SAP_{X({\mathbb {R}},w)}\). In particular, every function \(a\in SAP_{X({\mathbb {R}},w)}\) is of the form
with \(a_l,a_r\in AP_{X({\mathbb {R}},w)}\) and \(a_0\in C_{0,X({\mathbb {R}},w)}\). We infer from (6.26) that the representation (6.28) is unique for the function u. Moreover, the proof of Lemma 6.5 shows that \(a_l,a_r\in AP_{X({\mathbb {R}},w)}\) are independent of the particular choice of the function u. By Lemma 6.1, the mappings \(a\mapsto a_l\) and \(a\mapsto a_r\) are algebraic homomorphisms of \({\mathcal {A}}_u=SAP_{X({\mathbb {R}},w)}\) onto \(AP_{X({\mathbb {R}},w)}\). In view of (6.26), these homomorphisms are Banach algebra homomorphisms of the Banach algebra \(SAP_{X({\mathbb {R}},w)}\) onto the Banach algebra \(AP_{X({\mathbb {R}},w)}\) and the norms of these homomorphisms are not greater than one. For every function \(a\in AP_{X({\mathbb {R}},w)}\), we have equalities in (6.26) because
Thus, the norms of the homomorphisms \(a\mapsto a_l\) and \(a\mapsto a_l\) are equal to one. \(\square \)
References
Bennett, C., Sharpley, R.: Interpolation of Operators. Academic Press, Boston (1988)
Bogveradze, G., Castro, L.P.: On the Fredholm index of matrix Wiener–Hopf plus/minus Hankel operators with semi-almost periodic symbols. Oper. Theor. Adv. Appl. 181, 143–158 (2008)
Böttcher, A., Grudsky, S.M., Spitkovsky, I.M.: The spectrum is discontinuous on the manifold of Toeplitz operators. Arch. Math. 75, 46–52 (2000)
Böttcher, A., Grudsky, S.M., Spitkovsky, I.M.: Toeplitz operators with frequency modulated semi-almost periodic symbols. J. Fourier Anal. Appl. 7, 523–535 (2001)
Böttcher, A., Grudsky, S.M., Spitkovsky, I.M.: On the essential spectrum of Toeplitz operators with semi-almost periodic symbols. Oper. Theor. Adv. Appl. 142, 59–77 (2003)
Böttcher, A., Grudsky, S.M., Spitkovsky, I.M.: Block Toeplitz operators with frequency-modulated semi-almost periodic symbols. Int. J. Math. Math. Sci. 34, 2157–2176 (2003)
Böttcher, A., Karlovich, Y.I.: Carleson Curves, Muckenhoupt Weights, and Toeplitz Operators. Birkhäuser, Basel (1997)
Böttcher, A., Karlovich, Y.I., Spitkovsky, I.M.: Toeplitz operators with semi-almost periodic symbols on spaces with Muckenhoupt weight. Integral Equ. Oper. Theory 18, 261–276 (1994)
Böttcher, A., Karlovich, Y.I., Spitkovsky, I.M.: Toeplitz operators with semi-almost-periodic matrix symbols on Hardy spaces. Acta Appl. Math. 65(1–3), 115–136 (2001)
Böttcher, A., Karlovich, Y.I., Spitkovsky, I.M.: Convolution Operators and Factorization of Almost Periodic Matrix Functions. Birkhäuser, Basel (2002)
Boyd, D.W.: Indices of function spaces and their relationship to interpolation. Can. J. Math. 21, 1245–1254 (1969)
Brezis, H.: Functional Analysis. Sobolev Spaces and Partial Differential Equations. Springer, New York (2011)
Duduchava, R.V.: Integral Equations with Fixed Singularities. Teubner, Leipzig (1979)
Duduchava, R.V., Saginashvili, A.I.: Convolution integral equations on a half-line with semi-almost periodic presymbols. Differ. Equ. 17, 207–216 (1981)
Fernandes, C.A., Karlovich, A.Y., Karlovich, Y.I.: Noncompactness of Fourier convolution operators on Banach function spaces. Ann. Funct. Anal. AFA 10, 553–561 (2019)
Fernandes, C.A., Karlovich, A.Y., Karlovich, Y.I.: Fourier convolution operators with symbols equivalent to zero at infinity on Banach function spaces. Submitted. Preprint is available at arXiv:1909.13538 [math.FA] (2019)
Folland, G.B.: A Guide to Advanced Real Analysis. The Mathematical Association of America, Washington, DC (2009)
Grafakos, L.: Classical Fourier Analysis, 3rd edn. Springer, New York (2014)
Grudsky, S.M.; The Riemann boundary value problem with semi-almost periodic discontinuities in the space \(L_p(\Gamma ,\varrho )\). In: Integral and Differential Equations and Approximate Solutions, Elista, Kalmytsk. Gos. Univ., pp. 54–68, in Russian (1985)
Hörmander, L.: Estimates for translation invariant operators in \(L^p\) spaces. Acta Math. 104, 93–140 (1960)
Karlovich, A.Y.: Commutators of convolution type operators on some Banach function spaces. Ann. Funct. Anal. AFA 6, 191–205 (2015)
Karlovich, A.Y.: Banach algebra of the Fourier multipliers on weighted Banach function spaces. Concr. Oper. 2, 27–36 (2015)
Karlovich, A.Y.: Algebras of continuous Fourier multipliers on variable Lebesgue spaces. Submitted. Preprint is availabale at arXiv:1903.09696 [math.CA] (2019)
Karlovich, A., Shargorodsky, E.: When does the norm of a Fourier multiplier dominate its \(L^\infty \) norm? Proc. Lond. Math. Soc. 118, 901–941 (2019)
Karlovich, A.Y., Spitkovsky, I.M.: On singular integral operators with semi-almost periodic coefficients on variable Lebesgue spaces. J. Math. Anal. Appl. 384, 706–725 (2011)
Karlovich, A.Y., Spitkovsky, I.M.: The Cauchy singular integral operator on weighted variable Lebesgue spaces. Oper. Theor. Adv. Appl. 236, 275–291 (2014)
Karlovich, Y.I., Loreto Hernández, J.: Wiener–Hopf operators with semi-almost periodic matrix symbols on weighted Lebesgue spaces. Integral Equ. Oper. Theor. 62, 85–128 (2008)
Karlovich, Y.I., Loreto Hernández, J.: Wiener-Hopf operators with slowly oscillating matrix symbols on weighted Lebesgue spaces. Integral Equ. Oper. Theory 64, 203–237 (2009)
Karlovich, Y.I., Spitkovsky, I.M.: (Semi)-Fredholmness of convolution operators on the spaces of Bessel potentials. Oper. Theor. Adv. Appl. 71, 122–152 (1994)
Lindenstrauss, J., Tzafriri, L.: Classical Banach Spaces. Function Spaces. Springer, Berlin, II (1979)
Maligranda, L., Persson, L.E.: Generalized duality of some Banach function spaces. Indag. Math. 51, 323–338 (1989)
Nolasco, A.P., Castro, L.P.: A Duduchava–Saginashvili’s type theory for Wiener–Hopf plus Hankel operators. J. Math. Anal. Appl. 331, 329–341 (2007)
Nolasco, A.P., Castro, L.P.: A stronger version of the Sarason’s type theorem for Wiener–Hankel operators with \(SAP\) Fourier symbols. Bull. Greek Math. Soc. 54, 59–77 (2007)
Roch, S., Santos, P., Silbermann, B.: Non-Commutative Gelfand Theories. A Tool-kit for Operator Theorists and Numerical Analysts. Springer, Berlin (2011)
Saginashvili, A.I.: Singular integral equations with coefficients having discontinuities of semi-almost-periodic type. Trudy Tbilis. Mat. Inst. Razmadze 66, 84–95, in Russian (1980). English translation: Transl., Ser. 2, Am. Math. Soc. 127, 49–59 (1985)
Sarason, D.: Toeplitz operators with semi-almost periodic symbols. Duke Math. J. 44, 357–364 (1977)
Author information
Authors and Affiliations
Corresponding author
Additional information
Dedicated to Professor Yuri I. Karlovich on the occasion of his 70th birthday.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This work was partially supported by the Fundação para a Ciência e a Tecnologia (Portuguese Foundation for Science and Technology) through the projects UID/MAT/00297/2019 (Centro de Matemática e Aplicações).
Rights and permissions
About this article
Cite this article
Fernandes, C.A., Karlovich, A.Y. Semi-almost periodic Fourier multipliers on rearrangement-invariant spaces with suitable Muckenhoupt weights. Bol. Soc. Mat. Mex. 26, 1135–1162 (2020). https://doi.org/10.1007/s40590-020-00276-1
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s40590-020-00276-1
Keywords
- Rearrangement-invariant Banach function space
- Boyd indices
- Muckenhoupt weight
- Almost periodic function
- Semi-almost periodic function
- Fourier multiplier