Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Asymptotic boundedness and moment asymptotic expansion in ultradistribution spaces

Asymptotic boundedness and moment asymptotic expansion in ultradistribution spaces ASYMPTOTIC BOUNDEDNESS AND MOMENT ASYMPTOTIC EXPANSION IN ULTRADISTRIBUTION SPACES LENNY NEYT AND JASSON VINDAS Dedicated to the memory of Prof. Bogoljub Stankovi´c Abstract. We obtain structural theorems for the so-called S-asymptotic and quasi- asymptotic boundedness of ultradistributions. Using these results, we then analyze the moment asymptotic expansion (MAE), providing a full characterization of those ultradistributions satisfying this asymptotic formula in the one-dimensional case. We also introduce and study a uniform variant of the MAE. 1. Introduction In this article we study asymptotic properties of ultradistributions. Asymptotic analysis is an important subject in the theory of generalized functions and provides powerful tools for applications in areas such as mathematical physics, number theory, and differential equations. The theory of asymptotic behavior of generalized functions has been particularly useful in the study of Tauberian theorems for integral transforms. We refer to the monographs [7, 18, 22] for complete accounts on the subject and its many applications. The asymptotic behavior of a generalized function is usually analyzed via its para- metric behavior, mostly with respect to translation or dilation. The idea of looking at the translates of a distribution goes back to Schwartz [21, Chapter VII], who used it to measure the order of growth of tempered distributions at infinity. Pilipovi´c and Stankovi´c later introduced a generalization, the so called S-asymptotic behavior, and thoroughly investigated its properties for distributions, ultradistributions, and Fourier hyperfunctions. There are deep connections between S-asymptotics and Wiener Taube- rian theorems for generalized functions [15]. In Section 3 of this article we study the structure of S-asymptotically bounded ultradistributions; in fact, we discuss a coun- terpart of the main structural result from [17] (cf. [18, Theorem 1.10, p. 46]) for S-asymptotic boundedness, which we prove to hold under much weaker assumptions than those employed there. There are two very well-established approaches to asymptotics of generalized func- tions related to dilation. The first one is the quasiasymptotic behavior, which employs regularly varying functions [1] as gauges in the asymptotic comparisons. This concept 2010 Mathematics Subject Classification. 46F05; 46F10. Key words and phrases. Asymptotic behavior of generalized functions; moment asymptotic expan- sions; S-asymptotics; quasiasymptotics; ultradistribution spaces. L. Neyt gratefully acknowledges support by Ghent University, through the BOF-grant 01J11615. J. Vindas was supported by Ghent University through the BOF-grants 01J11615 and 01J04017. arXiv:1906.06232v2 [math.FA] 21 Nov 2019 2 L. NEYT AND J. VINDAS was first introduced by Zav’yalov for Schwartz distributions [25] and further devel- oped by him, Drozhzhinov, and Vladimirov in [22]. The quasiasymptotic behavior was naturally extended to the context of one-dimensional ultradistributions in [16]. Here we shall provide complete structural theorems for quasiasymptotically bounded ultra- distributions in Section 4, both at infinity and the origin, using the above-mentioned results for S-asymptotic boundedness and techniques developed by the authors in [13]. These results are the ultradistributional analogs of the structural theorems from [23] for Schwartz distributions (cf. [18, Section 2.12, p. 160]). The second important approach to asymptotic behavior related to dilation is the so- called moment asymptotic expansion (MAE), whose properties have been extensively investigated by Estrada and Kanwal [6, 7]. Some recent contributions can be found in [20, 24]. A generalized function f is said to satisfy the MAE if there is a certain multisequence tµ u d, called the moments of f, such that the following asymptotic αPN expansion holds |α| pαq p´1q µ δ pxq (1.1) fpλxq „ , λ Ñ 8. |α|`d α!λ αPN As is shown in the monograph [7], the MAE supplies a unified approach to several aspects of asymptotic analysis and its applications. Interestingly, Estrada characterized [5] the largest space of distributions where the MAE holds as the dual of the space of so-called GLS symbols [9]. We will consider the MAE for ultradistributions, providing in Section 5 a counterpart of Estrada’s full characterization in the one-dimensional case. We shall also study a uniform version of (1.1) in Section 6, which we call the UMAE. Our considerations naturally lead to introduce the ultradistribution spaces 1˚ d 1˚ d K pR q and K pR q, which are intimately connected with the MAE and UMAE. We 1˚ 2d note that in even dimension our space K pR q arises as the dual of one of the spaces of symbols of ‘infinite order’ pseudo-differential operators from [19]. 2. Preliminaries In this section we fix the notation and briefly collect some background material on ultradistributions. Given a weight sequence tM u of positive real numbers, we will p pPN often make use of one or more of the following conditions: pM.1q: M ď M M , p ě 1; p´1 p`1 1 p pM.2q : M ď AH M , p P N, for certain constants A and H; p`1 p p`q pM.2q: M ď AH M M , p, q P N, for certain constants A and H; p`q p q pM.3q : M {M ă 8; p´1 p p“1 pM.3q: M {M ď c qM {M , q ě 1, for a certain constant c . p´1 p 0 q´1 q 0 p“q The meaning of all these conditions is very well explained in [10]. For multi-indices α P N , we will simply denote M as M . |α| α d d Let Ω Ă R be open and let K Ť R be a (regular) compact subset. For any ℓ ą 0 M ,ℓ we denote as E pKq the space of all smooth functions ϕ for which pαq |ϕ pxq| kϕk :“ sup sup M ,ℓ E pKq |α| ℓ M xPK α αPN ASYMPTOTIC BOUNDEDNESS AND MOMENT ASYMPTOTIC EXPANSION 3 is finite. Then, as customary [10], we set pM q M ,ℓ tM u M ,ℓ p p p p E pΩq “ lim lim E pKq and E pΩq “ lim lim E pKq, ÐÝ ÐÝ ÐÝ ÝÑ KŤΩ ℓą0 KŤΩ ℓą0 and use E pΩq as the common notation for the Beurling and Roumieu case (when a separate treatment is needed, we will always first state assertions about the Beurling case pM q, then followed by the Roumieu case tM u in parenthesis). p p tM u We shall make use of Komatsu’s well-known projective limit description of E pΩq. Consider the directed set R “ pℓ q : ℓ Õ 8 and ℓ ą 0, @p P Z . For any pℓ q P p pPZ p p ` p R we write L “ 1 and L “ ℓ for p P Z . Then, it is shown in [11, Proposition 0 p j ` j“1 1 1 3.5, p. 675] that if the weight sequence satisfies pM.1q, pM.2q , and pM.3q , then tM u pL M q p p p (2.1) E pΩq “ lim E pΩq ÐÝ pℓ qPR as locally convex spaces. For compactly supported test functions our notation is standard, we write D pΩq “ M ,ℓ ˚ M ,ℓ E pΩq X DpΩq and D “ E pKq X D with K Ť Ω, these spaces are topologized in the canonical way [10]. (Naturally, by the Denjoy-Carleman theorem and under ˚ 1 1˚ pM.1q, the non-triviality of D pΩq is equivalent to pM.3q .) The strong dual D pΩq is 1˚ the space of ultradistributions of class ˚, while the elements of E pΩq are exactly those ultradistributions with compact support. 3. S-asymptotic boundedness Our aim in this section is to study the structure of those ultradistributions that satisfy 1˚ d (3.1) fpx ` hq “ O pωphqq, h P W, in D pR q, where W Ď R is simply an unbounded set and ω is a positive function. The S- asymptotic relation is to be interpreted in the ultradistributional sense [18], that is, it ˚ d explicitly means that for each test function ϕ P D pR q, hfpx ` hq, ϕpxqi pf ˚ ϕˇqphq (3.2) sup “ sup ă 8. ωphq ωphq hPW hPW We begin with a simple auxiliary lemma that allows us to preserve certain growth properties when regularizing functions. Given R ą 0 and a set W , we denote as W the open R-neighborhood of W , that is, the set W “ W ` Bp0, Rq. Lemma 3.1. Given R ą 0 there are absolute constants c and c such that each 0,R 1,R function g P L pW q satisfying the bound sup |gpx ` hq|{ωpxq ă 8, where loc xPW, |h|ăR W Ă R and ω is a positive function defined on W , can be written as g “ Δg ` g in 1 0 W for some functions g P CpR q that satisfy R j |g pxq| |gpx ` hq| sup ď c sup , j “ 0, 1. j,R ωpxq ωpxq xPW xPW, |h|ăR 4 L. NEYT AND J. VINDAS Proof. To show this, we make use of the fact that the fundamental solutions of the 1 d 8 d Laplacian belong to L pR q X C pR zt0uq. By cutting-off a fundamental solution in loc 1 d d the ball Bp0, Rq, this implies we can select functions χ P L pR q and χ P DpR q both 1 0 supported on Bp0, Rq, such that δ “ Δχ ` χ . Extend g off W as 0 and keep calling 1 0 R this extension by g. We obtain the claim if we set g “ g ˚ χ so that the desired j j inequalities hold with c “ |χ p´xq|dx. j,R j |x|ďR We can now obtain our first structural theorem for (3.1) in the case where the weight sequence satisfies mild hypotheses. For it, we shall need to impose the following regularity condition on the gauge function ω, ωp ¨ ` xq 8 d (3.3) sup P L pR q. loc d ωpxq xPR In the proof of the next theorem we employ a technique of Go´mez-Collado that she applied to obtain various characterizations of the space of bounded ultradistributions in [8]. Theorem 3.2. Let W Ă R be an unbounded set and let ω be a positive measurable d 1 1 function on R that satisfies (3.3). Suppose pM.1q, pM.2q , and pM.3q hold. Then, 1˚ d an ultradistribution f P D pR q satisfies (3.1) if and only if for each R ą 0 there are continuous functions tf u d defined on W such that for some ℓ ą 0 (for each ℓ ą 0) α R αPN there exists C ą 0 for which the bounds |α| (3.4) |f pxq| ď C ωpxq, x P W , α P N , α ℓ R hold and pαq (3.5) f “ f in W . αPN Proof. The sufficiency of the conditions is easily verified. For the necessity, we start by making some reductions. In view of Lemma 3.1 and the assumption pM.2q , it suffices to establish (3.5) with (3.4) for functions that are merely measurable. Next, we show that we may assume that W “ R . Let R ą 0 be arbitrary and let χ be a non-negative smooth function on R such that 0 ď χ ď 1, χ “ 1 on W while χ “ 0 outside R R R R W , and such that 2R pαq |χ pξq| sup ă 8, @ℓ ą 0. |α| d d ℓ M αPN , ξPR α ˚ d r r Then, we set f :“ χ ¨f and notice that f and f coincide on W . Take any ϕ P D pR q R R and let r ą 0 be such that supp ϕ Ă Bp0, rq. Take any h P R . If h R W then r`2R D E fpx ` hq{ω rphq, ϕpxq “ 0. Suppose now h P W , then h “ h ` h with h P W r`2R 1 2 1 ASYMPTOTIC BOUNDEDNESS AND MOMENT ASYMPTOTIC EXPANSION 5 and h P Bp0, r ` 2Rq. Then, employing (3.3) and the Banach-Steinhaus theorem, D E D E r r | fpx ` hq, ϕpxq | | fpx ` h q, ϕpx ´ h q | 1 2 ď C “ O p1q , R`2ε ω rphq ωph q ˚ d because tT ϕ : h P Bp0, r ` 2Rqu is a bounded family in D pR q. Consequently, h 2 d 1˚ d fpx ` hq “ Opωphqq, h P R , in D pR q We may therefore assume w.l.o.g. that f “ f and derive (3.4) and (3.5) on the whole d d R for measurable functions f under the hypothesis that tfpx ` hq{ωphq : h P R u is 1˚ d ˚ a bounded subset of D pR q. We now reason as in [8]. Let ψ P D be such that r´1,1s d d ψpx ´ nq “ 1 for each x P R . We have that tψfp ¨ ` nq{ωpnq : n P Z u is now nPZ 1˚ d 1 a bounded set in the space E pR q. Using pM.2q , we obtain in the Beurling case the existence of ℓ ą 0 and in the Roumieu case of pℓ q P R such that for some C ą 0 and d ˚ d all n P Z and φ P E pR q ωpnq pαq (3.6) |xfpxq, ψpx ´ nqφpx ´ nqy| ď C |φ pxq|dx, L M d α α r´1,1s αPN where in the Roumieu case we have used the projective description (2.1) and we have set L “ ℓ in the Beurling case and L “ ℓ in the Roumieu one. We further on p p p jďp 8 d consider the Banach space X of all ϕ P C pR q such that ωpxq pαq kϕk “ |ϕ pxq| dx ă 8. d L M α α αPN ˚ d Let ϕ P D pR q be arbitrary. Applying (3.6) to each φpxq “ ϕpx ` nq and using the hypothesis (3.3), we obtain, with C “ C sup ωpx ` yq{ωpxq, d d xPR , yPr´1,1s d 1 |xf, ϕy| ď |xfpxq, ψpx ´ nqϕpxqy| ď 3 C kϕk . nPZ 1 ˚ d Note that pM.1q and pM.3q ensure that D pR q is dense in X and hence we conclude f P 1 1 d d |α| pαq X . Embedding X into L pN ˆ R , dµ q via the isometry jpϕqpα, xq “ p´1q ϕ pxq, where the measure is given by dµ “ ωpxq{pL M qdαdx with dα the natural counting α α measure on N , we can apply the Hanh-Banach theorem to get the representation (3.5) d 2 with measurable functions f on R that satisfy bounds |f pxq| ď C ωpxq{pL M q. α α α α This yields already the result in the Beurling case. In the Roumieu case we finally employ [11, Lemma 3.4(ii), p. 674] to obtain the bounds (3.4) for each ℓ ą 0 and some C ą 0. In applications it is very useful to combine Theorem 3.2 with the ensuing proposition, which provides conditions under which one might essentially apply Theorem 3.2 with a function ω that is just defined on the set W . 6 L. NEYT AND J. VINDAS Proposition 3.3. Let W Ă R be a closed convex set. Any positive function ω on W satisfying ωpx ` hq (3.7) p@R ą 0q sup ă 8 ωpxq x, x`hPW |h|ďR can be extended to a positive function on R satisfying (3.3). In addition, if ω is measurable (or continuous), the extension can be chosen measurable (or continuous) as well. Proof. For any x P R we denote by xr P W the (unique in view of convexity) closest point to x in W . Then, we set ω rpxq :“ ωpxrq. Since x ÞÑ xr is continuous, ω r inherits measurability or continuity if ω has the property. We now verify (3.3) for ω r. Let R ą 0 and let C an upper bound for ωpt ` yq{ωpyq, where y, t ` y P W and t P Bp0, Rq. Let x P R and h P Bp0, Rq be arbitrary. Consider the points x, x ` h, xr and x ` h. By the obtuse angle criterion, the angles defined by the line segments rx, xr, x ` hs and Ć Ć rx`h, x ` h, xrs are at least π{2, whence |xr´x ` h| ď |x´px`hq| ď R. It then follows that ω rpx ` hq ď C ω rpxq, as required. If the weight sequence satisfies stronger assumption, one can drop any regularity assumption on ω, as stated in the next result. Theorem 3.4. Let W Ă R be an unbounded set and let ω be a positive function on 1˚ d W . Suppose that pM.1q, pM.2q, and pM.3q hold. An ultradistribution f P D pR q satisfies (3.1) if and only if for each R ą 0 there are continuous functions tf u d αPN defined on W such that for some ℓ ą 0 (for each ℓ ą 0) there exists C ą 0 such that R ℓ |α| (3.8) |f px ` hq| ď C ωpxq, x P W, |h| ă R, α P N , α ℓ and the representation (3.5) holds. Proof. The proof is similar to that of [18, Theorem 1.10, p. 46], but we provide some simplifications. The converse is easy to show, so we concentrate on showing the necessity of the conditions for the S-asymptotic boundedness relation (3.1). Let R ą 0. ˚ d We consider the linear mapping A : D pR q Ñ X, with values in the Banach space X “ tg : W Ñ C : sup |gpxq|{ωpxq ă 8u, given by Aϕ “ f ˚ ϕ. It follows from xPW the closed graph theorem that A is continuous. Consequently, we obtain from the Banach-Steinhaus theorem the existence of ℓ ą 0 in the Beurling case or pℓ q P R in M L ,1 M L ,1 p p p p the Roumieu case such that A P pD q and f ˚ ϕ P X for each ϕ P D , where Bp0,2Rq Bp0,2Rq we set L “ ℓ in the Beurling case or L “ ℓ in the Roumieu case. Since for p p p j“1 M L ,1 M L ,1 p p p p each ϕ P D the set tT ϕ : |x| ď Ru is compact in D , we conclude that for Bp0,Rq Bp0,2Rq any such a ϕ the function f ˚ ϕ is continuous on W and pf ˚ ϕqpx ` hq sup ă 8. pωphqq hPW, |x|ăR ASYMPTOTIC BOUNDEDNESS AND MOMENT ASYMPTOTIC EXPANSION 7 We now employ the parametrix method. As shown in [12, p. 199], there is an ultradif- M L ,1 p p ferential operator PpDq of class ˚ that admits a D -parametrix, namely, for which Bp0,Rq M L ,1 p p there are χ P D and ϕ P D such that δ “ PpDqϕ`χ. Setting f “ f ˚χ and ¯ ¯ Bp0,Rq Bp0,Rq g “ f ˚ϕ, we obtain the decomposition f “ PpDqg `f , which in particular establishes the representation (3.5) with functions f satisfying the bounds (3.8). 4. Quasiasymptotic boundedness Our results from the previous section can be applied to obtain structural theorems for ultradistributions being quasiasymptotically bounded in dimension 1. Let ρ be a positive function defined on an interval of the form rλ , 8q. We are interested in the relation fpλxq “ Opρpλqq as λ Ñ 8 in ultradistribution spaces. The analog of the condition (3.7) for a function ρ in this multiplicative setting is being O-regularly varying (at infinity) [1, p. 65]. The latter means (cf. [1, Theorem 2.0.4, p. 64]) that ρ is measurable and for each R ą 1 ρpλxq lim sup sup ă 8. ´1 ρpλq xÑ8 λPrR ,Rs The next proposition can be established with the aid of Theorem 3.2 and Theorem 3.4 via an exponential change of variables as in the authors’ work [13, Lemma 3.4]; we omit details. 1˚ Proposition 4.1. Let f P D pRq and ρ be a positive function. Suppose that fpλxq “ 1˚ Opρpλqq as λ Ñ 8 in D pRzt0uq. 1 1 (i) If pM.1q, pM.2q , and pM.3q hold and ρ is O-regularly varying at infinity, then there are continuous functions f and x ą 0 such that m 0 pmq (4.1) f “ f on Rzr´x , x s 0 0 m“0 and for some ℓ ą 0 (for any ℓ ą 0) there is C ą 0 such that |x| ρp|x|q, |x| ą x , m P N. (4.2) |f pxq| ď C m ℓ 0 (ii) If pM.1q, pM.2q, and pM.3q hold, for each R ą 1 one can find x and continuous functions such that f has the representation (4.1), where the f satisfy the bounds m ´1 (4.3) |f paxq| ď C |x| ρp|x|q, |x| ą x , R ă a ă R, m P N, m ℓ for some ℓ ą 0 (for any ℓ ą 0). Remark 4.2. Clearly, (4.2) implies (4.3) for an O-regularly varying function ρ. Assume 1 1 pM.1q, pM.2q , and pM.3q hold. Notice the representations (4.1) with bounds (4.3) are 1˚ also sufficient to yield fpλxq “ Opρpλqq as λ Ñ 8 in D pRzt0uq, so that the converses of both parts (i) and (ii) of Proposition 4.1 are valid. 8 L. NEYT AND J. VINDAS In the rest of the section we are interested in describing quasiasymptotic boundedness 1˚ in the full space D pRq. For it, we need to impose stronger variation assumptions on the gauge function ρ. We call a positive measurable function O-slowly varying at infintiy if for each ε ą 0 there are C , c , R ą 0 such that ε ε ε c Lpλxq (4.4) ď ď C λ , λ ě 1, x ą R . ε ε λ Lpxq In the terminology from [1] this means that the upper and lower Matuszewska indices of L are both equal to 0. Thus, a function of the form ρpλq “ λ Lpλq is an O-regularly varying function with both upper and lower Matuszewska indices equal to q P R. The authors have found in [13] complete structural theorems for the quasiasymp- totic behavior of ultradistributions with respect to regularly varying functions. If we exchange [13, Lemma 3.4] with Proposition 4.1, the same technique from [13, Sections 3 and 4] leads to two ensuing structural theorems for quasiasymptotic boundedness. 1 1 1˚ Theorem 4.3. Assume pM.1q, pM.2q , and pM.3q hold. Let f P D pRq, q P R, and let L P L r0, 8q be O-slowly varying at infinity. Let k be the smallest positive integer loc such that ´k ď q. Then, q 1˚ (4.5) fpλxq “ O pλ Lpλqq as λ Ñ 8 in D pRq holds if and only if there are continuous functions f on R such that pmq f “ f , m“k´1 for some ℓ ą 0 (for any ℓ ą 0) there exists C ą 0 such that q`m (4.6) |f pxq| ď C p1 ` |x|q Lp|x|q, m ě k ´ 1, m ℓ and additionally (only) when q “ ´k (4.7) f pxqdx “ O pLpxqq , x Ñ 8. k´1 ´x A function L is O-regularly varying at the origin if Lp1{xq is O-regularly varying at infinity. 1 1 1˚ Theorem 4.4. Assume pM.1q, pM.2q , and pM.3q . Let f P D pRq, q P R, and let L be O-slowly varying at the origin. Let k be the smallest positive integer such that ´k ď q. Then, we have that q ` 1˚ (4.8) fpεxq “ O pε Lpεqq as ε Ñ 0 in D pRq One still needs an O-version of the integration lemma [13, Lemma 3.2]; however, careful inspec- tion in the arguments given in [18, Subsection 2.10.2 and Proposition 2.17] shows that having the inequalities (4.4) is all one needs to establish the validity of such an O-version. ASYMPTOTIC BOUNDEDNESS AND MOMENT ASYMPTOTIC EXPANSION 9 holds if and only if there exist x ą 0 and continuous functions F and f on r´x , x szt0u, 0 m 0 0 m ě k, such that pkq pmq fpxq “ F ` f , on p´x , x q, 0 0 m“k for some ℓ ą 0 (for any ℓ ą 0) there exists C ą 0 such that q`m |f pxq| ď C |x| Lp|x|q, 0 ă |x| ď x , m ℓ 0 for all m ě k, and F “ 0 when q ą ´k while if q “ ´k the function F satisfies, for each a ą 0, the bounds Fpaxq ´ Fp´xq “ O pLpxqq, x Ñ 0 . We end this section with a remark that briefly indicates further properties of quasi- asymptotic boundedness. Remark 4.5. Let Z pRq be the space of ultradifferentiable functions introduced in [13, Section 5]. Similarly as in the quoted article, one can show under the assumptions 1 1 pM.1q, pM.2q and pM.3q : 1˚ (i) If (4.5) holds with an O-regularly varying function at infinity L , then f P Z pRq 1˚ and the quasiasymptotic boundedness relation (4.5) actually holds true in Z pRq. 1˚ From here one derives the following characterization of Z pRq. An ultradis- 1˚ 1˚ tribution f P D pRq belongs to Z pRq if and only if there is some q P R such q 1˚ that fpλxq “ Opλ q as λ Ñ 8 in D pRq. We leave the verification of the direct implication to the reader. 1˚ (ii) If f P Z pRq and (4.8) holds with an O-regularly varying function at the origin 1˚ L, then (4.8) is actually valid in Z pRq. (iii) From these two properties, the Banach-Steinhaus theorem, and the fact that ˚ ˚ D pRq is dense in Z pRq, one concludes that [13, Theorem 5.1 and Theorem 5.3] still hold true if one replaces pM.2q and pM.3q there by the weaker assumptions 1 1 pM.2q and pM.3q . 5. The moment asymptotic expansion This section is devoted to study the moment asymptotic expansion (1.1), which in general we interpret in the sense of the following definition. Definition 5.1. Let X be a l.c.s. of smooth functions provided with continuous actions of the dilation operators and the Dirac delta and all its partial derivatives. An element 1 1 f P X is said to satisfy the moment asymptotic expansion (MAE) in X if there are µ P C, α P N , called its moments, such that for any ϕ P X and k P N we have ˆ ˙ pαq µ ϕ p0q 1 (5.1) hfpλxq, ϕpxqi “ ` O , λ Ñ 8. |α|`d k`d α!λ λ |α|ăk Similarly as in the case of compactly supported distributions [6, 7] or analytic func- tionals [20], one can show that any compactly supported distribution satisfies the MAE 1˚ d in E pR q (we will actually state a stronger result in Proposition 6.3 below). Naturally, 10 L. NEYT AND J. VINDAS as in the distribution case, we expect the MAE to be also valid in larger ultradistri- bution spaces. In dimension 1, Estrada gave in [5, Theorem 7.1] (cf. [7]) a full char- acterization of the largest distribution space where the moment asymptotic expansion 1 1 holds; in fact, he showed that f P D pRq satisfies the MAE (in D pRq) if and only if 1 1 f P K pRq (and the MAE holds in this space), where K pRq is the dual of the so-called space of GLS symbols of pseudodifferential operators [9]. One of our goals here is to give an ultradistributional counterpart of Estrada’s result. We start by introducing an ultradistributional version of KpR q. For each q P N and M ,ℓ ℓ ą 0 we denote by K pR q the Banach space of all smooth functions ϕ for which the norm |α|´q pαq p1 ` |x|q |ϕ pxq| kϕk M ,ℓ “ sup sup K pR q |α| d d ℓ M αPN xPR is finite. From this we construct the spaces pM q d M ,ℓ d tM u d M ,ℓ d p p p p K pR q “ lim K pR q, K pR q “ lim K pR q, q q q q ÐÝ ÝÑ ℓÑ0 ℓÑ8 and finally the test function space ˚ d ˚ d K pR q “ lim K pR q. ÝÑ qPN ˚ d It should be noticed that this is space is never trivial; in fact, K pR q contains the space of polynomials. 1˚ d Our first important result in this subsection asserts that the elements of K pR q automatically satisfy the MAE. Interestingly, no restriction on the weight sequence M is needed to achieve this. 1˚ d 1˚ d Theorem 5.2. Any element f P K pR q satisfies the MAE in K pR q and its moments α d are exactly µ “ hfpxq, x i, α P N . 1˚ d ˚ d Proof. Let f P K pR q We keep λ ě 1 and fix k P N. Take any arbitrary ϕ P K pR q, where we may assume q ě k. Consider the pk ´ 1qth order Taylor polynomial of ϕ at pαq α ˚ d the origin, that is, ϕ pxq :“ ϕ p0qx {α!. Since ϕ P K pR q, k k |α|ăk pαq µ ϕ p0q hfpλxq, ϕpxqi “ ` hfpλxq, ϕpxq ´ ϕ pxqi . |α|`d α!λ |α|ăk k`d Thus, we need to show hfpλxq, ϕpxq ´ ϕ pxqi “ Op1{λ q. This bound does not pαq require any uniformity in k; therefore, we may just assume that ϕ p0q “ 0 for any |α| ă k so that our problem reduces to estimate | hfpλxq, ϕpxqi |. There exists some M ,ℓ ℓ “ ℓ ą 0 (some ℓ “ ℓ ą 0) such that ϕ P K pR q and f φ q kfk M ,ℓ |α|´q pαq d 1 p1 ` |x|q |ϕ px{λq| pK pR qq |hfpλxq, ϕpxqi| ď sup sup . d |α| |α| λ d d λ ℓ M αPN xPR If |α| ě q, we have ASYMPTOTIC BOUNDEDNESS AND MOMENT ASYMPTOTIC EXPANSION 11 ˆ ˙ |α|´q |α|´q pαq |α|´q pαq p1 ` |x|q |ϕ px{λq| 1 1 ` |x| p1 ` |x|{λq |ϕ px{λq| sup “ sup |α| |α| q |α| d λ ℓ M λ λ ` |x| ℓ M xPR α α kϕk M ,ℓ K pR q ď . We further consider |α| ă q. When |x| ě λ, obviously 1 1 ` |x| 2 λ ` |x| and we obtain |α|´q pαq kϕk M ,ℓ p1 ` |x|q |ϕ px{λq| K pR q sup ď 2 . |α| |α| k λ ℓ M λ |x|ěλ α We are left with the case |x| ď λ and |α| ă q. If k ď |α| ă q we get |α|´q pαq pαq kϕk M ,ℓ p d p1 ` |x|q |ϕ px{λq| 1 |ϕ px{λq| K pR q q´k sup ď sup ď 2 . |α| |α| k |α| k λ ℓ M λ ℓ M λ |x|ďλ α |x|ďλ α Finally, for |α| ă k, the Taylor formula yields |α|´q pαq |α|´q pβq |β´α| |ϕ pξ q| p1 ` |x|q |ϕ px{λq| p1 ` |x|q |x| x{λ sup ď sup |α| |α| |α| |α| |β´α| λ ℓ M λ ℓ M pβ ´ αq! λ |x|ďλ α |x|ďλ α αďβ |β|“k ℓ,k ď 2 kϕk M ,ℓ . k K pR q The proof is now complete. 1˚ d The next proposition describes the structure of the elements of K pR q. The proof in the Beurling case is standard, while in the Roumieu case it can be established via the dual Mittag-Leffler lemma in a similar fashion as in [10, Section 8], we therefore leave details to the reader. We point out that the converse of Proposition 5.3 holds unconditionally, that is, without having to impose any assumption on M . 1 1˚ d Proposition 5.3. Let M satisfy pM.1q and pM.2q . Let f P K pR q. Then, given any q P N one can find a multi-sequence of continuous functions f “ f P CpR q α q,α such that pαq (5.2) f “ f αPN and for some ℓ ą 0 (for any ℓ ą 0) there is C “ C ą 0 such that q,ℓ |α| |α|´q d d (5.3) |f pxq| ď C p1 ` |x|q , x P R , α P N . Notice that when pM.1q and pM.3q hold, then one has the continuous and dense in- ˚ d ˚ d ˚ d 1˚ d 1˚ d clusions D pR q ãÑ K pR q ãÑ E pR q, so that in particular K pR q Ă D pR q. Upon combining Proposition 4.1(i) with Theorem 5.2, one obtains the following complete 12 L. NEYT AND J. VINDAS 1˚ characterization of those one-dimensional ultradistributions f P D pRq satisfying the MAE: 1 1 Theorem 5.4. Suppose M satisfies pM.1q, pM.2q , and pM.3q . An ultradistribution 1˚ 1˚ 1˚ f P D pRq satisfies the MAE in D pRq if and only if f P K pRq. ´q 1˚ Proof. If f satisfies the MAE, then in particular fpλxq “ Opλ q in D pRzt0uq for each q P N. Hence, for a fixed but arbitrary q P N, using Proposition 4.1(i) and Komatsu’s first structural theorem in the case of compactly supported ultradistributions, we can 8 pmq 1˚ write f “ f in D pRq with f “ f P CpRq such that for some (for m q,m m“1 m m´q´2 each) ℓ ą 0 they fulfill bounds f pxq “ O pℓ p|x| ` 1q {M q. Clearly, this m q,ℓ m 1˚ 1˚ representation yields f P K pRq. Since q was arbitrary, we conclude that f P K pRq. For the converse, Theorem 5.2 shows that a stronger conclusion actually holds. Remark 5.5. In dimension d “ 1, this argument gives an alternative way for proving Proposition 5.3 in the non-quasianalytic case without having to resort in the dual Mittag-Leffler lemma. 6. The uniform MAE The bound in (5.1) is not uniform in general, but in the ultradistributional case it is natural to expect that some sort of uniformity could be present. For instance, we see below in Proposition 6.3 that this is the case for compactly supported ultradis- tributions. Let us introduce the following uniform variant of the MAE. Throughout this section we work with three weight sequences M , N , and A , and simultaneously p p p denote in short their Beurling or Roumieu cases by ˚, :, and 7, respectively; the as- sociated functions of M and N are M and N. The associated function [10] of N is p p p given by t N Nptq :“ sup log , t ě 0. pPN p Definition 6.1. Let A be a weight sequence and let X be a l.c.s. of smooth functions provided with with continuous actions of the dilation operators and the Dirac delta and all its partial derivatives. An element f P X satisfies the uniform moment asymptotic 1 d expansion (UMAE) in X with respect to 7 if there are µ P C, α P N , such that for any ϕ P X and each ℓ ą 0 (for some ℓ “ ℓ ą 0) the asymptotic formula ˆ ˙ pαq k µ ϕ p0q ℓ A α k (6.1) hfpλxq, ϕpxqi “ ` O , λ Ñ 8, |α|`d k`d α!λ λ |α|ăk holds uniformly for k P N. We now introduce ultradistribution spaces that are closely related to the UMAE. M ,ℓ d 8 d Given q, ℓ ą 0 we denote by K pR q the Banach space of all ϕ P C pR q for which N ,q ˇ ˇ ´Npq|x|q |α| pαq ˇ ˇ e p1 ` |x|q ϕ pxq (6.2) kϕk M ,ℓ :“ sup sup K pR q |α| N ,q ℓ M p d d αPN xPR ASYMPTOTIC BOUNDEDNESS AND MOMENT ASYMPTOTIC EXPANSION 13 is finite. We then define pM q M ,ℓ tM u M ,ℓ p d p d p d p d K pR q “ lim lim K pR q, K pR q “ lim lim K pR q, N ,q N ,q pN q p tN u p p p ÝÑ ÐÝ ÝÑ ÐÝ ` ` qÑ8 ℓÑ8 ℓÑ0 qÑ0 1˚ d and consider the dual K pR q, whose elements satisfy the UMAE as stated in the next theorem. Theorem 6.2. Suppose M and N satisfy pM.1q and pM.2q; in addition we assume p p 1˚ d that inf N ą 0. Set A “ N max pM {j!q. Then, any element f P K pR q pPN p p p jďp j 1˚ d satisfies the UMAE in K pR q w.r.t. 7. Proof. By replacing it by an equivalent sequence, we may assume that N ą 1 for each p. Fix an arbitrary 0 ă ε ď 1 in the Beurling case, while we put ε “ 1 in the Roumieu case. We will always assume λ ě H ě 1, where H is the parameter in pM.2q (for 1˚ d ˚ d both sequences). Take any f P K pR q and ϕ P K pR q. Arguing as in the proof of : : Theorem 5.2, we need to find a uniform bound for | hfpλxq, ϕpxq ´ ϕ pxqi |, where ϕ k k is the pk ´ 1qth order Taylor polynomial of ϕ at the origin. There exist q “ q ą 0 and M ,ℓ ℓ “ ℓ ą 0 (ℓ “ ℓ ą 0 and q “ q ą 0) such that ϕ P K pR q and for some C ą 0 f ϕ f Np,q ´Npq|x|q |α| α C e p1 ` |x|q |B rϕpx{λq ´ ϕ px{λqs| |hfpλxq, ϕpxq ´ ϕ pxqi| ď sup sup . d |α| λ d d ℓ M αPN xPR α We split according to the size of α P N . First suppose that |α| ă k. Set ℓ :“ maxp1, ℓq. From the Taylor expansion and [10, Proposition 3.6, p. 51] applied to the sequence N , pαq |α| pαq px{λq| p1 ` |x|q |ϕ px{λq ´ ϕ Npq|x|q |α| |α| e λ ℓ M ˆ ˙ pβq |β´α| ´Npq|x|q |α| ÿ x |ϕ pξ q| e p1 ` |x|q |x| |α| pλℓq M pβ ´ αq! λ αďβ |β|“k |β´α| ℓ M ´k k k Npq|x|{λq´Npq|x|q ď λ kϕk M ,εℓ ε p1 ` |x|q e K pR q N ,q p M pβ ´ αq! αďβ |β|“k β´α ´k k k ´Npq|x|{Hq ď λ AN kϕk M ,εℓ pdHℓ εq p1 ` |x|q e 0 p 0 K pR q N ,q p |β ´ α|! αďβ |β|“k ` ˘ k j ´k d´1 2 ´1 2 ď λ 2 A kϕk M ,εℓ 4dq H ℓ ε N max . p 0 k K pR q N ,q 0ďjďk p j! Now let |α| ě k. For |x| ě λ, one has " * p p p p k pq|x|{λq N q |x| N q |x| N 0 0 0 k k λ exprNpq|x|{λqs “ λ sup ď max sup , sup N N N pPN p pěk p 0ďpăk p ´k ď q N exprNpq|x|qs, 0 14 L. NEYT AND J. VINDAS |α| |α| |α| d where q “ minp1, qq. Then, since p1 ` |x|q {p1 ` |x|{λq ď λ for any α P N , we have ˇ ˇ ´Npq|x|q |α| pαq ˇ ˇ e p1 ` |x|q ϕ px{λq ´k k sup sup ď λ kϕk M ,εℓ pε{q q N . 0 k |α| K pR q pλℓq M N ,q |α|ěk |x|ěλ In the case |x| ď λ, we have for |α| ě k, ˇ ˇ ´Npq|x|q |α| pαq ´Npq|x|q k |α|´k pαq ˇ ˇ e p1 ` |x|q ϕ px{λq e p1 ` |x|q p1 ` |x|q |ϕ px{λq| |α| k |α|´k |α| pλℓq M λ λ ℓ M α α ´k ´1 Npqq k ď λ N e kϕk M ,εℓ p2ε{qq N , 0 d K pR q N ,q which concludes the proof. The next result describes the UMAE for compactly supported ultradistributions. The proof goes alone the same lines as that of Theorem 6.2 and we therefore leave details to the reader. 1˚ d 1˚ d Proposition 6.3. Any element f P E pR q satisfies the UMAE in E pR q w.r.t. 7, where A “ max pM {j!q. p jďp j 1˚ d A standard argument shows the ensuing structural description for K pR q. 1 1˚ d Proposition 6.4. Let M satisfy pM.1q and pM.2q . Let f P K pR q. Then, for each q ą 0, there is some ℓ “ ℓ (for each ℓ there some q ą 0) such that one can find a q ℓ pαq multi-sequence of continuous functions f “ f P CpR q for which f “ f d α α q,ℓ,α αPN and there is C “ C ą 0 such that q,ℓ |α| |α| ´Npq|x|q d d (6.3) |f pxq| ď C p1 ` |x|q e , x P R , α P N . Let us now consider the one-dimensional case. The ensuing theorem is a counterpart of Theorem 5.4 for the UMAE; notice however that a full characterization is lacking 1 ˚ d in this case. We mention that if pM.1q and pM.3q hold, one verifies that D pR q ãÑ ˚ d ˚ d K pR q ãÑ E pR q. Theorem 6.5. Suppose that N satisfies pM.1q and that pM.1q, pM.2q, and pM.3q hold 1˚ for the weight sequence M . Set A “ M N {p!. If f P D pRq satisfies the UMAE in p p p p 1˚ 1˚ D pRq with respect to :, then f P K pRq and if in addition N satisfies pM.2q and 1˚ N ą 0, the UMAE holds for f in K pRq w.r.t. 7. inf pPN p 1˚ Proof. It suffices to show that f P K pRq. In the Beurling case we take an arbitrary constant sequence r “ 1{q ą 0 and in the Roumieu case an arbitrary pr q P R. We p p ˚ d have that, whenever ϕ P D pR zt0uq, ˆ ˙ R N k´1 k´1 |hfpλxq, ϕpxqi| ď O , k`d λ N which implies, taking infimum over k, ` ` ˘˘ ´pd`1q |hfpλxq, ϕpxqi| “ O λ exp ´N pλq . p ASYMPTOTIC BOUNDEDNESS AND MOMENT ASYMPTOTIC EXPANSION 15 pmq Appliying Proposition 4.1(ii), we can write f “ f with continuous functions mPN f satisfying the bounds m´d´1 ´N p|x|q |f pxq| ď C p1 ` |x|q e , x P R, m P N, m ℓ for some ℓ ą 0 (for each ℓ ą 0), where N stands for the associated function of R N . r p p 1˚ This yields f P K pRq in both cases, as required. (In the Roumieu case we apply [3, Lemma 4.5(i), p. 417].) It has been proved by Petzsche [14, Proposition 1.1] that pM.3q implies the so-called Rudin condition, namely, there is C such that ˆ ˙ ˆ ˙ 1{j 1{p M M j p max ď C , p P N; jďp j! p! therefore, the rest follows from Theorem 6.2. References [1] N. Bingham, C. Goldie, and J. Teugels, Regular variation, Encyclopedia of Mathematics and its Applications, 27, Cambridge University Press, Cambridge, 1987. [2] I. Cioranescu, The characterization of the almost periodic ultradistributions of Beurling type, Proc. Amer. Math. Soc. 116 (1992), 127–134. [3] A. Debrouwere, H. Vernaeve, and J. Vindas, Optimal embeddings of ultradistributions into differential algebras, Monatsh. Math. 186 (2018), 407–438. [4] P. Dimovski, S. Pilipovic, B. Prangoski, and J. Vindas, Convolution of ultradistributions and ultradistribution spaces associated to translation-invariant Banach spaces, Kyoto J. Math. 56 (2016), 401–440. [5] R. Estrada, The Cesa`ro behaviour of distributions, Proc. R. Soc. Lond. A 454 (1998), 2425– [6] R. Estrada and R. Kanwal, A distributional theory of asymptotic expansions, Proc. Roy. Soc. London Ser. A 428 (1990), 399–430. [7] R. Estrada and R. Kanwal, A distributional approach to asymptotics. Theory and applica- tions, Birkh¨auser Boston, Boston, MA, 2002. [8] M. C. Gomez-Collado, Almost periodic ultradistributions of Beurling and of Roumieu type, Proc. Amer. Math. Soc. 129 (2001), 2319–2329. [9] A. Grossmann, G. Loupias, and E. M. Stein, An algebra of pseudodifferential operators and quantum mechanics in phase space, Ann. Inst. Fourier 18 (1968), 343–368. [10] H. Komatsu, Ultradistributions. I. Structure theorems and a characterization, J. Fac. Sci. Univ. Tokyo Sect. IA Math. 20 (1973), 25–105. [11] H. Komatsu, Ultradistributions, III. Vector valued ultradistributions and the theory of kernels, J. Fac. Sci. Univ. Tokyo Sect. IA Math. 29 (1982), 653–717. [12] H. Komatsu, Microlocal analysis in Gevrey classes and in complex domains, in: Microlocal anal- ysis and applications (Montecatini Terme, 1989), pp. 426–493. Lect. Not. Math. 1726, Springer- Verlag, Berlin, 1991. [13] L. Neyt and J. Vindas, Structural theorems for quasiasymptotics of ultradistributions, Asymp- tot. Anal. 114 (2019), 1–18. [14] H.-J. Petzsche, On E. Borel’s theorem, Math. Ann. 282 (1988), 299–313. ´ ´ [15] S. Pilipovic and B. Stankovic, Wiener Tauberian theorems for ultradistributions, Rend. Sem. Mat. Univ. Padova 92 (1994), 209–220. ´ ´ [16] S. Pilipovic and B. Stankovic, Quasi-asymptotics and S-asymptotics of ultradistributions, Bull. Acad. Serbe Sci. Arts Cl. Sci. Math. Natur. 20 (1995), 61–74. 16 L. NEYT AND J. VINDAS ´ ´ [17] S. Pilipovic and B. Stankovic, Properties of ultradistributions having the S-asymptotics, Bull. Acad. Serbe. Sci. 21 (1996), 47–59. ´ ´ [18] S. Pilipovic, B. Stankovic, and J. Vindas, Asymptotic behavior of generalized functions, Series on Analysis, Applications and Computation, 5, World Scientific Publishing Co. Pte. Ltd., Hackensack, NJ, 2012. [19] B. Prangoski, Pseudodifferential operators of infinite order in spaces of tempered ultradistribu- tions, J. Pseudo-Differ. Oper. Appl. 4 (2013), 495–549. [20] A. Schmidt, Asymptotic hyperfunctions, tempered hyperfunctions, and asymptotic expansions, Int. J. Math. Math. Sci. 2005, 755–788. [21] L. Schwartz, Th´eorie des distributions, Hermann, Paris, 1966. [22] V. S. Vladimirov, Yu. N. Drozhzhinov, and B. I. Zavialov, Tauberian theorems for generalized functions, Kluwer Academic Publishers, Dordrecht, 1988. [23] J. Vindas, The structure of quasiasymptotics of Schwartz distributions, Banach Center Publ. 88 (2010), 297–314. [24] Y. Yang and R. Estrada, Asymptotic expansion of thick distributions, Asymptot. Anal., 95 (2015), 1–19. [25] B. I. Zav’yalov, Self-similar asymptotic form of electromagnetic form-factors and the behavior of their Fourier transforms in the neighborhood of the light cone, Teoret. Mat. Fiz. 17 (1973), 178–188; English translation in: Theoret. and Math. Phys. 17 (1973), 1074–1081. Department of Mathematics: Analysis, Logic and Discrete Mathematics, Ghent Uni- versity, Krijgslaan 281, 9000 Gent, Belgium E-mail address: lenny.neyt@UGent.be E-mail address: jasson.vindas@UGent.be http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Mathematics arXiv (Cornell University)

Asymptotic boundedness and moment asymptotic expansion in ultradistribution spaces

Mathematics , Volume 2020 (1906) – Jun 14, 2019

Loading next page...
 
/lp/arxiv-cornell-university/asymptotic-boundedness-and-moment-asymptotic-expansion-in-acyeG1YDqz

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

ISSN
1452-8630
eISSN
ARCH-3343
DOI
10.2298/AADM191023021N
Publisher site
See Article on Publisher Site

Abstract

ASYMPTOTIC BOUNDEDNESS AND MOMENT ASYMPTOTIC EXPANSION IN ULTRADISTRIBUTION SPACES LENNY NEYT AND JASSON VINDAS Dedicated to the memory of Prof. Bogoljub Stankovi´c Abstract. We obtain structural theorems for the so-called S-asymptotic and quasi- asymptotic boundedness of ultradistributions. Using these results, we then analyze the moment asymptotic expansion (MAE), providing a full characterization of those ultradistributions satisfying this asymptotic formula in the one-dimensional case. We also introduce and study a uniform variant of the MAE. 1. Introduction In this article we study asymptotic properties of ultradistributions. Asymptotic analysis is an important subject in the theory of generalized functions and provides powerful tools for applications in areas such as mathematical physics, number theory, and differential equations. The theory of asymptotic behavior of generalized functions has been particularly useful in the study of Tauberian theorems for integral transforms. We refer to the monographs [7, 18, 22] for complete accounts on the subject and its many applications. The asymptotic behavior of a generalized function is usually analyzed via its para- metric behavior, mostly with respect to translation or dilation. The idea of looking at the translates of a distribution goes back to Schwartz [21, Chapter VII], who used it to measure the order of growth of tempered distributions at infinity. Pilipovi´c and Stankovi´c later introduced a generalization, the so called S-asymptotic behavior, and thoroughly investigated its properties for distributions, ultradistributions, and Fourier hyperfunctions. There are deep connections between S-asymptotics and Wiener Taube- rian theorems for generalized functions [15]. In Section 3 of this article we study the structure of S-asymptotically bounded ultradistributions; in fact, we discuss a coun- terpart of the main structural result from [17] (cf. [18, Theorem 1.10, p. 46]) for S-asymptotic boundedness, which we prove to hold under much weaker assumptions than those employed there. There are two very well-established approaches to asymptotics of generalized func- tions related to dilation. The first one is the quasiasymptotic behavior, which employs regularly varying functions [1] as gauges in the asymptotic comparisons. This concept 2010 Mathematics Subject Classification. 46F05; 46F10. Key words and phrases. Asymptotic behavior of generalized functions; moment asymptotic expan- sions; S-asymptotics; quasiasymptotics; ultradistribution spaces. L. Neyt gratefully acknowledges support by Ghent University, through the BOF-grant 01J11615. J. Vindas was supported by Ghent University through the BOF-grants 01J11615 and 01J04017. arXiv:1906.06232v2 [math.FA] 21 Nov 2019 2 L. NEYT AND J. VINDAS was first introduced by Zav’yalov for Schwartz distributions [25] and further devel- oped by him, Drozhzhinov, and Vladimirov in [22]. The quasiasymptotic behavior was naturally extended to the context of one-dimensional ultradistributions in [16]. Here we shall provide complete structural theorems for quasiasymptotically bounded ultra- distributions in Section 4, both at infinity and the origin, using the above-mentioned results for S-asymptotic boundedness and techniques developed by the authors in [13]. These results are the ultradistributional analogs of the structural theorems from [23] for Schwartz distributions (cf. [18, Section 2.12, p. 160]). The second important approach to asymptotic behavior related to dilation is the so- called moment asymptotic expansion (MAE), whose properties have been extensively investigated by Estrada and Kanwal [6, 7]. Some recent contributions can be found in [20, 24]. A generalized function f is said to satisfy the MAE if there is a certain multisequence tµ u d, called the moments of f, such that the following asymptotic αPN expansion holds |α| pαq p´1q µ δ pxq (1.1) fpλxq „ , λ Ñ 8. |α|`d α!λ αPN As is shown in the monograph [7], the MAE supplies a unified approach to several aspects of asymptotic analysis and its applications. Interestingly, Estrada characterized [5] the largest space of distributions where the MAE holds as the dual of the space of so-called GLS symbols [9]. We will consider the MAE for ultradistributions, providing in Section 5 a counterpart of Estrada’s full characterization in the one-dimensional case. We shall also study a uniform version of (1.1) in Section 6, which we call the UMAE. Our considerations naturally lead to introduce the ultradistribution spaces 1˚ d 1˚ d K pR q and K pR q, which are intimately connected with the MAE and UMAE. We 1˚ 2d note that in even dimension our space K pR q arises as the dual of one of the spaces of symbols of ‘infinite order’ pseudo-differential operators from [19]. 2. Preliminaries In this section we fix the notation and briefly collect some background material on ultradistributions. Given a weight sequence tM u of positive real numbers, we will p pPN often make use of one or more of the following conditions: pM.1q: M ď M M , p ě 1; p´1 p`1 1 p pM.2q : M ď AH M , p P N, for certain constants A and H; p`1 p p`q pM.2q: M ď AH M M , p, q P N, for certain constants A and H; p`q p q pM.3q : M {M ă 8; p´1 p p“1 pM.3q: M {M ď c qM {M , q ě 1, for a certain constant c . p´1 p 0 q´1 q 0 p“q The meaning of all these conditions is very well explained in [10]. For multi-indices α P N , we will simply denote M as M . |α| α d d Let Ω Ă R be open and let K Ť R be a (regular) compact subset. For any ℓ ą 0 M ,ℓ we denote as E pKq the space of all smooth functions ϕ for which pαq |ϕ pxq| kϕk :“ sup sup M ,ℓ E pKq |α| ℓ M xPK α αPN ASYMPTOTIC BOUNDEDNESS AND MOMENT ASYMPTOTIC EXPANSION 3 is finite. Then, as customary [10], we set pM q M ,ℓ tM u M ,ℓ p p p p E pΩq “ lim lim E pKq and E pΩq “ lim lim E pKq, ÐÝ ÐÝ ÐÝ ÝÑ KŤΩ ℓą0 KŤΩ ℓą0 and use E pΩq as the common notation for the Beurling and Roumieu case (when a separate treatment is needed, we will always first state assertions about the Beurling case pM q, then followed by the Roumieu case tM u in parenthesis). p p tM u We shall make use of Komatsu’s well-known projective limit description of E pΩq. Consider the directed set R “ pℓ q : ℓ Õ 8 and ℓ ą 0, @p P Z . For any pℓ q P p pPZ p p ` p R we write L “ 1 and L “ ℓ for p P Z . Then, it is shown in [11, Proposition 0 p j ` j“1 1 1 3.5, p. 675] that if the weight sequence satisfies pM.1q, pM.2q , and pM.3q , then tM u pL M q p p p (2.1) E pΩq “ lim E pΩq ÐÝ pℓ qPR as locally convex spaces. For compactly supported test functions our notation is standard, we write D pΩq “ M ,ℓ ˚ M ,ℓ E pΩq X DpΩq and D “ E pKq X D with K Ť Ω, these spaces are topologized in the canonical way [10]. (Naturally, by the Denjoy-Carleman theorem and under ˚ 1 1˚ pM.1q, the non-triviality of D pΩq is equivalent to pM.3q .) The strong dual D pΩq is 1˚ the space of ultradistributions of class ˚, while the elements of E pΩq are exactly those ultradistributions with compact support. 3. S-asymptotic boundedness Our aim in this section is to study the structure of those ultradistributions that satisfy 1˚ d (3.1) fpx ` hq “ O pωphqq, h P W, in D pR q, where W Ď R is simply an unbounded set and ω is a positive function. The S- asymptotic relation is to be interpreted in the ultradistributional sense [18], that is, it ˚ d explicitly means that for each test function ϕ P D pR q, hfpx ` hq, ϕpxqi pf ˚ ϕˇqphq (3.2) sup “ sup ă 8. ωphq ωphq hPW hPW We begin with a simple auxiliary lemma that allows us to preserve certain growth properties when regularizing functions. Given R ą 0 and a set W , we denote as W the open R-neighborhood of W , that is, the set W “ W ` Bp0, Rq. Lemma 3.1. Given R ą 0 there are absolute constants c and c such that each 0,R 1,R function g P L pW q satisfying the bound sup |gpx ` hq|{ωpxq ă 8, where loc xPW, |h|ăR W Ă R and ω is a positive function defined on W , can be written as g “ Δg ` g in 1 0 W for some functions g P CpR q that satisfy R j |g pxq| |gpx ` hq| sup ď c sup , j “ 0, 1. j,R ωpxq ωpxq xPW xPW, |h|ăR 4 L. NEYT AND J. VINDAS Proof. To show this, we make use of the fact that the fundamental solutions of the 1 d 8 d Laplacian belong to L pR q X C pR zt0uq. By cutting-off a fundamental solution in loc 1 d d the ball Bp0, Rq, this implies we can select functions χ P L pR q and χ P DpR q both 1 0 supported on Bp0, Rq, such that δ “ Δχ ` χ . Extend g off W as 0 and keep calling 1 0 R this extension by g. We obtain the claim if we set g “ g ˚ χ so that the desired j j inequalities hold with c “ |χ p´xq|dx. j,R j |x|ďR We can now obtain our first structural theorem for (3.1) in the case where the weight sequence satisfies mild hypotheses. For it, we shall need to impose the following regularity condition on the gauge function ω, ωp ¨ ` xq 8 d (3.3) sup P L pR q. loc d ωpxq xPR In the proof of the next theorem we employ a technique of Go´mez-Collado that she applied to obtain various characterizations of the space of bounded ultradistributions in [8]. Theorem 3.2. Let W Ă R be an unbounded set and let ω be a positive measurable d 1 1 function on R that satisfies (3.3). Suppose pM.1q, pM.2q , and pM.3q hold. Then, 1˚ d an ultradistribution f P D pR q satisfies (3.1) if and only if for each R ą 0 there are continuous functions tf u d defined on W such that for some ℓ ą 0 (for each ℓ ą 0) α R αPN there exists C ą 0 for which the bounds |α| (3.4) |f pxq| ď C ωpxq, x P W , α P N , α ℓ R hold and pαq (3.5) f “ f in W . αPN Proof. The sufficiency of the conditions is easily verified. For the necessity, we start by making some reductions. In view of Lemma 3.1 and the assumption pM.2q , it suffices to establish (3.5) with (3.4) for functions that are merely measurable. Next, we show that we may assume that W “ R . Let R ą 0 be arbitrary and let χ be a non-negative smooth function on R such that 0 ď χ ď 1, χ “ 1 on W while χ “ 0 outside R R R R W , and such that 2R pαq |χ pξq| sup ă 8, @ℓ ą 0. |α| d d ℓ M αPN , ξPR α ˚ d r r Then, we set f :“ χ ¨f and notice that f and f coincide on W . Take any ϕ P D pR q R R and let r ą 0 be such that supp ϕ Ă Bp0, rq. Take any h P R . If h R W then r`2R D E fpx ` hq{ω rphq, ϕpxq “ 0. Suppose now h P W , then h “ h ` h with h P W r`2R 1 2 1 ASYMPTOTIC BOUNDEDNESS AND MOMENT ASYMPTOTIC EXPANSION 5 and h P Bp0, r ` 2Rq. Then, employing (3.3) and the Banach-Steinhaus theorem, D E D E r r | fpx ` hq, ϕpxq | | fpx ` h q, ϕpx ´ h q | 1 2 ď C “ O p1q , R`2ε ω rphq ωph q ˚ d because tT ϕ : h P Bp0, r ` 2Rqu is a bounded family in D pR q. Consequently, h 2 d 1˚ d fpx ` hq “ Opωphqq, h P R , in D pR q We may therefore assume w.l.o.g. that f “ f and derive (3.4) and (3.5) on the whole d d R for measurable functions f under the hypothesis that tfpx ` hq{ωphq : h P R u is 1˚ d ˚ a bounded subset of D pR q. We now reason as in [8]. Let ψ P D be such that r´1,1s d d ψpx ´ nq “ 1 for each x P R . We have that tψfp ¨ ` nq{ωpnq : n P Z u is now nPZ 1˚ d 1 a bounded set in the space E pR q. Using pM.2q , we obtain in the Beurling case the existence of ℓ ą 0 and in the Roumieu case of pℓ q P R such that for some C ą 0 and d ˚ d all n P Z and φ P E pR q ωpnq pαq (3.6) |xfpxq, ψpx ´ nqφpx ´ nqy| ď C |φ pxq|dx, L M d α α r´1,1s αPN where in the Roumieu case we have used the projective description (2.1) and we have set L “ ℓ in the Beurling case and L “ ℓ in the Roumieu one. We further on p p p jďp 8 d consider the Banach space X of all ϕ P C pR q such that ωpxq pαq kϕk “ |ϕ pxq| dx ă 8. d L M α α αPN ˚ d Let ϕ P D pR q be arbitrary. Applying (3.6) to each φpxq “ ϕpx ` nq and using the hypothesis (3.3), we obtain, with C “ C sup ωpx ` yq{ωpxq, d d xPR , yPr´1,1s d 1 |xf, ϕy| ď |xfpxq, ψpx ´ nqϕpxqy| ď 3 C kϕk . nPZ 1 ˚ d Note that pM.1q and pM.3q ensure that D pR q is dense in X and hence we conclude f P 1 1 d d |α| pαq X . Embedding X into L pN ˆ R , dµ q via the isometry jpϕqpα, xq “ p´1q ϕ pxq, where the measure is given by dµ “ ωpxq{pL M qdαdx with dα the natural counting α α measure on N , we can apply the Hanh-Banach theorem to get the representation (3.5) d 2 with measurable functions f on R that satisfy bounds |f pxq| ď C ωpxq{pL M q. α α α α This yields already the result in the Beurling case. In the Roumieu case we finally employ [11, Lemma 3.4(ii), p. 674] to obtain the bounds (3.4) for each ℓ ą 0 and some C ą 0. In applications it is very useful to combine Theorem 3.2 with the ensuing proposition, which provides conditions under which one might essentially apply Theorem 3.2 with a function ω that is just defined on the set W . 6 L. NEYT AND J. VINDAS Proposition 3.3. Let W Ă R be a closed convex set. Any positive function ω on W satisfying ωpx ` hq (3.7) p@R ą 0q sup ă 8 ωpxq x, x`hPW |h|ďR can be extended to a positive function on R satisfying (3.3). In addition, if ω is measurable (or continuous), the extension can be chosen measurable (or continuous) as well. Proof. For any x P R we denote by xr P W the (unique in view of convexity) closest point to x in W . Then, we set ω rpxq :“ ωpxrq. Since x ÞÑ xr is continuous, ω r inherits measurability or continuity if ω has the property. We now verify (3.3) for ω r. Let R ą 0 and let C an upper bound for ωpt ` yq{ωpyq, where y, t ` y P W and t P Bp0, Rq. Let x P R and h P Bp0, Rq be arbitrary. Consider the points x, x ` h, xr and x ` h. By the obtuse angle criterion, the angles defined by the line segments rx, xr, x ` hs and Ć Ć rx`h, x ` h, xrs are at least π{2, whence |xr´x ` h| ď |x´px`hq| ď R. It then follows that ω rpx ` hq ď C ω rpxq, as required. If the weight sequence satisfies stronger assumption, one can drop any regularity assumption on ω, as stated in the next result. Theorem 3.4. Let W Ă R be an unbounded set and let ω be a positive function on 1˚ d W . Suppose that pM.1q, pM.2q, and pM.3q hold. An ultradistribution f P D pR q satisfies (3.1) if and only if for each R ą 0 there are continuous functions tf u d αPN defined on W such that for some ℓ ą 0 (for each ℓ ą 0) there exists C ą 0 such that R ℓ |α| (3.8) |f px ` hq| ď C ωpxq, x P W, |h| ă R, α P N , α ℓ and the representation (3.5) holds. Proof. The proof is similar to that of [18, Theorem 1.10, p. 46], but we provide some simplifications. The converse is easy to show, so we concentrate on showing the necessity of the conditions for the S-asymptotic boundedness relation (3.1). Let R ą 0. ˚ d We consider the linear mapping A : D pR q Ñ X, with values in the Banach space X “ tg : W Ñ C : sup |gpxq|{ωpxq ă 8u, given by Aϕ “ f ˚ ϕ. It follows from xPW the closed graph theorem that A is continuous. Consequently, we obtain from the Banach-Steinhaus theorem the existence of ℓ ą 0 in the Beurling case or pℓ q P R in M L ,1 M L ,1 p p p p the Roumieu case such that A P pD q and f ˚ ϕ P X for each ϕ P D , where Bp0,2Rq Bp0,2Rq we set L “ ℓ in the Beurling case or L “ ℓ in the Roumieu case. Since for p p p j“1 M L ,1 M L ,1 p p p p each ϕ P D the set tT ϕ : |x| ď Ru is compact in D , we conclude that for Bp0,Rq Bp0,2Rq any such a ϕ the function f ˚ ϕ is continuous on W and pf ˚ ϕqpx ` hq sup ă 8. pωphqq hPW, |x|ăR ASYMPTOTIC BOUNDEDNESS AND MOMENT ASYMPTOTIC EXPANSION 7 We now employ the parametrix method. As shown in [12, p. 199], there is an ultradif- M L ,1 p p ferential operator PpDq of class ˚ that admits a D -parametrix, namely, for which Bp0,Rq M L ,1 p p there are χ P D and ϕ P D such that δ “ PpDqϕ`χ. Setting f “ f ˚χ and ¯ ¯ Bp0,Rq Bp0,Rq g “ f ˚ϕ, we obtain the decomposition f “ PpDqg `f , which in particular establishes the representation (3.5) with functions f satisfying the bounds (3.8). 4. Quasiasymptotic boundedness Our results from the previous section can be applied to obtain structural theorems for ultradistributions being quasiasymptotically bounded in dimension 1. Let ρ be a positive function defined on an interval of the form rλ , 8q. We are interested in the relation fpλxq “ Opρpλqq as λ Ñ 8 in ultradistribution spaces. The analog of the condition (3.7) for a function ρ in this multiplicative setting is being O-regularly varying (at infinity) [1, p. 65]. The latter means (cf. [1, Theorem 2.0.4, p. 64]) that ρ is measurable and for each R ą 1 ρpλxq lim sup sup ă 8. ´1 ρpλq xÑ8 λPrR ,Rs The next proposition can be established with the aid of Theorem 3.2 and Theorem 3.4 via an exponential change of variables as in the authors’ work [13, Lemma 3.4]; we omit details. 1˚ Proposition 4.1. Let f P D pRq and ρ be a positive function. Suppose that fpλxq “ 1˚ Opρpλqq as λ Ñ 8 in D pRzt0uq. 1 1 (i) If pM.1q, pM.2q , and pM.3q hold and ρ is O-regularly varying at infinity, then there are continuous functions f and x ą 0 such that m 0 pmq (4.1) f “ f on Rzr´x , x s 0 0 m“0 and for some ℓ ą 0 (for any ℓ ą 0) there is C ą 0 such that |x| ρp|x|q, |x| ą x , m P N. (4.2) |f pxq| ď C m ℓ 0 (ii) If pM.1q, pM.2q, and pM.3q hold, for each R ą 1 one can find x and continuous functions such that f has the representation (4.1), where the f satisfy the bounds m ´1 (4.3) |f paxq| ď C |x| ρp|x|q, |x| ą x , R ă a ă R, m P N, m ℓ for some ℓ ą 0 (for any ℓ ą 0). Remark 4.2. Clearly, (4.2) implies (4.3) for an O-regularly varying function ρ. Assume 1 1 pM.1q, pM.2q , and pM.3q hold. Notice the representations (4.1) with bounds (4.3) are 1˚ also sufficient to yield fpλxq “ Opρpλqq as λ Ñ 8 in D pRzt0uq, so that the converses of both parts (i) and (ii) of Proposition 4.1 are valid. 8 L. NEYT AND J. VINDAS In the rest of the section we are interested in describing quasiasymptotic boundedness 1˚ in the full space D pRq. For it, we need to impose stronger variation assumptions on the gauge function ρ. We call a positive measurable function O-slowly varying at infintiy if for each ε ą 0 there are C , c , R ą 0 such that ε ε ε c Lpλxq (4.4) ď ď C λ , λ ě 1, x ą R . ε ε λ Lpxq In the terminology from [1] this means that the upper and lower Matuszewska indices of L are both equal to 0. Thus, a function of the form ρpλq “ λ Lpλq is an O-regularly varying function with both upper and lower Matuszewska indices equal to q P R. The authors have found in [13] complete structural theorems for the quasiasymp- totic behavior of ultradistributions with respect to regularly varying functions. If we exchange [13, Lemma 3.4] with Proposition 4.1, the same technique from [13, Sections 3 and 4] leads to two ensuing structural theorems for quasiasymptotic boundedness. 1 1 1˚ Theorem 4.3. Assume pM.1q, pM.2q , and pM.3q hold. Let f P D pRq, q P R, and let L P L r0, 8q be O-slowly varying at infinity. Let k be the smallest positive integer loc such that ´k ď q. Then, q 1˚ (4.5) fpλxq “ O pλ Lpλqq as λ Ñ 8 in D pRq holds if and only if there are continuous functions f on R such that pmq f “ f , m“k´1 for some ℓ ą 0 (for any ℓ ą 0) there exists C ą 0 such that q`m (4.6) |f pxq| ď C p1 ` |x|q Lp|x|q, m ě k ´ 1, m ℓ and additionally (only) when q “ ´k (4.7) f pxqdx “ O pLpxqq , x Ñ 8. k´1 ´x A function L is O-regularly varying at the origin if Lp1{xq is O-regularly varying at infinity. 1 1 1˚ Theorem 4.4. Assume pM.1q, pM.2q , and pM.3q . Let f P D pRq, q P R, and let L be O-slowly varying at the origin. Let k be the smallest positive integer such that ´k ď q. Then, we have that q ` 1˚ (4.8) fpεxq “ O pε Lpεqq as ε Ñ 0 in D pRq One still needs an O-version of the integration lemma [13, Lemma 3.2]; however, careful inspec- tion in the arguments given in [18, Subsection 2.10.2 and Proposition 2.17] shows that having the inequalities (4.4) is all one needs to establish the validity of such an O-version. ASYMPTOTIC BOUNDEDNESS AND MOMENT ASYMPTOTIC EXPANSION 9 holds if and only if there exist x ą 0 and continuous functions F and f on r´x , x szt0u, 0 m 0 0 m ě k, such that pkq pmq fpxq “ F ` f , on p´x , x q, 0 0 m“k for some ℓ ą 0 (for any ℓ ą 0) there exists C ą 0 such that q`m |f pxq| ď C |x| Lp|x|q, 0 ă |x| ď x , m ℓ 0 for all m ě k, and F “ 0 when q ą ´k while if q “ ´k the function F satisfies, for each a ą 0, the bounds Fpaxq ´ Fp´xq “ O pLpxqq, x Ñ 0 . We end this section with a remark that briefly indicates further properties of quasi- asymptotic boundedness. Remark 4.5. Let Z pRq be the space of ultradifferentiable functions introduced in [13, Section 5]. Similarly as in the quoted article, one can show under the assumptions 1 1 pM.1q, pM.2q and pM.3q : 1˚ (i) If (4.5) holds with an O-regularly varying function at infinity L , then f P Z pRq 1˚ and the quasiasymptotic boundedness relation (4.5) actually holds true in Z pRq. 1˚ From here one derives the following characterization of Z pRq. An ultradis- 1˚ 1˚ tribution f P D pRq belongs to Z pRq if and only if there is some q P R such q 1˚ that fpλxq “ Opλ q as λ Ñ 8 in D pRq. We leave the verification of the direct implication to the reader. 1˚ (ii) If f P Z pRq and (4.8) holds with an O-regularly varying function at the origin 1˚ L, then (4.8) is actually valid in Z pRq. (iii) From these two properties, the Banach-Steinhaus theorem, and the fact that ˚ ˚ D pRq is dense in Z pRq, one concludes that [13, Theorem 5.1 and Theorem 5.3] still hold true if one replaces pM.2q and pM.3q there by the weaker assumptions 1 1 pM.2q and pM.3q . 5. The moment asymptotic expansion This section is devoted to study the moment asymptotic expansion (1.1), which in general we interpret in the sense of the following definition. Definition 5.1. Let X be a l.c.s. of smooth functions provided with continuous actions of the dilation operators and the Dirac delta and all its partial derivatives. An element 1 1 f P X is said to satisfy the moment asymptotic expansion (MAE) in X if there are µ P C, α P N , called its moments, such that for any ϕ P X and k P N we have ˆ ˙ pαq µ ϕ p0q 1 (5.1) hfpλxq, ϕpxqi “ ` O , λ Ñ 8. |α|`d k`d α!λ λ |α|ăk Similarly as in the case of compactly supported distributions [6, 7] or analytic func- tionals [20], one can show that any compactly supported distribution satisfies the MAE 1˚ d in E pR q (we will actually state a stronger result in Proposition 6.3 below). Naturally, 10 L. NEYT AND J. VINDAS as in the distribution case, we expect the MAE to be also valid in larger ultradistri- bution spaces. In dimension 1, Estrada gave in [5, Theorem 7.1] (cf. [7]) a full char- acterization of the largest distribution space where the moment asymptotic expansion 1 1 holds; in fact, he showed that f P D pRq satisfies the MAE (in D pRq) if and only if 1 1 f P K pRq (and the MAE holds in this space), where K pRq is the dual of the so-called space of GLS symbols of pseudodifferential operators [9]. One of our goals here is to give an ultradistributional counterpart of Estrada’s result. We start by introducing an ultradistributional version of KpR q. For each q P N and M ,ℓ ℓ ą 0 we denote by K pR q the Banach space of all smooth functions ϕ for which the norm |α|´q pαq p1 ` |x|q |ϕ pxq| kϕk M ,ℓ “ sup sup K pR q |α| d d ℓ M αPN xPR is finite. From this we construct the spaces pM q d M ,ℓ d tM u d M ,ℓ d p p p p K pR q “ lim K pR q, K pR q “ lim K pR q, q q q q ÐÝ ÝÑ ℓÑ0 ℓÑ8 and finally the test function space ˚ d ˚ d K pR q “ lim K pR q. ÝÑ qPN ˚ d It should be noticed that this is space is never trivial; in fact, K pR q contains the space of polynomials. 1˚ d Our first important result in this subsection asserts that the elements of K pR q automatically satisfy the MAE. Interestingly, no restriction on the weight sequence M is needed to achieve this. 1˚ d 1˚ d Theorem 5.2. Any element f P K pR q satisfies the MAE in K pR q and its moments α d are exactly µ “ hfpxq, x i, α P N . 1˚ d ˚ d Proof. Let f P K pR q We keep λ ě 1 and fix k P N. Take any arbitrary ϕ P K pR q, where we may assume q ě k. Consider the pk ´ 1qth order Taylor polynomial of ϕ at pαq α ˚ d the origin, that is, ϕ pxq :“ ϕ p0qx {α!. Since ϕ P K pR q, k k |α|ăk pαq µ ϕ p0q hfpλxq, ϕpxqi “ ` hfpλxq, ϕpxq ´ ϕ pxqi . |α|`d α!λ |α|ăk k`d Thus, we need to show hfpλxq, ϕpxq ´ ϕ pxqi “ Op1{λ q. This bound does not pαq require any uniformity in k; therefore, we may just assume that ϕ p0q “ 0 for any |α| ă k so that our problem reduces to estimate | hfpλxq, ϕpxqi |. There exists some M ,ℓ ℓ “ ℓ ą 0 (some ℓ “ ℓ ą 0) such that ϕ P K pR q and f φ q kfk M ,ℓ |α|´q pαq d 1 p1 ` |x|q |ϕ px{λq| pK pR qq |hfpλxq, ϕpxqi| ď sup sup . d |α| |α| λ d d λ ℓ M αPN xPR If |α| ě q, we have ASYMPTOTIC BOUNDEDNESS AND MOMENT ASYMPTOTIC EXPANSION 11 ˆ ˙ |α|´q |α|´q pαq |α|´q pαq p1 ` |x|q |ϕ px{λq| 1 1 ` |x| p1 ` |x|{λq |ϕ px{λq| sup “ sup |α| |α| q |α| d λ ℓ M λ λ ` |x| ℓ M xPR α α kϕk M ,ℓ K pR q ď . We further consider |α| ă q. When |x| ě λ, obviously 1 1 ` |x| 2 λ ` |x| and we obtain |α|´q pαq kϕk M ,ℓ p1 ` |x|q |ϕ px{λq| K pR q sup ď 2 . |α| |α| k λ ℓ M λ |x|ěλ α We are left with the case |x| ď λ and |α| ă q. If k ď |α| ă q we get |α|´q pαq pαq kϕk M ,ℓ p d p1 ` |x|q |ϕ px{λq| 1 |ϕ px{λq| K pR q q´k sup ď sup ď 2 . |α| |α| k |α| k λ ℓ M λ ℓ M λ |x|ďλ α |x|ďλ α Finally, for |α| ă k, the Taylor formula yields |α|´q pαq |α|´q pβq |β´α| |ϕ pξ q| p1 ` |x|q |ϕ px{λq| p1 ` |x|q |x| x{λ sup ď sup |α| |α| |α| |α| |β´α| λ ℓ M λ ℓ M pβ ´ αq! λ |x|ďλ α |x|ďλ α αďβ |β|“k ℓ,k ď 2 kϕk M ,ℓ . k K pR q The proof is now complete. 1˚ d The next proposition describes the structure of the elements of K pR q. The proof in the Beurling case is standard, while in the Roumieu case it can be established via the dual Mittag-Leffler lemma in a similar fashion as in [10, Section 8], we therefore leave details to the reader. We point out that the converse of Proposition 5.3 holds unconditionally, that is, without having to impose any assumption on M . 1 1˚ d Proposition 5.3. Let M satisfy pM.1q and pM.2q . Let f P K pR q. Then, given any q P N one can find a multi-sequence of continuous functions f “ f P CpR q α q,α such that pαq (5.2) f “ f αPN and for some ℓ ą 0 (for any ℓ ą 0) there is C “ C ą 0 such that q,ℓ |α| |α|´q d d (5.3) |f pxq| ď C p1 ` |x|q , x P R , α P N . Notice that when pM.1q and pM.3q hold, then one has the continuous and dense in- ˚ d ˚ d ˚ d 1˚ d 1˚ d clusions D pR q ãÑ K pR q ãÑ E pR q, so that in particular K pR q Ă D pR q. Upon combining Proposition 4.1(i) with Theorem 5.2, one obtains the following complete 12 L. NEYT AND J. VINDAS 1˚ characterization of those one-dimensional ultradistributions f P D pRq satisfying the MAE: 1 1 Theorem 5.4. Suppose M satisfies pM.1q, pM.2q , and pM.3q . An ultradistribution 1˚ 1˚ 1˚ f P D pRq satisfies the MAE in D pRq if and only if f P K pRq. ´q 1˚ Proof. If f satisfies the MAE, then in particular fpλxq “ Opλ q in D pRzt0uq for each q P N. Hence, for a fixed but arbitrary q P N, using Proposition 4.1(i) and Komatsu’s first structural theorem in the case of compactly supported ultradistributions, we can 8 pmq 1˚ write f “ f in D pRq with f “ f P CpRq such that for some (for m q,m m“1 m m´q´2 each) ℓ ą 0 they fulfill bounds f pxq “ O pℓ p|x| ` 1q {M q. Clearly, this m q,ℓ m 1˚ 1˚ representation yields f P K pRq. Since q was arbitrary, we conclude that f P K pRq. For the converse, Theorem 5.2 shows that a stronger conclusion actually holds. Remark 5.5. In dimension d “ 1, this argument gives an alternative way for proving Proposition 5.3 in the non-quasianalytic case without having to resort in the dual Mittag-Leffler lemma. 6. The uniform MAE The bound in (5.1) is not uniform in general, but in the ultradistributional case it is natural to expect that some sort of uniformity could be present. For instance, we see below in Proposition 6.3 that this is the case for compactly supported ultradis- tributions. Let us introduce the following uniform variant of the MAE. Throughout this section we work with three weight sequences M , N , and A , and simultaneously p p p denote in short their Beurling or Roumieu cases by ˚, :, and 7, respectively; the as- sociated functions of M and N are M and N. The associated function [10] of N is p p p given by t N Nptq :“ sup log , t ě 0. pPN p Definition 6.1. Let A be a weight sequence and let X be a l.c.s. of smooth functions provided with with continuous actions of the dilation operators and the Dirac delta and all its partial derivatives. An element f P X satisfies the uniform moment asymptotic 1 d expansion (UMAE) in X with respect to 7 if there are µ P C, α P N , such that for any ϕ P X and each ℓ ą 0 (for some ℓ “ ℓ ą 0) the asymptotic formula ˆ ˙ pαq k µ ϕ p0q ℓ A α k (6.1) hfpλxq, ϕpxqi “ ` O , λ Ñ 8, |α|`d k`d α!λ λ |α|ăk holds uniformly for k P N. We now introduce ultradistribution spaces that are closely related to the UMAE. M ,ℓ d 8 d Given q, ℓ ą 0 we denote by K pR q the Banach space of all ϕ P C pR q for which N ,q ˇ ˇ ´Npq|x|q |α| pαq ˇ ˇ e p1 ` |x|q ϕ pxq (6.2) kϕk M ,ℓ :“ sup sup K pR q |α| N ,q ℓ M p d d αPN xPR ASYMPTOTIC BOUNDEDNESS AND MOMENT ASYMPTOTIC EXPANSION 13 is finite. We then define pM q M ,ℓ tM u M ,ℓ p d p d p d p d K pR q “ lim lim K pR q, K pR q “ lim lim K pR q, N ,q N ,q pN q p tN u p p p ÝÑ ÐÝ ÝÑ ÐÝ ` ` qÑ8 ℓÑ8 ℓÑ0 qÑ0 1˚ d and consider the dual K pR q, whose elements satisfy the UMAE as stated in the next theorem. Theorem 6.2. Suppose M and N satisfy pM.1q and pM.2q; in addition we assume p p 1˚ d that inf N ą 0. Set A “ N max pM {j!q. Then, any element f P K pR q pPN p p p jďp j 1˚ d satisfies the UMAE in K pR q w.r.t. 7. Proof. By replacing it by an equivalent sequence, we may assume that N ą 1 for each p. Fix an arbitrary 0 ă ε ď 1 in the Beurling case, while we put ε “ 1 in the Roumieu case. We will always assume λ ě H ě 1, where H is the parameter in pM.2q (for 1˚ d ˚ d both sequences). Take any f P K pR q and ϕ P K pR q. Arguing as in the proof of : : Theorem 5.2, we need to find a uniform bound for | hfpλxq, ϕpxq ´ ϕ pxqi |, where ϕ k k is the pk ´ 1qth order Taylor polynomial of ϕ at the origin. There exist q “ q ą 0 and M ,ℓ ℓ “ ℓ ą 0 (ℓ “ ℓ ą 0 and q “ q ą 0) such that ϕ P K pR q and for some C ą 0 f ϕ f Np,q ´Npq|x|q |α| α C e p1 ` |x|q |B rϕpx{λq ´ ϕ px{λqs| |hfpλxq, ϕpxq ´ ϕ pxqi| ď sup sup . d |α| λ d d ℓ M αPN xPR α We split according to the size of α P N . First suppose that |α| ă k. Set ℓ :“ maxp1, ℓq. From the Taylor expansion and [10, Proposition 3.6, p. 51] applied to the sequence N , pαq |α| pαq px{λq| p1 ` |x|q |ϕ px{λq ´ ϕ Npq|x|q |α| |α| e λ ℓ M ˆ ˙ pβq |β´α| ´Npq|x|q |α| ÿ x |ϕ pξ q| e p1 ` |x|q |x| |α| pλℓq M pβ ´ αq! λ αďβ |β|“k |β´α| ℓ M ´k k k Npq|x|{λq´Npq|x|q ď λ kϕk M ,εℓ ε p1 ` |x|q e K pR q N ,q p M pβ ´ αq! αďβ |β|“k β´α ´k k k ´Npq|x|{Hq ď λ AN kϕk M ,εℓ pdHℓ εq p1 ` |x|q e 0 p 0 K pR q N ,q p |β ´ α|! αďβ |β|“k ` ˘ k j ´k d´1 2 ´1 2 ď λ 2 A kϕk M ,εℓ 4dq H ℓ ε N max . p 0 k K pR q N ,q 0ďjďk p j! Now let |α| ě k. For |x| ě λ, one has " * p p p p k pq|x|{λq N q |x| N q |x| N 0 0 0 k k λ exprNpq|x|{λqs “ λ sup ď max sup , sup N N N pPN p pěk p 0ďpăk p ´k ď q N exprNpq|x|qs, 0 14 L. NEYT AND J. VINDAS |α| |α| |α| d where q “ minp1, qq. Then, since p1 ` |x|q {p1 ` |x|{λq ď λ for any α P N , we have ˇ ˇ ´Npq|x|q |α| pαq ˇ ˇ e p1 ` |x|q ϕ px{λq ´k k sup sup ď λ kϕk M ,εℓ pε{q q N . 0 k |α| K pR q pλℓq M N ,q |α|ěk |x|ěλ In the case |x| ď λ, we have for |α| ě k, ˇ ˇ ´Npq|x|q |α| pαq ´Npq|x|q k |α|´k pαq ˇ ˇ e p1 ` |x|q ϕ px{λq e p1 ` |x|q p1 ` |x|q |ϕ px{λq| |α| k |α|´k |α| pλℓq M λ λ ℓ M α α ´k ´1 Npqq k ď λ N e kϕk M ,εℓ p2ε{qq N , 0 d K pR q N ,q which concludes the proof. The next result describes the UMAE for compactly supported ultradistributions. The proof goes alone the same lines as that of Theorem 6.2 and we therefore leave details to the reader. 1˚ d 1˚ d Proposition 6.3. Any element f P E pR q satisfies the UMAE in E pR q w.r.t. 7, where A “ max pM {j!q. p jďp j 1˚ d A standard argument shows the ensuing structural description for K pR q. 1 1˚ d Proposition 6.4. Let M satisfy pM.1q and pM.2q . Let f P K pR q. Then, for each q ą 0, there is some ℓ “ ℓ (for each ℓ there some q ą 0) such that one can find a q ℓ pαq multi-sequence of continuous functions f “ f P CpR q for which f “ f d α α q,ℓ,α αPN and there is C “ C ą 0 such that q,ℓ |α| |α| ´Npq|x|q d d (6.3) |f pxq| ď C p1 ` |x|q e , x P R , α P N . Let us now consider the one-dimensional case. The ensuing theorem is a counterpart of Theorem 5.4 for the UMAE; notice however that a full characterization is lacking 1 ˚ d in this case. We mention that if pM.1q and pM.3q hold, one verifies that D pR q ãÑ ˚ d ˚ d K pR q ãÑ E pR q. Theorem 6.5. Suppose that N satisfies pM.1q and that pM.1q, pM.2q, and pM.3q hold 1˚ for the weight sequence M . Set A “ M N {p!. If f P D pRq satisfies the UMAE in p p p p 1˚ 1˚ D pRq with respect to :, then f P K pRq and if in addition N satisfies pM.2q and 1˚ N ą 0, the UMAE holds for f in K pRq w.r.t. 7. inf pPN p 1˚ Proof. It suffices to show that f P K pRq. In the Beurling case we take an arbitrary constant sequence r “ 1{q ą 0 and in the Roumieu case an arbitrary pr q P R. We p p ˚ d have that, whenever ϕ P D pR zt0uq, ˆ ˙ R N k´1 k´1 |hfpλxq, ϕpxqi| ď O , k`d λ N which implies, taking infimum over k, ` ` ˘˘ ´pd`1q |hfpλxq, ϕpxqi| “ O λ exp ´N pλq . p ASYMPTOTIC BOUNDEDNESS AND MOMENT ASYMPTOTIC EXPANSION 15 pmq Appliying Proposition 4.1(ii), we can write f “ f with continuous functions mPN f satisfying the bounds m´d´1 ´N p|x|q |f pxq| ď C p1 ` |x|q e , x P R, m P N, m ℓ for some ℓ ą 0 (for each ℓ ą 0), where N stands for the associated function of R N . r p p 1˚ This yields f P K pRq in both cases, as required. (In the Roumieu case we apply [3, Lemma 4.5(i), p. 417].) It has been proved by Petzsche [14, Proposition 1.1] that pM.3q implies the so-called Rudin condition, namely, there is C such that ˆ ˙ ˆ ˙ 1{j 1{p M M j p max ď C , p P N; jďp j! p! therefore, the rest follows from Theorem 6.2. References [1] N. Bingham, C. Goldie, and J. Teugels, Regular variation, Encyclopedia of Mathematics and its Applications, 27, Cambridge University Press, Cambridge, 1987. [2] I. Cioranescu, The characterization of the almost periodic ultradistributions of Beurling type, Proc. Amer. Math. Soc. 116 (1992), 127–134. [3] A. Debrouwere, H. Vernaeve, and J. Vindas, Optimal embeddings of ultradistributions into differential algebras, Monatsh. Math. 186 (2018), 407–438. [4] P. Dimovski, S. Pilipovic, B. Prangoski, and J. Vindas, Convolution of ultradistributions and ultradistribution spaces associated to translation-invariant Banach spaces, Kyoto J. Math. 56 (2016), 401–440. [5] R. Estrada, The Cesa`ro behaviour of distributions, Proc. R. Soc. Lond. A 454 (1998), 2425– [6] R. Estrada and R. Kanwal, A distributional theory of asymptotic expansions, Proc. Roy. Soc. London Ser. A 428 (1990), 399–430. [7] R. Estrada and R. Kanwal, A distributional approach to asymptotics. Theory and applica- tions, Birkh¨auser Boston, Boston, MA, 2002. [8] M. C. Gomez-Collado, Almost periodic ultradistributions of Beurling and of Roumieu type, Proc. Amer. Math. Soc. 129 (2001), 2319–2329. [9] A. Grossmann, G. Loupias, and E. M. Stein, An algebra of pseudodifferential operators and quantum mechanics in phase space, Ann. Inst. Fourier 18 (1968), 343–368. [10] H. Komatsu, Ultradistributions. I. Structure theorems and a characterization, J. Fac. Sci. Univ. Tokyo Sect. IA Math. 20 (1973), 25–105. [11] H. Komatsu, Ultradistributions, III. Vector valued ultradistributions and the theory of kernels, J. Fac. Sci. Univ. Tokyo Sect. IA Math. 29 (1982), 653–717. [12] H. Komatsu, Microlocal analysis in Gevrey classes and in complex domains, in: Microlocal anal- ysis and applications (Montecatini Terme, 1989), pp. 426–493. Lect. Not. Math. 1726, Springer- Verlag, Berlin, 1991. [13] L. Neyt and J. Vindas, Structural theorems for quasiasymptotics of ultradistributions, Asymp- tot. Anal. 114 (2019), 1–18. [14] H.-J. Petzsche, On E. Borel’s theorem, Math. Ann. 282 (1988), 299–313. ´ ´ [15] S. Pilipovic and B. Stankovic, Wiener Tauberian theorems for ultradistributions, Rend. Sem. Mat. Univ. Padova 92 (1994), 209–220. ´ ´ [16] S. Pilipovic and B. Stankovic, Quasi-asymptotics and S-asymptotics of ultradistributions, Bull. Acad. Serbe Sci. Arts Cl. Sci. Math. Natur. 20 (1995), 61–74. 16 L. NEYT AND J. VINDAS ´ ´ [17] S. Pilipovic and B. Stankovic, Properties of ultradistributions having the S-asymptotics, Bull. Acad. Serbe. Sci. 21 (1996), 47–59. ´ ´ [18] S. Pilipovic, B. Stankovic, and J. Vindas, Asymptotic behavior of generalized functions, Series on Analysis, Applications and Computation, 5, World Scientific Publishing Co. Pte. Ltd., Hackensack, NJ, 2012. [19] B. Prangoski, Pseudodifferential operators of infinite order in spaces of tempered ultradistribu- tions, J. Pseudo-Differ. Oper. Appl. 4 (2013), 495–549. [20] A. Schmidt, Asymptotic hyperfunctions, tempered hyperfunctions, and asymptotic expansions, Int. J. Math. Math. Sci. 2005, 755–788. [21] L. Schwartz, Th´eorie des distributions, Hermann, Paris, 1966. [22] V. S. Vladimirov, Yu. N. Drozhzhinov, and B. I. Zavialov, Tauberian theorems for generalized functions, Kluwer Academic Publishers, Dordrecht, 1988. [23] J. Vindas, The structure of quasiasymptotics of Schwartz distributions, Banach Center Publ. 88 (2010), 297–314. [24] Y. Yang and R. Estrada, Asymptotic expansion of thick distributions, Asymptot. Anal., 95 (2015), 1–19. [25] B. I. Zav’yalov, Self-similar asymptotic form of electromagnetic form-factors and the behavior of their Fourier transforms in the neighborhood of the light cone, Teoret. Mat. Fiz. 17 (1973), 178–188; English translation in: Theoret. and Math. Phys. 17 (1973), 1074–1081. Department of Mathematics: Analysis, Logic and Discrete Mathematics, Ghent Uni- versity, Krijgslaan 281, 9000 Gent, Belgium E-mail address: lenny.neyt@UGent.be E-mail address: jasson.vindas@UGent.be

Journal

MathematicsarXiv (Cornell University)

Published: Jun 14, 2019

There are no references for this article.