![Arxiv:1905.13477V2 [Math.CA] 9 Jan 2020 Aiyblnsto Belongs Sarily a Ata Ovre Aeyif Zyg Namely [12], Converse, in Stein Partial by a Shown Has As Moreover, Integrable](http://data.docslib.org/img/824af28c1d98713ac7d4051289a44f89-1.webp)
arXiv:1905.13477v2 [math.CA] 9 Jan 2020 aiyblnsto belongs sarily a ata ovre aeyif Zyg namely [12], converse, in Stein partial by a shown has as Moreover, integrable. are transforms ad space Hardy hr,a sa,log usual, as where, set htteproi ibr rnfr maps transform Hilbert periodic the that asserts htthis that f function integrable locally a for defined analysis, rm28i hpe I f[6.Ti mle that implies This [16]. of VII Chapter in 2.8 orem oefiieball finite some ad-iteodmxmlfnto,i h olwn es:if sense: following the in function, maximal Hardy-Littlewood se the isfies | one from bounded where L ntne 1]fra ndphdiscussion). in-depth an for [14] instance, uhrwsprilyspotdb h pns oenetgr Government Wallen Spanish Alice the and by Knut supported the partially by was financed author 2017.0425, KAW no. grant A o log log 7→ | nte lsia eutta novstespace the involves that result classical Another h ad-iteodmxmlfnto safnaetlojc nh in object fundamental a is function maximal Hardy-Littlewood The 2010 Date However, h rtato a atal upre yte“alnegM “Wallenberg the by supported partially was author first The phrases. and words Key eoe h eegemaueof measure Lebesgue the denotes M B aur 0 2020. 10, January : ahmtc ujc Classification. Subject Mathematics DSESBKS AVDRRODR SALVADOR BAKAS, ODYSSEAS ( o ucin eogn oteHryspace Hardy the to belonging functions for rizsae.Ti ilssih xesoso eut pre results of extensions others. slight and Iwaniec-Verde, yields Bonami-Madan, This con the in spaces. applications Orlicz further present We Ky. and Grellier, Abstract. L ( f Z L ,r x, L sbuddon bounded is ) B ESOSO TI’ N YMN’ THEOREMS ZYGMUND’S AND STEIN’S OF VERSIONS log log M M eoe h pnbl in ball open the denotes ) H L ( L ( L 1 f f M B ( 1 condition ) slclyitgal provided integrable locally is ) odto sbt ucetadncsayfritgaiiyo the of integrability for necessary and sufficient both is condition T L x< dx ( ( = eoti esoso oecasclrslso ymn n S and Zygmund of results classical some of versions obtain We ossigo nerbefntoso h ou hs Hilbert whose torus the on functions integrable of consisting ) R f + O H AD SPACE HARDY THE FOR log d )( B | oweak- to ) x x ( L | ∞ r =sup := ) aia ucin elHrysae,Olc spaces. Orlicz spaces, Hardy Real function, Maximal max = ( frdu 0 radius of ) T Z .Teeoe nve fteaoeetoe eut of results aforementioned the of view in Therefore, ). L R f n nyif, only and if, r> p d ( | R 0 f { 1. d log | L ( B o 1 for ) x f 1 ) Introduction ( u osntmap not does but , | | ,r x, ∈ 1 x log | < r < A H , 22 piay;4B5 63 (secondary). 46E30 42B35, (primary); 42B25 ) 0 + | ⊆ 1 } 1 p < R ( Z | na16 ae,EM ti 1]proved [12] Stein E.M. paper, 1969 a In . f T d B R ( ´ and ) IGUEZ-L ( ∞ x d etrdat centered Z x,r ∞ ≤ ti ai atta h mapping the that fact basic a is It . ) H B then , | x< dx ) log | f | f L f f L ( f ( scmatyspotdadsat- and supported compactly is h aia prtri also is operator maximal The . R ( L log : x log y PZ N LNSOLA ALAN AND OPEZ, d ´ snnngtv,then non-negative, is ) R log ) ∞ nrdcdb Bonami, by introduced ) n MTM2016-75196-P. ant | | H d L log yx dy L , ( → log L ( T L x T egFudto.Tesecond The Foundation. berg + iul bandby obtained viously eto oegeneral more of text to ) sdet ymn,and Zygmund, to due is ( 1 scnandi h real the in contained is ) C teaisPorm2018”, Program athematics | ihradius with R ( f R ∈ d ysetting by ( d ) x L oisl se for (see, itself to ) R ) 1 | f d x< dx ( , T udstheorem mund’s sspotdin supported is ;seeg The- e.g. see ); ∞ tein > r . f armonic neces- and 0 2 BAKAS,RODR´IGUEZ-LOPEZ,´ AND SOLA
Zygmund and Stein, the Hardy space H1(T) is, in terms of magnitude, associated with the Orlicz space L log L(T). In this note, we obtain versions of these results for the Musielak-Orlicz Hardy space Hlog(Rd) that was recently introduced by A. Bonami, S. Grellier, and L.D. Ky in [3] and further studied by Ky in [9]. See also [5] and [15]. To do this, we identify the correct analog of L log L in this context, which turns out to be L log log L: given a measurable subset B of Rd, L log log L(B) denotes the class of all locally integrable functions f with supp(f) ⊆ B and
|f(x)| log+ log+ |f(x)|dx < ∞. ZB In order to formally state our results, we now give the definition of the space Hlog(Rd). Let Ψ: Rd × [0, ∞) → [0, ∞) denote the function given by t Ψ(x, t) := , (x, t) ∈ Rd × [0, ∞). log(e + t) + log(e + |x|)
d If B is a subset of R , one defines LΨ(B) to be the space of all locally integrable functions f on B satisfying
Ψ(x, |f(x)|)dx < ∞. ZB We shall also fix a non-negative function φ ∈ C∞(Rd), which is supported in the d unit ball of R and has Rd φ(y)dy = 1 and φ(x) = c0 for all |x| ≤ 1/2, where c0 is a constant. Given an ǫ> 0, we use the standard notation φ (x) := ǫ−dφ(ǫ−1x), R ǫ x ∈ Rd. Definition (Hlog, see [3, 15]). If φ is as above, consider the maximal function
d Mφ(f)(x) := sup |(f ∗ φǫ)(x)|, x ∈ R . ǫ>0 The Hardy space Hlog(Rd) is defined to be the space of tempered distributions f d d on R such that Mφ(f) ∈ LΨ(R ), that is, Mφ(f) satisfies
Ψ(x, Mφ(f)(x))dx < ∞. Rd Z The motivation for defining the space Hlog comes from the study of products of functions in the real Hardy space H1(Rd) and functions in BMO(Rd), the class of functions of bounded mean oscillation. Following earlier work by Bonami, T. Iwaniec, P. Jones, and M. Zinsmeister in [2], it was shown by Bonami, Grellier, and Ky [3] that the product fg, in the sense of distributions, of a function f ∈ H1(Rd) and a function g ∈ BMO(Rd) can be represented as a sum of a continuous bilinear mapping into L1(Rd) and a continuous bilinear operator into Hlog(Rd). Here is our version of Stein’s lemma for LΨ. Theorem 1. Let f be a measurable function supported in a closed ball B (Rd. Then M(f) ∈ LΨ(B) if, and only if, f ∈ L log log L(B). Our proof in fact leads to a more general version of Theorem 1. We discuss this, and give a proof of Theorem 1 in Section 2. Next is the analog of Zygmund’s result for Hlog(Rd). Theorem 2. Let B denote the closed unit ball in Rd.
If f is a measurable function satisfying f ∈ L log log L(B) and B f(y)dy = 0, then f ∈ Hlog(Rd). R d STEIN AND ZYGMUND THEOREMS FOR Hlog (R ) 3
We remark that the mean-zero condition in the hypothesis is in fact necessary in order to place a compactly supported function in Hlog. The proof of Theorem 2 is presented in Section 3. In Section 4, we discuss further extensions to the periodic setting.
Remark 3. After posting a first version of this note, the authors were informed that our main results can be derived from results previously obtained in the setting of Orlicz spaces; see for instance [4, 8]. We are grateful for having been directed to the appropriate sources. In this note, we give a self-contained account, including a discussion of sharpness, and indicate some minor modifications that need to be made to obtain results in the Musielak-Orlicz setting.
2. Proof of the Stein-type Theorem for LΨ and further extensions We begin with an elementary observation that will be implicitly used several times in the sequel: if Φ : [0, ∞) → [0, ∞) is an increasing function, then for every positive constant α0 one has
Φ(|g(x)|)dx ≤ Φ(α0)|B| + Φ(|g(x)|)dx ZB Z{|g|>α0} for each measurable set B in Rd with finite measure. We now turn to the proof of our first theorem.
Proof of Theorem 1. Assume first that f ∈ L log log L(B). The main observation is that locally the space LΨ essentially coincides with the Orlicz space defined in −1 terms of the function Ψ0(t) := t · [log(e + t)] , t ≥ 0 and so, one can employ the arguments of Stein [12]. In view of this observation, we remark that the fact that f ∈ L log log L(B) implies M(f) ∈ LΨ0 (B) is well-known; see for instance [4, p.242], [8, Sections 4 and 7]. We shall also include the proof of this implication here for the convenience of the reader. To be more precise, we note that for x ∈ B one has
(2.1) log(e + M(f)(x)) ≤ log ((e + |x|)(e + M(f)(x))) ≤ c log(e + M(f)(x)), for a constant c that only depends on B. Next, an integration by parts yields y 1 y y 1 dα = − e + dα, log α log y log2 α Ze Ze so that y y 1 ≤ e + dα, for y > e. log y log α Ze Together, these two observations imply that
M(f)(x) 1 Ψ(x, M(f)(x))dx . 1+ dα dx B log α ZB ZB∩{M(f)>e} Ze ! ∞ 1 =1+ · |{x ∈ B : M(f)(x) > α}|dα. log α Ze To estimate the last integral, note that there exists an absolute constant Cd > 0 such that C (2.2) |{x ∈ Rd : M(f)(x) > α}| ≤ d |f(x)|dx α Z{|f|>α/2} 4 BAKAS,RODR´IGUEZ-LOPEZ,´ AND SOLA for all α > 0; see e.g. [12, (5)] or Section 5.2 (a) in Chapter I in [13]. We thus deduce from (2.2) that
2|f(x)| 1 Ψ(x, M(f)(x))dx . 1+ |f(x)| · dα dx B α log α ZB ZB Ze ! . 1+ |f(x)| log+ log+ |f(x)|dx, ZB which implies that M(f) ∈ LΨ(B). To prove the reverse implication, assume that for some f supported in B with 1 f ∈ L (B) we have M(f) ∈ LΨ(B). Our task is to show that f ∈ L log log L(B). In order to accomplish this, we shall make use of the fact that there exists a ρ> 2, depending only on kfkL1(B) and B, such that we also have M(f) ∈ LΨ(ρB) and moreover, for every α ≥ ee,
c (2.3) |{x ∈ ρB : M(f)(x) >c · α}| ≥ 2 |f(x)|dx, 1 α ZB∩{|f|>α} where c1, c2 are positive constants that can be taken to be independent of f and α. Indeed, arguing as in the proof of [12, Lemma 1], note that for every r > 2 one has
1 d (2.4) M(f)(x) . ·kfk 1 for all x ∈ R \ rB. (r − 1)d|B| L (B)
Hence, if we choose ρ > 2 to be large enough, then M(f)(x) < ee ≤ α for all x ∈ Rd \ ρB and so, (2.3) follows from [12, Inequality (6)]. Furthermore, one can check that M(f) ∈ LΨ(ρB). Indeed, if we write B = B(x0, r0) then, as in [12], it follows from the definition of M and the fact that supp(f) ⊆ B that there exists a constant c0 > 0, depending only on the dimension, such that for every x ∈ 2B \ B one has
x − x (2.5) M(f)(x) ≤ c · M(f) x + r2 · 0 0 0 0 |x − x |2 0 and so, M(f) ∈ LΨ(2B). To show that (2.5) implies that M(f) ∈ LΨ(B), observe first that the function Ψ0(s) = s/ log(e + s) is increasing on [0, +∞), and for all t ≥ 1 and all s> 0,
log(e + s) log(e + s) log(e + s) 1 1 ≥ = ≥ ≥ , log(e + ts) log(e/t + s) + log t log(e + s) + log t 1 + log t so Ψ0 satisfies
−1 (2.6) t(1 + log t) Ψ0(s) ≤ Ψ0(st) ≤ tΨ0(s), which implies that for all c> 0 and all s> 0
Ψ0(cs) ∼c Ψ0(s). d STEIN AND ZYGMUND THEOREMS FOR Hlog (R ) 5
Observe that a change to polar coordinates, followed by another a change of variables and elementary estimates yield
2r0 d−1 2 Ψ0(Mf(x))dx . s Ψ0(Mf(x0 + r0θ/s))dσ(θ)ds d−1 Z2B\B Zr0 ZS 1 −1 −1−d ∼ r0 t Ψ0(Mf(x0 + r0tθ))dσ(θ)dt 1 Sd−1 Z 2 Z 1 d−1 ∼ t Ψ0(Mf(x0 + r0tθ))dσ(θ)dt 1 Sd−1 Z 2 Z . Ψ(x,Mf(x))dx. ZB Moreover, we deduce from (2.4) that M(f) belongs to LΨ(ρB \ 2B) and it thus follows that M(f) ∈ LΨ(ρB), as desired. Next, note that by the same reasoning as in the proof of sufficiency and by Fubini’s theorem, M(f)(x) Ψ(x, M(f)(x))dx & dx e log(M(f)(x)) ZρB ZρB∩{M(f)>max{e ,|x0|+r0}} M(f)(x) 1 & dα dx e e log α ZρB∩{M(f)>max{e ,|x0|+r0}} Ze ! ∞ 1 & · |{x ∈ ρB : M(f)(x) >c2 · α}|dα. e log α Zmax{e ,|x0|+r0} By using (2.3), we now get
|f(x)| 1 ∞ > Ψ(x, M(f)(x))dx & |f(x)| · dα dx e α log α ZρB ZB Zmax{e ,|x0|+r0} ! & 1+ |f(x)| log+ log+ |f(x)|dx ZB and this completes the proof of Theorem 1.
d −e Remark 4. Let B0 denote the closed unit ball in R . Given a small δ ∈ (0,e ), −d −d if, as on pp. 58–59 in [7], one considers f := δ χ{|x|<δ} then M(f)(x) ∼ |x| for all |x| > 2δ and so,
(2.7) |f(x)| log+ log+ |f(x)|dx ∼ log(log(δ−1)) ∼ Ψ(x, M(f)(x))dx. ZB0 ZB0 This shows that given LΨ(B0), the space L log log L(B0) in the statement of Theo- rem 1 is best possible in general, in terms of size. Indeed, the left-hand side of (2.7) follows by direct calculation. On the other hand, using (2.1), (2.6), a change to polar coordinates, and further change of vari- ables yield 1 1 ds Ψ(x, M(f)(x))dx ∼ 1+ −d B0 2δ log(e + s ) s Z Z −1 −1 (2δ) 1 du (2δ) 1 du ∼ 1+ ∼ 1+ , log(e + ud) u log(u) u Z1 Ze from where the right-hand side of (2.7) follows. 6 BAKAS,RODR´IGUEZ-LOPEZ,´ AND SOLA
2.1. Further generalizations. Assume that Ψ : Rd × [0, ∞) is a non-negative function satisfying the following properties: d (1) For every x ∈ R fixed, Ψ(x, t)=Ψx(t) is Orlicz in t ∈ [0, ∞), namely Ψx(0) = 0, Ψx is increasing on [0, ∞) with Ψx(t) > 0 for all t > 0 and Ψx(t) → ∞ as t → ∞. Moreover, assume that there exists an absolute constant C0 > 0 such d that Ψx(2t) ≤ C0Ψx(t) for all x ∈ R and every t ∈ [0, ∞). d (2) If K is a compact set in R , then there exist x1, x2 ∈ K and a constant CK > 0 such that −1 CK < Ψ(x1,t) ≤ Ψ(x, t) ≤ Ψ(x2,t) < CK for every x ∈ K and for all t> 0. t (3) If we write Ψ(x, t)=Ψx(t) = 0 ψx(s)ds, then for every α0, β0 with 0 < α <β one has 0 0 R β0 ψ (s) x ds < ∞ s Zα0 for every x ∈ Rd. By carefully examining the proof of Theorem 1, one obtains the following result.
t d Theorem 5. Let Ψ(x, t)= 0 ψx(s)ds, (x, t) ∈ R × [0, ∞), be as above. Fix a closed ball B with B ( Rd and let f be such that supp(f) ⊆ B. Then, R M(f) ∈ LΨ(B) if, and only if, |f(x)| ψ (s) |f(x)| · x ds dx < ∞ s Z{|f|>α0} Zα0 ! for every α0 > 0. Theorem 5 applies to certain Orlicz spaces considered in connection with con- vergence of Fourier series, see e.g. [1, 11], and the recent paper by V. Lie [10]; we give some sample applications in Subsection 4.1.
3. Proof of the Zygmund-type Theorem for Hlog(Rd) We begin with the following elementary lemmas. Lemma 6. Consider the function g : [0, ∞)2 → [0, ∞) given by 1 g(s,t) := , (s,t) ∈ [0, ∞)2. log(e + t) + log(e + s) Then one has t Ψ(x, t) ≤ g(|x|, τ)dτ ≤ 2Ψ(x, t) Z0 for all (x, t) ∈ Rd × [0, ∞). 1 Proof. The function t 7→ g(s,t)= log((e+t)(e+s)) is decreasing, so clearly t g(|x|,s)ds ≥ tg(|x|,t)=Ψ(x, t). Z0 We now address the upper bound. A calculation yields that tǫ ǫ 1 ∂ (tǫg(|x|,t)) = − , t log(e + t) + log(1 + |x|) t (e + t)(log(e + t) + log(e + |x|) and we observe that the term within the parenthesis is positive if, and only if, ǫ 1 − > 0, t (e + t)(log(e + t) + log(e + |x|) d STEIN AND ZYGMUND THEOREMS FOR Hlog (R ) 7
1 which for ǫ = 2 is equivalent to the inequality (e + t)(log(e + t) + log(e + |x|)) > 2t. But clearly (e + t)(log(e + t) + log(e + |x|)) ≥ 2(e + t) > 2t. Thus s 7→ sǫg(|x|,s) is increasing for ǫ =1/2, which implies that t t 1 g(|x|,s)ds = s−ǫsǫg(|x|,s)ds ≤ Ψ(x, t) = 2Ψ(x, t) 1 − ǫ Z0 Z0 and this completes the proof of the lemma. d d Lemma 7. Let x0 ∈ R be fixed and for u ∈ S(R ) define hτx0 f,ui := hf, τ−x0 ui, d where τ−x0 u(x) := u(x − x0), x ∈ R . log d log d Then f ∈ H (R ) if, and only if, τx0 f ∈ H (R ). d log d Proof. Note that it suffices to prove that for any x0 ∈ R and f ∈ H (R ) one log d also has that τx0 f ∈ H (R ). d log d Towards this aim, fix an x0 ∈ R and an f ∈ H (R ). Observe that, by using a change of variables and the translation invariance of Mφ, we may write M (τ f)(x) I := φ x0 dx Rd log(e + |x|) + log(e + M (τ f)(x)) Z φ x0 as M (f)(x) I = φ dx. Rd log(e + |x − x |) + log(e + M (f)(x)) Z 0 φ To prove that I < ∞, we split I = I1 + I2, where M (f)(x) I := φ dx 1 log(e + |x − x |) + log(e + M (f)(x)) Z|x|>4|x0| 0 φ and M (f)(x) I := φ dx. 2 log(e + |x − x |) + log(e + M (f)(x)) Z|x|≤4|x0| 0 φ To show that I1 < ∞, observe that for |x| > 4|x0| one has 4|x − x | 4|x − x | 0 < |x| < 0 5 3 and so, M (f)(x) I . φ dx 1 log(e + |x|) + log(e + M (f)(x)) Z|x|>4|x0| φ M (f)(x) ≤ φ dx. Rd log(e + |x|) + log(e + M (f)(x)) Z φ log d Since f ∈ H (R ), the last integral is finite and we thus deduce that I1 < ∞. Next, to show that I2 < ∞, we have M (f)(x) I ≤ φ dx 2 1 + log(e + M (f)(x)) Z|x|≤4|x0| φ M (f)(x) . φ dx |x0| log(e + |x|) + log(e + M (f)(x)) Z|x|≤4|x0| φ M (f)(x) ≤ φ dx Rd log(e + |x|) + log(e + M (f)(x)) Z φ log d and so, I2 < ∞, as f ∈ H (R ). Therefore, I < ∞ and it thus follows that log d τx0 f ∈ H (R ). 8 BAKAS,RODR´IGUEZ-LOPEZ,´ AND SOLA
To obtain the desired variant of Zygmund’s theorem, we shall use the fact that functions in Hlog(Rd) have mean zero; see Lemma 1.4 in [2]. For the convenience of the reader, we present a detailed proof of this fact below. Lemma 8 ([2]). If f ∈ Hlog(Rd) is a compactly supported integrable function, then Rd f(y)dy =0.
RProof. Let f be a given function in Hlog(Rd) with compact support. In light of Lemma 7, we may assume, without loss of generality, that f is supported in a closed d ball Br centered at 0 with radius r> 0, i.e. supp(f) ⊆ Br := {x ∈ R : |x|≤ r}. To prove the lemma, take an x ∈ Rd with |x| > 2r and observe that, by the definition of φǫ, we can take ǫ =4|x| to get 1 x − y 1 |f ∗ φ (x)| = f(y)φ dy & · f(y)dy ǫ ǫd ǫ |x|d ZBr ZBr
−1 as we then have φ(ǫ (x − y)) = c0 for y ∈ Br. Therefore, for all |x | > 2r and ǫ =4|x|, we have 1 Mφ(f)(x) & d · f(y)dy |x| Br Z and so, we deduce from Lemma 6 that
1 Ψ(x, Mφ(f)(x)) & d · f(y)dy |x| log(e + |x|) Br Z
for |x| large enough. Hence, if f(y)dy 6= 0, then the function Ψ(x, Mφ(f)(x)) does not belong to L1(Rd), which is a contradiction. R We are now ready to prove Theorem 2.
Proof of Theorem 2. Let B denote the unit closed ball in Rd. Fix a function f with supp(f) ⊆ B, B f(y)dy = 0 and f ∈ L log log L(B). First of all, observe that d R Mφ(f)(x) . M(f)(x) for all x ∈ R , where M(f) denotes the Hardy-Littlewood maximal function of f; see e.g. Theorem 2 on pp. 62–63 in [13]. We thus deduce from Lemma 6 that
d Ψ(x, Mφ(f)(x)) . Ψ(x, M(f)(x)) for all x ∈ R and hence, by using Theorem 1, we obtain
+ + (3.1) Ψ(x, Mφ(f)(x))dx . 1+ |f(x)| log log |f(x)|dx, Z2B Z2B where 2B := {x ∈ Rd : |x|≤ 2}. d To estimate the integral of Ψ(x, Mφ(f)(x)) for x ∈ R \ 2B, we shall make use of the cancellation of f. To be more specific, observe that if |x| > 2 then for every ǫ< |x|/2, one has that 1 x − y f ∗ φǫ(x)= d f(y)φ dy =0 ǫ B ǫ Z since |x−y|/ǫ > 1 whenever y ∈ B. Therefore, we may restrict ourselves to ǫ ≥ |x|/2 when |x| > 2. Hence, for ǫ ≥ |x|/2, by exploiting the cancellation of f and using a d STEIN AND ZYGMUND THEOREMS FOR Hlog (R ) 9
Lipschitz estimate on φǫ, we obtain 1 x − y 1 x − y x |f ∗ φǫ(x)| = d f(y)φ dy = d f(y) φ − φ dy ǫ B ǫ ǫ B ǫ ǫ Z Z h i
1 1 . |y · f(y)|dy . 1+ |f(y)| log+ log+ |f(y)|dy . φ ǫd+1 |x|d+1 ZB " ZB # We thus deduce that, for every x ∈ Rd \ 2B, 1 |M (f)(x)| . 1+ |f(y)| log+ log+ |f(y)|dy φ |x|d+1 " ZB # and so,
Ψ(x, Mφ(x))dx Rd Z \2B 1 . 1+ |f(y)| log+ log+ |f(y)|dy · dx Rd |x|d+1 log(e + |x|) " ZB # Z \2B . 1+ |f(y)| log+ log+ |f(y)|dy, ZB as desired. Therefore, Theorem 2 is now established by using the last estimate combined with (3.1). 3.1. A partial converse. As in the classical setting of the real Hardy space H1, see [12], Theorem 2 has a partial converse. To be more precise, if a function f is positive on an open set U and f belongs to Hlog(Rd), then the function f ∈ L log log L(K) for every compact set K ⊂ U. Indeed, to see this, note that if f is as above then
Mφ(f)(x) & M(f · ηK )(x) for all x ∈ K, where ηK is an appropriate Schwartz function with ηK ∼ 1 on K; see e.g. Section 5.3 in Chapter III in [14]. Hence, by using Lemma 6 and Theorem 1, we get
Ψ(x, Mφ(f)(x))dx ≥ Ψ(x, Mφ(f)(x))dx & Ψ(x, M(ηK · f)(x))dx Rd Z ZK ZK & 1+ |f(x)| log+ log+ |f(x)|dx. ZK 4. Variants in the periodic setting Following [2], define Hlog(D) to be the space of all holomorphic functions F on the unit disk D of C such that 2π |F (reiθ)| sup dθ < ∞. log(e + |F (reiθ)|) 0 Theorem 2.2] for details. Moreover, by using [2, Proposition 8.2], one may identify Hlog(D) with the space of all measurable functions f on the torus such that 2π Ψ0 sup |Pr ∗ f(θ)| dθ < ∞, Z0 0 There is a periodic version of Theorem 1, namely M(f) ∈ LΨ0 (T) if, and only if, f ∈ L log log L(T). Combining this with Lemma 6, one obtains the following result. Proposition 9. One has the inclusion L log log L(T) ⊆ Hlog(T). Moreover, arguing as in the previous section and using the necessity in Theorem 1 as well as Proposition 9 and Lemma 6, one can show that if f ∈ Hlog(T) and f is non-negative, then f ∈ L log log L(T). Proposition 10. One has {f ∈ L log log L(T): f ≥ 0 a.e. on T} = {f ∈ Hlog(T): f ≥ 0 a.e. on T}. Proof. Note that Proposition 9 implies that (4.1) {f ∈ L log log L(T): f ≥ 0 a.e. on T}⊆{f ∈ Hlog(T): f ≥ 0 a.e. on T}. To prove the reverse inclusion, take a non-negative function f ∈ Hlog(T) and notice that it follows from the work of Stein [12] that c (4.2) |{θ ∈ T : M(f)(θ) >c α}| ≥ 2 |f(θ)|dθ, 1 α Z{|f|>α} where c1,c2 > 0 are absolute constants. Hence, by arguing as in the proof of Theorem 1, it follows from (4.2) (noting that the periodic case is easier as one does not need to consider the contribution away from the support of f) that + + (4.3) Ψ0(M(f))(θ)dθ & 1+ |f(x)| log log |f(θ)|dθ. T T Z Z Since f ≥ 0 a.e. on T, as in the Euclidean case, one has (4.4) sup |Pr ∗ f(θ)| & M(f)(θ) fora.e. θ ∈ T. 0 This allows us to conclude that, for this choice of Ψ, 2 M(f) ∈ LΨ(T) if, and only if, f ∈ L log L log log L(T). Turning to the space L log log L log log log log L appearing in Lie’s paper [10], we can check where the maximal operator maps this space. Performing the appropriate computations, we obtain that M(f) log+ log+ log+ log+ M(f)dx < ∞ T log(M(f)+ e) Z if, and only if, f ∈ L log log L log log log log L(T). Roughly speaking, the contents of Theorem 5 and the computations presented above can be summarized as follows. Let Φ0 be a given Orlicz function, namely Φ0 : [0, ∞) → [0, ∞) is an increasing function with Φ0(0) = 0 and Φ0(t) → ∞ as t → ∞. Suppose that one can find non-negative, increasing functions M,S with Φ0(t)= M(t) · S(t) (t> 0) and such that, for 0 <α f ∈ LΦ0 (T) if, and only if,M(f) ∈ LFα·S(T), for any α ≥ α0. Acknowledgments. AS extends his thanks to Kelly Bickel and the rest of the Department of Mathematics at Bucknell University (Lewisburg, PA) for hospitality during a visit where part of this work was carried out. References [1] Antonov, Nikolai Yu. Convergence of Fourier series. East J. Approx. 2 (1996), 187–196. [2] Bonami, Aline, Tadeusz Iwaniec, Peter Jones, and Michel Zinsmeister. On the Product of Functions in BMO and H1. Ann. Inst. Fourier (Grenoble) 57, no. 5 (2007), 1405–1439. [3] Bonami, Aline, Sandrine Grellier, and Luong Dang Ky. Paraproducts and prod- ucts of functions in BMO(Rn) and H1(Rn) through wavelets. J. Math. Pures Appl. 97, no. 3 (2012), 230–241. [4] Bonami, Aline, and Shobha Madan. Balayage of Carleson measures and Hankel operators on generalized Hardy spaces. Math. Nachr. 153 (1991), no. 1, 237–245. [5] Bonami, Aline, Jun Cao, Luong Dang Ky, Liguang Liu, Dachun Yang, and Wen Yuan. Multiplication between Hardy spaces and their dual spaces. J. Math. Pures Appl. 131 (2019), 130–170. [6] Duren, Peter L. Theory of Hp spaces. Pure and Applied Mathematics, Vol. 38 Academic Press, New York-London 1970 xii+258 pp. [7] Fefferman, Robert. Multiparameter Fourier analysis. Beijing lectures in har- monic analysis (Beijing, 1984), 47-130, Ann. of Math. Stud., 112, Princeton Univ. Press, Princeton, NJ, 1986. 12 BAKAS, RODR´IGUEZ-LOPEZ,´ AND SOLA [8] Iwaniec, Tadeusz, and Anna Verde. A study of Jacobians in Hardy-Orlicz spaces. Proc. Roy. Soc. Edinburgh Sect. A 129 (1999), no. 3, 539–570. [9] Ky, Luong Dang. New Hardy spaces of Musielak-Orlicz type and boundedness of sublinear operators. Integral Equations Operator Theory 78 (2014), no. 1, 115–150. [10] Lie, Victor. Pointwise convergence of Fourier series (I). On a conjecture of Konyagin. J. Eur. Math. Soc. 19 (2017), 1655-1728. [11] Sj¨olin, Per. An inequality of Paley and convergence a.e. of Walsh-Fourier se- ries. Ark. Mat. 7 (1969), 551–570. [12] Stein, Elias M. Note on the class L log L. Studia Math. 32, no. 3 (1969), 305– 310. [13] Stein, Elias M. Singular integrals and differentiability properties of functions. Princeton Math. Series Vol. 30. Princeton University Press, 2016. [14] Stein, Elias M. Harmonic Analysis: Real-Variable Methods, Orthogonality, and Oscillatory Integrals. Princeton Math. Series Vol. 43. Princeton University Press, 2016. [15] Yang, Dachun, Yiyu Liang, and Luong Dang Ky. Real-variable theory of Musielak-Orlicz Hardy spaces. Lecture Notes in Math. Vol. 2182. Springer, 2017. [16] Zygmund, Antoni. Trigonometric series. Vol. I, II. Cambridge University Press, 2002. Centre for Mathematical Sciences, Lund University, 221 00 Lund, Sweden E-mail address: [email protected] Department of Mathematics, Stockholm University, 106 91 Stockholm, Sweden E-mail address: [email protected] Department of Mathematics, Stockholm University, 106 91 Stockholm, Sweden E-mail address: [email protected]