Endpoint regularity of the dyadic and the fractional maximal function
Julian Weigt
1/16
Background
Forf :Rd →Rthe uncentered Hardy-Littlewood maximal function is defined by
Mf(x) = sup
B3x
fB with fB = 1
L(B) ˆ
B
|f|.
The Hardy-Littlewood maximal function theorem:
kMfkLp(Rd)≤Cd,pkfkLp(Rd) if and only ifp >1 Juha Kinnunen (1997):
k∇MfkLp(Rd)≤Cd,pk∇fkLp(Rd) ifp >1
Question (Haj lasz and Onninen 2004) Is it true that
k∇MfkL1(Rd)≤Cdk∇fkL1(Rd)?
Background
Forf :Rd →Rthe uncentered Hardy-Littlewood maximal function is defined by
Mf(x) = sup
B3x
fB with fB = 1
L(B) ˆ
B
|f|.
The Hardy-Littlewood maximal function theorem:
kMfkLp(Rd)≤Cd,pkfkLp(Rd) if and only ifp >1 Juha Kinnunen (1997):
k∇MfkLp(Rd)≤Cd,pk∇fkLp(Rd) ifp >1
Question (Haj lasz and Onninen 2004) Is it true that
k∇MfkL1(Rd)≤Cdk∇fkL1(Rd)?
1/16
Background
Forf :Rd →Rthe uncentered Hardy-Littlewood maximal function is defined by
Mf(x) = sup
B3x
fB with fB = 1
L(B) ˆ
B
|f|.
The Hardy-Littlewood maximal function theorem:
kMfkLp(Rd)≤Cd,pkfkLp(Rd) if and only ifp >1 Juha Kinnunen (1997):
k∇MfkLp(Rd)≤Cd,pk∇fkLp(Rd) ifp >1
Question (Haj lasz and Onninen 2004) Is it true that
k∇MfkL1(Rd)≤Cdk∇fkL1(Rd)?
Progress on classical Hardy-Littlewood
d = 1 [Tanaka 2002, Aldaz+P´erez L´azaro 2007]
block decreasing f [Aldaz+P´erez L´azaro 2009]
centered M,d = 1 [Kurka 2015]
radial f [Luiro 2018]
characteristic f [W 2020]
3/16
The dyadic and the fractional maximal function
Forα∈(0,d) define the uncentered fractional Hardy-Littlewood maximal function by
Mαf(x) = sup
B3x
r(B)αfB, wherer(B) is the radius ofB.
The corresponding bound is k∇MαfkLd/(d−α)(Rd)≤Cd,αk∇fk1
The version withp >1 follows the same way as for α= 0 in Kinnunen (1997). The dyadic maximal function is
Mdf(x) = sup
Qdyadic cube,x∈Q
fQ The corresponding bound is
varMdf ≤Cdvarf.
The dyadic and the fractional maximal function
Forα∈(0,d) define the uncentered fractional Hardy-Littlewood maximal function by
Mαf(x) = sup
B3x
r(B)αfB,
wherer(B) is the radius ofB. The corresponding bound is k∇MαfkLd/(d−α)(Rd)≤Cd,αk∇fk1
The version withp >1 follows the same way as for α= 0 in Kinnunen (1997).
The dyadic maximal function is Mdf(x) = sup
Qdyadic cube,x∈Q
fQ The corresponding bound is
varMdf ≤Cdvarf.
3/16
The dyadic and the fractional maximal function
Forα∈(0,d) define the uncentered fractional Hardy-Littlewood maximal function by
Mαf(x) = sup
B3x
r(B)αfB,
wherer(B) is the radius ofB. The corresponding bound is k∇MαfkLd/(d−α)(Rd)≤Cd,αk∇fk1
The version withp >1 follows the same way as for α= 0 in Kinnunen (1997). The dyadic maximal function is
Mdf(x) = sup
Qdyadic cube,x∈Q
fQ
The corresponding bound is
varMdf ≤Cdvarf.
The dyadic and the fractional maximal function
Forα∈(0,d) define the uncentered fractional Hardy-Littlewood maximal function by
Mαf(x) = sup
B3x
r(B)αfB,
wherer(B) is the radius ofB. The corresponding bound is k∇MαfkLd/(d−α)(Rd)≤Cd,αk∇fk1
The version withp >1 follows the same way as for α= 0 in Kinnunen (1997). The dyadic maximal function is
Mdf(x) = sup
Qdyadic cube,x∈Q
fQ The corresponding bound is
4/16
Progress on related operators
convolution, d = 1 [Carneiro+Svaiter 2013]
uncentered fractional, α >1 [Kinnunen+Saksman 2003, Carneiro+Madrid 2017]
fractional smooth convolution [Beltran+Ramos+Saari 2018]
fractional lacunary [Beltran+Ramos+Saari 2018]
convolution, radial f [Carneiro+Svaiter 2019]
fractional, radialf [Beltran+Madrid 2019]
dyadic [W 2020]
fractional [W 2020]
related: Continuity off 7→ ∇Mf as W1,1(Rd)→L1(Rd) (does not follow from boundedness), boundedness on other spaces, local maximal operator.
Progress on related operators
convolution, d = 1 [Carneiro+Svaiter 2013]
uncentered fractional, α >1 [Kinnunen+Saksman 2003, Carneiro+Madrid 2017]
fractional smooth convolution [Beltran+Ramos+Saari 2018]
fractional lacunary [Beltran+Ramos+Saari 2018]
convolution, radial f [Carneiro+Svaiter 2019]
fractional, radialf [Beltran+Madrid 2019]
dyadic [W 2020]
fractional [W 2020]
related: Continuity off 7→ ∇Mf as W1,1(Rd)→L1(Rd) (does not follow from boundedness), boundedness on other spaces, local maximal operator.
4/16
Progress on related operators
convolution, d = 1 [Carneiro+Svaiter 2013]
uncentered fractional, α >1 [Kinnunen+Saksman 2003, Carneiro+Madrid 2017]
fractional smooth convolution [Beltran+Ramos+Saari 2018]
fractional lacunary [Beltran+Ramos+Saari 2018]
convolution, radial f [Carneiro+Svaiter 2019]
fractional, radialf [Beltran+Madrid 2019]
dyadic [W 2020]
fractional [W 2020]
related: Continuity off 7→ ∇Mf as W1,1(Rd)→L1(Rd) (does not follow from boundedness), boundedness on other spaces, local maximal operator.
1. The Dyadic Maximal Operator
By the coarea formula
and{f > λ} ⊂ {Mdf > λ} up to measure zero
varMdf = ˆ
R
Hd−1(∂{Mdf > λ})dλ
≤ ˆ
R
Hd−1(∂{Mdf > λ} \ {f > λ}) +Hd−1(∂{f > λ})dλ
Since ˆ
R
Hd−1(∂{f > λ})dλ= varf it suffices to bound the first summand.
5/16
1. The Dyadic Maximal Operator
By the coarea formula and{f > λ} ⊂ {Mdf > λ} up to measure zero
varMdf = ˆ
R
Hd−1(∂{Mdf > λ})dλ
≤ ˆ
R
Hd−1(∂{Mdf > λ} \ {f > λ}) +Hd−1(∂{f > λ})dλ
Since ˆ
R
Hd−1(∂{f > λ})dλ= varf it suffices to bound the first summand.
1. The Dyadic Maximal Operator
By the coarea formula and{f > λ} ⊂ {Mdf > λ} up to measure zero
varMdf = ˆ
R
Hd−1(∂{Mdf > λ})dλ
≤ ˆ
R
Hd−1(∂{Mdf > λ} \ {f > λ}) +Hd−1(∂{f > λ})dλ
Since ˆ
R
Hd−1(∂{f > λ})dλ= varf it suffices to bound the first summand.
6/16
Definition
Q is maximal forλ <fQ if for allP )Q we havefP ≤λ.
GivenQ, let λQ be the smallest suchλ. ˆ
R
Hd−1(∂{Mdf > λ} \ {f > λ})dλ
= ˆ
R
Hd−1(∂[
{maximal Q :fQ > λ} \ {f > λ})dλ
≤ ˆ
R
X
maximalQ:fQ>λ
Hd−1(∂Q\ {f > λ})dλ
= ˆ
R
X
Q:λQ<λ<fQ
Hd−1(∂Q\ {f > λ})dλ
= ˆ
R
X
Q:λQ<λ<λ˜Q
+ X
Q:˜λQ<λ<fQ
Hd−1(∂Q\ {f > λ})dλ
where ”L(Q∩ {f >˜λQ}) = 2−d−2L(Q)”.
Definition
Q is maximal forλ <fQ if for allP )Q we havefP ≤λ.
GivenQ, let λQ be the smallest suchλ.
ˆ
R
Hd−1(∂{Mdf > λ} \ {f > λ})dλ
= ˆ
R
Hd−1(∂[
{maximal Q :fQ > λ} \ {f > λ})dλ
≤ ˆ
R
X
maximalQ:fQ>λ
Hd−1(∂Q\ {f > λ})dλ
= ˆ
R
X
Q:λQ<λ<fQ
Hd−1(∂Q\ {f > λ})dλ
= ˆ
R
X
Q:λQ<λ<λ˜Q
+ X
Q:˜λQ<λ<fQ
Hd−1(∂Q\ {f > λ})dλ
where ”L(Q∩ {f >˜λQ}) = 2−d−2L(Q)”.
6/16
Definition
Q is maximal forλ <fQ if for allP )Q we havefP ≤λ.
GivenQ, let λQ be the smallest suchλ.
ˆ
R
Hd−1(∂{Mdf > λ} \ {f > λ})dλ
= ˆ
R
Hd−1(∂[
{maximal Q :fQ > λ} \ {f > λ})dλ
≤ ˆ
R
X
maximalQ:fQ>λ
Hd−1(∂Q\ {f > λ})dλ
= ˆ
R
X
Q:λQ<λ<fQ
Hd−1(∂Q\ {f > λ})dλ
= ˆ
R
X
Q:λQ<λ<λ˜Q
+ X
Q:˜λQ<λ<fQ
Hd−1(∂Q\ {f > λ})dλ
where ”L(Q∩ {f >˜λQ}) = 2−d−2L(Q)”.
Definition
Q is maximal forλ <fQ if for allP )Q we havefP ≤λ.
GivenQ, let λQ be the smallest suchλ.
ˆ
R
Hd−1(∂{Mdf > λ} \ {f > λ})dλ
= ˆ
R
Hd−1(∂[
{maximal Q :fQ > λ} \ {f > λ})dλ
≤ ˆ
R
X
maximalQ:fQ>λ
Hd−1(∂Q\ {f > λ})dλ
= ˆ
R
X
Q:λQ<λ<fQ
Hd−1(∂Q\ {f > λ})dλ
= ˆ
R
X
Q:λQ<λ<λ˜Q
+ X
Q:˜λQ<λ<fQ
Hd−1(∂Q\ {f > λ})dλ
where ”L(Q∩ {f >˜λQ}) = 2−d−2L(Q)”.
6/16
Definition
Q is maximal forλ <fQ if for allP )Q we havefP ≤λ.
GivenQ, let λQ be the smallest suchλ.
ˆ
R
Hd−1(∂{Mdf > λ} \ {f > λ})dλ
= ˆ
R
Hd−1(∂[
{maximal Q :fQ > λ} \ {f > λ})dλ
≤ ˆ
R
X
maximalQ:fQ>λ
Hd−1(∂Q\ {f > λ})dλ
= ˆ
R
X
Q:λQ<λ<fQ
Hd−1(∂Q\ {f > λ})dλ
= ˆ
R
X
Q:λQ<λ<λ˜Q
+ X
Q:˜λQ<λ<fQ
Hd−1(∂Q\ {f > λ})dλ
where ”L(Q∩ {f >˜λQ}) = 2−d−2L(Q)”.
Definition
Q is maximal forλ <fQ if for allP )Q we havefP ≤λ.
GivenQ, let λQ be the smallest suchλ.
ˆ
R
Hd−1(∂{Mdf > λ} \ {f > λ})dλ
= ˆ
R
Hd−1(∂[
{maximal Q :fQ > λ} \ {f > λ})dλ
≤ ˆ
R
X
maximalQ:fQ>λ
Hd−1(∂Q\ {f > λ})dλ
= ˆ
R
X
Q:λQ<λ<fQ
Hd−1(∂Q\ {f > λ})dλ
=
ˆ X
+ X
Hd−1(∂Q\ {f > λ})dλ
7/16
High density case
Proposition (High density) ForL(Q∩E)≥εL(Q) we have
Hd−1(∂Q\E).εHd−1(Q∩∂E).
ˆ
R
X
Q:λQ<λ<˜λQ
Hd−1(∂Q\ {f > λ})dλ
. ˆ
R
X
Q:λQ<λ<˜λQ
Hd−1(Q∩∂{f > λ})dλ
≤ ˆ
R
Hd−1(∂{f > λ})dλ
= varf
High density case
Proposition (High density) ForL(Q∩E)≥εL(Q) we have
Hd−1(∂Q\E).εHd−1(Q∩∂E).
ˆ
R
X
Q:λQ<λ<˜λQ
Hd−1(∂Q\ {f > λ})dλ
. ˆ
R
X
Q:λQ<λ<˜λQ
Hd−1(Q∩∂{f > λ})dλ
≤ ˆ
R
Hd−1(∂{f > λ})dλ
= varf
7/16
High density case
Proposition (High density) ForL(Q∩E)≥εL(Q) we have
Hd−1(∂Q\E).εHd−1(Q∩∂E).
ˆ
R
X
Q:λQ<λ<˜λQ
Hd−1(∂Q\ {f > λ})dλ
. ˆ
R
X
Q:λQ<λ<˜λQ
Hd−1(Q∩∂{f > λ})dλ
≤ ˆ
R
Hd−1(∂{f > λ})dλ
= varf
High density case
Proposition (High density) ForL(Q∩E)≥εL(Q) we have
Hd−1(∂Q\E).εHd−1(Q∩∂E).
ˆ
R
X
Q:λQ<λ<˜λQ
Hd−1(∂Q\ {f > λ})dλ
. ˆ
R
X
Q:λQ<λ<˜λQ
Hd−1(Q∩∂{f > λ})dλ
≤ ˆ
R
Hd−1(∂{f > λ})dλ
E
Q
L(Q∩E)≥εL(Q)
=⇒ Hd−1(∂Q\E).εHd−1(Q∩∂E)
Low density case
Estimate ˆ
R
X
Q:˜λQ<λ<fQ
Hd−1(∂Q\ {f > λ})dλ
≤ ˆ
R
X
Q:˜λQ<λ<fQ
Hd−1(∂Q)dλ
=X
Q
(fQ−˜λQ)Hd−1(∂Q)
We have to estimate this by varf.
9/16
Low density case
Estimate ˆ
R
X
Q:˜λQ<λ<fQ
Hd−1(∂Q\ {f > λ})dλ
≤ ˆ
R
X
Q:˜λQ<λ<fQ
Hd−1(∂Q)dλ
=X
Q
(fQ−˜λQ)Hd−1(∂Q)
We have to estimate this by varf.
Low density case
Estimate ˆ
R
X
Q:˜λQ<λ<fQ
Hd−1(∂Q\ {f > λ})dλ
≤ ˆ
R
X
Q:˜λQ<λ<fQ
Hd−1(∂Q)dλ
=X
Q
(fQ−˜λQ)Hd−1(∂Q)
We have to estimate this by varf.
9/16
Low density case
Estimate ˆ
R
X
Q:˜λQ<λ<fQ
Hd−1(∂Q\ {f > λ})dλ
≤ ˆ
R
X
Q:˜λQ<λ<fQ
Hd−1(∂Q)dλ
=X
Q
(fQ−˜λQ)Hd−1(∂Q)
We have to estimate this by varf.
Relative isoperimetric inequality
ForL(Q∩E)≤ L(Q)/2 the relative isoperimetric inequality states L(Q∩E)d−1d .Hd−1(Q∩∂E)
Proposition
(fQ −λ˜Q)L(Q). ˆ
R
X
P(Q:¯λP<λ<fP
L(P ∩ {f > λ})dλ
whereP is maximal above ¯λP and
”L(P ∩ {f >¯λP}) = 2−1L(P)”
”L(Q∩ {f >˜λQ}) = 2−d−2L(Q)”
10/16
Relative isoperimetric inequality
ForL(Q∩E)≤ L(Q)/2 the relative isoperimetric inequality states L(Q∩E)d−1d .Hd−1(Q∩∂E)
Proposition
(fQ −λ˜Q)L(Q). ˆ
R
X
P(Q:¯λP<λ<fP
L(P ∩ {f > λ})dλ
whereP is maximal above ¯λP and
”L(P ∩ {f >¯λP}) = 2−1L(P)”
”L(Q∩ {f >λ˜Q}) = 2−d−2L(Q)”
fQ
fP1
fP2
P3
fP3
12/16
X
Q
(fQ−˜λQ)Hd−1(∂Q)
. ˆ
R
X
Q
l(Q)−1 X
P(Q:¯λP<λ<fP
L(P∩ {f > λ})dλ
= ˆ
R
X
P:¯λP<λ<fP
L(P∩ {f > λ})X
Q)P
l(Q)−1dλ
= ˆ
R
X
P:¯λP<λ<fP
L(P∩ {f > λ}) l(P)−1dλ
≤ ˆ
R
X
P:¯λP<λ<fP
L(P∩ {f > λ})d−1d dλ
. ˆ
R
X
P:¯λP<λ<fP
Hd−1(P∩∂{f > λ})dλ
≤ ˆ
R
Hd−1(∂{f > λ})dλ= varf
X
Q
(fQ−˜λQ)Hd−1(∂Q)
. ˆ
R
X
Q
l(Q)−1 X
P(Q:¯λP<λ<fP
L(P∩ {f > λ})dλ
= ˆ
R
X
P:¯λP<λ<fP
L(P ∩ {f > λ})X
Q)P
l(Q)−1dλ
= ˆ
R
X
P:¯λP<λ<fP
L(P∩ {f > λ}) l(P)−1dλ
≤ ˆ
R
X
P:¯λP<λ<fP
L(P∩ {f > λ})d−1d dλ
. ˆ
R
X
P:¯λP<λ<fP
Hd−1(P∩∂{f > λ})dλ
≤ ˆ
R
Hd−1(∂{f > λ})dλ= varf
12/16
X
Q
(fQ−˜λQ)Hd−1(∂Q)
. ˆ
R
X
Q
l(Q)−1 X
P(Q:¯λP<λ<fP
L(P∩ {f > λ})dλ
= ˆ
R
X
P:¯λP<λ<fP
L(P ∩ {f > λ})X
Q)P
l(Q)−1dλ
= ˆ
R
X
P:¯λP<λ<fP
L(P ∩ {f > λ}) l(P)−1dλ
≤ ˆ
R
X
P:¯λP<λ<fP
L(P∩ {f > λ})d−1d dλ
. ˆ
R
X
P:¯λP<λ<fP
Hd−1(P∩∂{f > λ})dλ
≤ ˆ
R
Hd−1(∂{f > λ})dλ= varf
X
Q
(fQ−˜λQ)Hd−1(∂Q)
. ˆ
R
X
Q
l(Q)−1 X
P(Q:¯λP<λ<fP
L(P∩ {f > λ})dλ
= ˆ
R
X
P:¯λP<λ<fP
L(P ∩ {f > λ})X
Q)P
l(Q)−1dλ
= ˆ
R
X
P:¯λP<λ<fP
L(P ∩ {f > λ}) l(P)−1dλ
≤ ˆ
R
X
P:¯λP<λ<fP
L(P ∩ {f > λ})d−1d dλ
. ˆ
R
X
P:¯λP<λ<fP
Hd−1(P∩∂{f > λ})dλ
≤ ˆ
R
Hd−1(∂{f > λ})dλ= varf
12/16
X
Q
(fQ−˜λQ)Hd−1(∂Q)
. ˆ
R
X
Q
l(Q)−1 X
P(Q:¯λP<λ<fP
L(P∩ {f > λ})dλ
= ˆ
R
X
P:¯λP<λ<fP
L(P ∩ {f > λ})X
Q)P
l(Q)−1dλ
= ˆ
R
X
P:¯λP<λ<fP
L(P ∩ {f > λ}) l(P)−1dλ
≤ ˆ
R
X
P:¯λP<λ<fP
L(P ∩ {f > λ})d−1d dλ
. ˆ
R
X
P:¯λP<λ<fP
Hd−1(P∩∂{f > λ})dλ
≤ ˆ
R
Hd−1(∂{f > λ})dλ= varf
X
Q
(fQ−˜λQ)Hd−1(∂Q)
. ˆ
R
X
Q
l(Q)−1 X
P(Q:¯λP<λ<fP
L(P∩ {f > λ})dλ
= ˆ
R
X
P:¯λP<λ<fP
L(P ∩ {f > λ})X
Q)P
l(Q)−1dλ
= ˆ
R
X
P:¯λP<λ<fP
L(P ∩ {f > λ}) l(P)−1dλ
≤ ˆ
R
X
P:¯λP<λ<fP
L(P ∩ {f > λ})d−1d dλ
. ˆ
R
X
P:¯λP<λ<fP
Hd−1(P∩∂{f > λ})dλ ˆ
12/16
X
Q
(fQ−˜λQ)Hd−1(∂Q)
. ˆ
R
X
Q
l(Q)−1 X
P(Q:¯λP<λ<fP
L(P∩ {f > λ})dλ
= ˆ
R
X
P:¯λP<λ<fP
L(P ∩ {f > λ})X
Q)P
l(Q)−1dλ
= ˆ
R
X
P:¯λP<λ<fP
L(P ∩ {f > λ}) l(P)−1dλ
≤ ˆ
R
X
P:¯λP<λ<fP
L(P ∩ {f > λ})d−1d dλ
. ˆ
R
X
P:¯λP<λ<fP
Hd−1(P∩∂{f > λ})dλ
≤ ˆ
R
Hd−1(∂{f > λ})dλ= varf
2. The fractional maximal operator
Recall 0< α <d and
Mαf(x) = sup
B3x
r(B)αfB.
Then for almost everyx ∈Rd the supremum is attained in some ballB with x∈B. Denote byBα the set of all optimal balls forf. Want to show
k∇Mαfkd/(d−α)≤Cd,αk∇fk1.
Kinnunen and Saksman (2003) For an optimal ballB for x we have
|∇Mαf(x)| ≤(d −α)r(B)α−1fB. Conclude
|∇Mαf(x)|. sup
B∈Bα,x∈B
r(B)α−1fB =:Mα,−1f(x).
13/16
2. The fractional maximal operator
Recall 0< α <d and
Mαf(x) = sup
B3x
r(B)αfB.
Then for almost everyx ∈Rd the supremum is attained in some ballB with x∈B. Denote byBα the set of all optimal balls forf.
Want to show
k∇Mαfkd/(d−α)≤Cd,αk∇fk1.
Kinnunen and Saksman (2003) For an optimal ballB for x we have
|∇Mαf(x)| ≤(d −α)r(B)α−1fB. Conclude
|∇Mαf(x)|. sup
B∈Bα,x∈B
r(B)α−1fB =:Mα,−1f(x).
2. The fractional maximal operator
Recall 0< α <d and
Mαf(x) = sup
B3x
r(B)αfB.
Then for almost everyx ∈Rd the supremum is attained in some ballB with x∈B. Denote byBα the set of all optimal balls forf. Want to show
k∇Mαfkd/(d−α)≤Cd,αk∇fk1.
Kinnunen and Saksman (2003) For an optimal ballB for x we have
|∇Mαf(x)| ≤(d −α)r(B)α−1fB. Conclude
|∇Mαf(x)|. sup
B∈Bα,x∈B
r(B)α−1fB =:Mα,−1f(x).
13/16
2. The fractional maximal operator
Recall 0< α <d and
Mαf(x) = sup
B3x
r(B)αfB.
Then for almost everyx ∈Rd the supremum is attained in some ballB with x∈B. Denote byBα the set of all optimal balls forf. Want to show
k∇Mαfkd/(d−α)≤Cd,αk∇fk1.
Kinnunen and Saksman (2003) For an optimal ballB for x we have
|∇Mαf(x)| ≤(d−α)r(B)α−1fB.
Conclude
|∇Mαf(x)|. sup
B∈Bα,x∈B
r(B)α−1fB =:Mα,−1f(x).
2. The fractional maximal operator
Recall 0< α <d and
Mαf(x) = sup
B3x
r(B)αfB.
Then for almost everyx ∈Rd the supremum is attained in some ballB with x∈B. Denote byBα the set of all optimal balls forf. Want to show
k∇Mαfkd/(d−α)≤Cd,αk∇fk1.
Kinnunen and Saksman (2003) For an optimal ballB for x we have
|∇Mαf(x)| ≤(d−α)r(B)α−1fB. Conclude
14/16
1. Make disjoint ˆ
(Mα,−1f)d−αd = ˆ
sup
B∈Bα
(r(B)α−1fB)d−αd 1B
.α,c1,c2
ˆ X
B∈Beα
(r(B)α−1fB)d−αd 1B
∼α X
B∈Beα
(fBHd−1(∂B))d−αd
whereBeα⊂ Bα such that for two ballsB,C ∈Bewe have c1B∩c1C =∅, or r(C)<r(B) andfC >c2fB.
Ifα−1≥0: Vitali covering argument suffices. Ifα−1<0: Use that ifB,C ∈ Bα with C ⊂B and
r(C)<r(B)/N we haver(C)αfC >r(B)αfB and thus fC >NαfB.
1. Make disjoint ˆ
(Mα,−1f)d−αd = ˆ
sup
B∈Bα
(r(B)α−1fB)d−αd 1B .α,c1,c2
ˆ X
B∈Beα
(r(B)α−1fB)d−αd 1B
∼α X
B∈Beα
(fBHd−1(∂B))d−αd
whereBeα⊂ Bα such that for two ballsB,C ∈Bewe have c1B∩c1C =∅, or r(C)<r(B) andfC >c2fB.
Ifα−1≥0: Vitali covering argument suffices. Ifα−1<0: Use that ifB,C ∈ Bα with C ⊂B and
r(C)<r(B)/N we haver(C)αfC >r(B)αfB and thus fC >NαfB.
14/16
1. Make disjoint ˆ
(Mα,−1f)d−αd = ˆ
sup
B∈Bα
(r(B)α−1fB)d−αd 1B .α,c1,c2
ˆ X
B∈Beα
(r(B)α−1fB)d−αd 1B
∼α X
B∈Beα
(fBHd−1(∂B))d−αd
whereBeα⊂ Bα such that for two ballsB,C ∈Bewe have c1B∩c1C =∅, or r(C)<r(B) andfC >c2fB.
Ifα−1≥0: Vitali covering argument suffices.
Ifα−1<0: Use that ifB,C ∈ Bα with C ⊂B and
r(C)<r(B)/N we haver(C)αfC >r(B)αfB and thus fC >NαfB.
1. Make disjoint ˆ
(Mα,−1f)d−αd = ˆ
sup
B∈Bα
(r(B)α−1fB)d−αd 1B .α,c1,c2
ˆ X
B∈Beα
(r(B)α−1fB)d−αd 1B
∼α X
B∈Beα
(fBHd−1(∂B))d−αd
whereBeα⊂ Bα such that for two ballsB,C ∈Bewe have c1B∩c1C =∅, or r(C)<r(B) andfC >c2fB.
Ifα−1≥0: Vitali covering argument suffices.
Ifα−1<0: Use that ifB,C ∈ Bα with C ⊂B and
14/16
1. Make disjoint ˆ
(Mα,−1f)d−αd = ˆ
sup
B∈Bα
(r(B)α−1fB)d−αd 1B .α,c1,c2
ˆ X
B∈Beα
(r(B)α−1fB)d−αd 1B
∼α X
B∈Beα
(fBHd−1(∂B))d−αd
whereBeα⊂ Bα such that for two ballsB,C ∈Bewe have c1B∩c1C =∅, or r(C)<r(B) andfC >c2fB.
Ifα−1≥0: Vitali covering argument suffices.
Ifα−1<0: Use that ifB,C ∈ Bα with C ⊂B and
r(C)<r(B)/N we haver(C)αfC >r(B)αfB and thus fC >NαfB.
2. Reduce to dyadic
X
B∈Beα
(fBHd−1(∂B))d−αd
≤
X
B∈B˜α
fBHd−1(∂B) d−αd
.α
X
Q∈Q˜α
fQHd−1(∂Q) d−αd
≤Cd,α(varf)d−αd
where l(Q)∼αr(B) and fQ ∼α fB so that alsocαQ∩cαP =∅, or l(P)<l(Q) andfP >2fQ.
15/16
2. Reduce to dyadic
X
B∈Beα
(fBHd−1(∂B))d−αd ≤
X
B∈B˜α
fBHd−1(∂B) d−αd
.α
X
Q∈Q˜α
fQHd−1(∂Q) d−αd
≤Cd,α(varf)d−αd
where l(Q)∼αr(B) and fQ ∼α fB so that alsocαQ∩cαP =∅, or l(P)<l(Q) andfP >2fQ.
2. Reduce to dyadic
X
B∈Beα
(fBHd−1(∂B))d−αd ≤
X
B∈B˜α
fBHd−1(∂B) d−αd
.α
X
Q∈Q˜α
fQHd−1(∂Q) d−αd
≤Cd,α(varf)d−αd
where l(Q)∼αr(B) and fQ ∼α fB so that alsocαQ∩cαP =∅, or l(P)<l(Q) andfP >2fQ.
15/16
2. Reduce to dyadic
X
B∈Beα
(fBHd−1(∂B))d−αd ≤
X
B∈B˜α
fBHd−1(∂B) d−αd
.α
X
Q∈Q˜α
fQHd−1(∂Q) d−αd
≤Cd,α(varf)d−αd
where l(Q)∼αr(B) and fQ ∼α fB so that alsocαQ∩cαP =∅, or l(P)<l(Q) andfP >2fQ.
2. Reduce to dyadic
X
B∈Beα
(fBHd−1(∂B))d−αd ≤
X
B∈B˜α
fBHd−1(∂B) d−αd
.α
X
Q∈Q˜α
fQHd−1(∂Q) d−αd
≤Cd,α(varf)d−αd
where l(Q)∼αr(B) and fQ ∼α fB so that alsocαQ∩cαP =∅, or l(P)<l(Q) andfP >2fQ.
16/16
Thank you