AAA Abstract and Applied Analysis 1687-0409 1085-3375 Hindawi Publishing Corporation 987214 10.1155/2014/987214 987214 Research Article Boundedness for a Class of Singular Integral Operators on Both Classical and Product Hardy Spaces Tan Chaoqiang Hudzik Henryk Department of Mathematics Shantou University Shantou 515063 China stu.edu.cn 2014 2722014 2014 10 09 2013 02 01 2014 27 2 2014 2014 Copyright © 2014 Chaoqiang Tan. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

We found that the classical Calderón-Zygmund singular integral operators are bounded on both the classical Hardy spaces and the product Hardy spaces. The purpose of this paper is to extend this result to a more general class. More precisely, we introduce a class of singular integral operators including the classical Calderón-Zygmund singular integral operators and show that they are bounded on both the classical Hardy spaces and the product Hardy spaces.

1. Introduction

The classical Hardy spaces and the product Hardy spaces play important roles in Harmonic analysis, which are due to the original work of Fefferman and Stein  and Gundy and Stein , respectively. It is well known that these two Hardy spaces are essentially different. For instance, see . It has been known that the classical Calderón-Zygmund singular integral operators are bounded on the classical Hardy spaces and the product singular integral operators are bounded on the product Hardy spaces. Surprisingly, in , we found that the classical Calderón-Zygmund singular integral operators are also bounded on the product Hardy spaces. More precisely, if T is a bounded operator on L 2 ( 2 ) with T f ( x ) = p · v · 𝒦 * f ( x ) , where the kernel 𝒦 C 2 ( 2 { 0 } ) and satisfies | x α 𝒦 ( x ) | C / | x | 2 + α for 0 | α | 2 and x 2 { 0 } , then T is bounded on both the classical Hardy spaces H p ( 2 ) and the product Hardy space H p ( × ) .

A natural question arises: weather there exist a more general class of operators that are bounded both on the classical Hardy spaces and the product Hardy spaces. The purpose of this paper is to answer this question. Now we first recall the definitions of the classical Hardy spaces H p ( 2 ) (see  for more details) and the product Hardy spaces H p ( × ) (see [2, 7] for more details).

We let 𝒮 1 be the set including all ψ 𝒮 ( 2 ) that satisfy 2 ψ ( x ) d x = 0 and j | ψ ^ ( 2 - j ξ ) | 2 = 1 for all ξ 2 { 0 } . And let 𝒮 2 be the set including all ψ ~ 𝒮 ( 2 ) that satisfy ψ ~ ( x 1 , x 2 ) d x 1 = ψ ~ ( x 1 , x 2 ) d x 2 = 0 and j , k | ψ ~ ^ ( 2 - j ξ 1 , 2 - k ξ 2 ) | 2 = 1 for all ξ = ( ξ 1 , ξ 2 ) 2 { 0 } .

Given an ψ 𝒮 1 , the Littlewood-Paley-Stein square function of f 𝒮 ( 2 ) is defined by g ψ ( f ) ( x ) = { j | ψ j * f ( x ) | 2 } 1 / 2 , where ψ j ( x ) = 2 2 j ψ ( 2 j x 1 , 2 j x 2 ) , x = ( x 1 , x 2 ) . And the discrete square function is defined by g ψ d ( f ) ( x ) = { j Q | ψ j * f ( c Q ) | 2 χ Q ( x ) } 1 / 2 , where Q are dyadic cubes in 2 with the side length l ( Q ) = 2 - j and the center c Q and χ Q is the characteristic function. It is well known that if 0 < p < , then g ψ ( f ) L p ( 2 ) g ψ d ( f ) L p ( 2 ) and, for different ψ , ϕ 𝒮 1 , g ψ ( f ) L p ( 2 ) g ϕ ( f ) L p ( 2 ) .

The classical Hardy space H p ( 2 ) is then defined by (1) H p ( 2 ) = { f 𝒮 𝒫 ( 2 ) : g ψ ( f ) ( x ) L p ( 2 ) } , where 𝒮 𝒫 denotes the space of distributions modulo polynomials. The H p ( 2 ) norm is defined by f H p ( 2 ) = g ψ ( f ) L p ( 2 ) .

Similarly, given a ψ ~ 𝒮 2 , the product square function of f 𝒮 ( 2 ) is defined by g ψ ~ ( f ) P ( x ) = { j , k | ψ ~ j , k * f ( x ) | 2 } 1 / 2 , where ψ ~ j , k ( x ) = 2 j + k ψ ~ ( 2 j x 1 , 2 k x 2 ) , x = ( x 1 , x 2 ) . And the discrete square function is defined by g ψ ~ d ( f ) P ( x 1 , x 2 ) = { j , k I , J | ψ ~ j , k * f ( c I , c J ) | 2 χ I ( x 1 ) χ I ( x 2 ) } 1 / 2 , where I , J are dyadic intervals in with the side length l ( I ) = 2 - j , l ( J ) = 2 - k and the center c I , c J , respectively. Also, for 0 < p < , g ψ ~ ( f ) P L p ( 2 ) g ψ ~ d ( f ) P L p ( 2 ) and, for different ψ ~ , ϕ ~ 𝒮 2 ,   g ψ ~ ( f ) P L p ( 2 ) g ϕ ~ ( f ) P L p ( 2 ) .

The product Hardy space H p ( × ) is then defined by (2) H p ( × ) = { f 𝒮 𝒫 ( 2 ) : g ψ ~ ( f ) P ( x ) L p ( 2 ) } . The H p ( × ) norm is defined by f H p ( × ) = g ψ ~ ( f ) P L p ( 2 ) .

The following theorem is our main result.

Theorem 1.

If 2 / 3 < p 1 , δ > 0 and T is an operator bounded on L 2 ( 2 ) with T f ( x ) = p · v · 𝒦 * f ( x ) , where the kernel 𝒦 C 2 ( 2 { ( 0 , ) ( , 0 ) } ) and satisfies | x 1 α x 2 β 𝒦 ( x 1 , x 2 ) | C ( 1 / | x 1 | 1 + α ) ( 1 / | x 2 | 1 + β ) ( ( | x 1 | / | x 2 | ) + ( | x 2 | / | x 1 | ) ) - δ for all | α | , | β | 1 , then T is bounded on both H p ( 2 ) and H p ( × ) .

Remark 2.

(I) In , we have shown that the operator T is bounded on L p ( 2 ) for all 1 < p < .

(II) It is easy to verify that the classical Calderón-Zygmund singular integral operators are contained in our class. Moreover, some more operators will be in our class. For example, the operator T = p · v · 𝒦 * f ( x ) with 𝒦 ( x 1 , x 2 ) = s g n ( x 1 + x 2 ) / ( | x 1 | | x 2 | ) 1 / 2 ( | x 1 | + | x 2 | ) .

Throughout this paper, we do the following conventions.

The notation A B means that C 1 A B C 2 A for some positive constants C 1 , C 2 .

If Q is a cube or interval, then we denote by c Q its center and by l ( Q ) its side length.

For j and a large positive integer N , we denote the set 𝒟 j ( 2 ) = { Q , where Q are dyadic cubes in 2 with side length l ( Q ) = 2 - j } and 𝒟 j N ( 2 ) = 𝒟 j + N ( 2 ) .

For j , k and a large positive integer N , we denote the set 𝒟 j , k ( 2 ) = { R = I × J , where I and J are dyadic intervals in with side length l ( I ) = 2 - j and l ( J ) = 2 - k , respectively} and 𝒟 j , k N ( 2 ) = 𝒟 j + N , k + N ( 2 ) .

j j means the minimum of j and j .

2. Proof of Theorem <xref ref-type="statement" rid="thm1.1">1</xref>

The first crucial tool in the proof of Theorem 1 is to apply the following discrete Calderón identity (see  for more details).

Lemma 3.

If 0 < p 1 , then consider the following.

( a ) Suppose that ϕ 𝒮 1 C 0 ( 2 ) with supp ( ϕ ) { x : | x | 1 } . Then for all f L 2 ( 2 ) H p ( 2 ) , there exist a large positive integer N (depending only on p ) and a function h L 2 ( 2 ) H p ( 2 ) such that f ( x ) = j Q 𝒟 j N ( 2 ) | Q | ϕ j ( x - c Q ) ( ϕ j * h ) ( c Q ) , where the series converges in L 2 ( 2 ) . Moreover, f L 2 ( 2 ) h L 2 ( 2 ) and f H p ( 2 ) h L 2 ( 2 ) .

( b ) Suppose that ϕ ~ 𝒮 2 C 0 ( 2 ) with supp ( ϕ ~ ) { x : | x | 1 } . Then for all f L 2 ( 2 ) H p ( × ) , there exist a large positive integer N (depending only on p ) and a function h L 2 ( 2 ) H p ( × ) such that f ( x ) = j , k R 𝒟 j , k N ( 2 ) | R | ϕ j , k ( x - c R ) ( ϕ j , k * h ) ( c R ) , where the series converges in L 2 ( 2 ) . Moreover, f H p ( 2 ) h L 2 ( 2 ) and f H p ( × ) h H p ( × ) P .

For the proof, we refer readers to .

The second crucial tool in the proof of Theorem 1 is the following orthogonal estimates.

Lemma 4.

Suppose that 0 < λ min ( δ , 1 / 2 ) and 𝒦 is the kernel as in Theorem 1; then

( a ) for ϕ C 0 ( 2 ) with 2 ϕ ( x ) d x = 0 , one has | 𝒦 * ϕ j ( x ) | C 2 2 j ( 1 / ( 1 + | 2 j x 1 | 1 + λ ) ) ( 1 / ( 1 + | 2 j x 2 | 1 + λ ) ) , for all x = ( x 1 , x 2 ) 2 and j , where C is a constant independent of j and x ;

( b ) for ϕ ~ C 0 ( 2 ) with ϕ ~ ( x 1 , x 2 ) d x 1 = ϕ ~ ( x 1 , x 2 ) d x 2 = 0 , one has | 𝒦 * ϕ ~ j , k ( x ) | C 2 j + k ( 1 / ( 1 + | 2 j x 1 | 1 + λ ) ) ( 1 / ( 1 + | 2 k x 2 | 1 + λ ) ) , for all x = ( x 1 , x 2 ) 2 and j , k , where C is a constant independent of j , k , and x .

Proof.

(a) Since 𝒦 is single-parameter dilation invariant, that is, for each δ > 0 , δ 2 𝒦 ( δ x 1 , δ x 2 ) satisfies the same hypotheses, with the same bounds as 𝒦 . We just need to show that | 𝒦 * ϕ ( x ) | C ( ( 1 / ( 1 + | x 1 | 1 + λ ) ) ) ( 1 / ( 1 + | x 2 | 1 + λ ) ) , for all x = ( x 1 , x 2 ) 2 . Without loss of generality, we may assume that supp ( ϕ ) { x : | x | 1 } . To get the required estimate, we will discuss it in the following three cases: (I) | x 1 | 2 , | x 2 | 2 ; (II) | x 1 | 2 , | x 2 | < 2 or | x 1 | < 2 , | x 2 | 2 ; (III) | x 1 | < 2 , | x 2 | < 2 .

For case (I), | x 1 | 2 , | x 2 | 2 , by the moment condition of ϕ , we have (3) | 𝒦 * ϕ ( x ) | = | 2 𝒦 ( x 1 - y 1 , x 2 - y 2 ) ϕ ( y 1 , y 2 ) d y 1 d y 2 | = | 2 ( 𝒦 ( x 1 - y 1 , x 2 - y 2 ) - 𝒦 ( x 1 , x 2 ) ) H H H h h H × ϕ ( y 1 , y 2 ) d y 1 d y 2 2 ( 𝒦 ( x 1 - y 1 , x 2 - y 2 ) - 𝒦 ( x 1 , x 2 ) ) 2 | C | 2 ( 1 | x 1 | 2 1 | x 2 | + 1 | x 1 | 1 | x 2 | 2 ) ( | x 1 | | x 2 | + | x 2 | | x 1 | ) - δ hhhhhhhhhh × | ϕ ( y 1 , y 2 ) | d y 1 d y 2 | 2 ( 1 | x 1 | 2 1 | x 2 | + 1 | x 1 | 1 | x 2 | 2 ) ( | x 1 | | x 2 | + | x 2 | | x 1 | ) - δ | C ( 1 | x 1 | 2 1 | x 2 | + 1 | x 1 | 1 | x 2 | 2 ) ( | x 1 | | x 2 | + | x 2 | | x 1 | ) - δ C 1 + | x 1 | 1 + λ 1 1 + | x 2 | 1 + λ .

For case (II), | x 1 | 2 , | x 2 | < 2 or | x 1 | < 2 , | x 2 | 2 , we have (4) | 𝒦 * ϕ ( x ) | 2 | 𝒦 ( x 1 - y 1 , x 2 - y 2 ) ϕ ( y 1 , y 2 ) | d y 1 d y 2 C | y 1 | 1 | y 2 | 1 1 | x 1 - y 1 | 1 | x 2 - y 2 | hhhhhhhh × ( | x 1 - y 1 | | x 2 - y 2 | + | x 2 - y 2 | | x 1 - x 2 | ) - δ d y 1 d y 2 C 1 + | x 1 | 1 + δ 1 1 + | x 2 | 1 + δ C 1 + | x 1 | 1 + λ 1 1 + | x 2 | 1 + λ .

For case (III), | x 1 | < 2 , | x 2 | < 2 , we let η C 0 ( 2 ) with 0 η ( x ) 1 and η ( x ) = 1 when | x | 4 and η ( x ) = 0 when | x | 8 . We have (5) | 𝒦 * ϕ ( x ) | = | 2 𝒦 ( y 1 , y 2 ) ϕ ( x 1 - y 1 , x 2 - y 2 ) η ( y 1 , y 2 ) d y 1 d y 2 | | 2 𝒦 ( y 1 , y 2 ) ( ϕ ( x 1 - y 1 , x 2 - y 2 ) - ϕ ( x 1 , x 2 ) ) h h h h h h × η ( y 1 , y 2 ) d y 1 d y 2 2 𝒦 ( y 1 , y 2 ) | m m l + | 2 𝒦 ( y 1 , y 2 ) ϕ ( x 1 , x 2 ) η ( y 1 , y 2 ) d y 1 d y 2 | C | | y 1 | 8 | y 2 | 8 1 | y 1 | 1 | y 2 | ( | y 2 | | y 1 | + | y 1 | | y 2 | ) - δ h h h h h h h × ( | y 1 | + | y 2 | ) d y 1 d y 2 | y 1 | 8 | y 2 | 8 | m m l + C | 2 𝒦 ~ ( ξ 1 , ξ 2 ) η ~ ( ξ 1 , ξ 2 ) d y 1 d y 2 | C C 1 + | x 1 | 1 + λ 1 1 + | x 2 | 1 + λ .

(b) Without loss of generality, we may assume that supp ( ϕ ~ ) { x : | x | 1 } . The required estimate will be discussed in the following four cases: (I) | x 1 | 2 - j + 1 , | x 2 | 2 - k + 1 ; (II) | x 1 | 2 - j + 1 , | x 2 | < 2 - k + 1 ; (III) | x 1 | < 2 - j + 1 , | x 2 | 2 - k + 1 ; (IV) | x 1 | < 2 - j + 1 , | x 2 | < 2 - k + 1 .

For case (I), | x 1 | 2 - j + 1 , | x 2 | 2 - k + 1 , by the moment condition of ϕ ~ , we have (6) | 𝒦 * ϕ ~ j , k ( x ) | = 2 j + k | 2 { ( 𝒦 ( x 1 - y 1 , x 2 - y 2 ) - 𝒦 ( x 1 , x 2 - y 2 ) ) hhhhhhhhhhh - ( 𝒦 ( x 1 - y 1 , x 2 ) - 𝒦 ( x 1 , x 2 ) ) } h h h h h h h h h h × ϕ ~ ( 2 j y 1 , 2 k y 2 ) d y 1 d y 2 2 | = 2 j + k | 2 x 2 x 2 - y 2 x 1 x 1 - y 1 z 1 1 z 2 1 𝒦 ( z 1 , z 2 ) d z 1 d z 2 h h h h h h h h h h h h h h h h h h h × ϕ ~ ( 2 j y 1 , 2 k y 2 ) d y 1 d y 2 x 2 x 2 - y 2 | C 2 j + k 2 1 | x 1 | 2 1 | x 2 | 2 ( | x 1 | | x 2 | + | x 2 | | x 1 | ) - δ h h h h h h h h h h × | y 1 | | y 2 | | ϕ ~ ( 2 j y 1 , 2 k y 2 ) | d y 1 d y 2 C 2 - j - k 1 | x 1 | 2 1 | x 2 | 2 C 2 j + k 1 1 + | 2 j x 1 | 1 + λ 1 1 + | 2 k x 2 | 1 + λ .

For case (II), | x 1 | 2 - j + 1 , | x 2 | < 2 - k + 1 , similarly, we have (7) | 𝒦 * ϕ ~ j , k ( x ) | = 2 j + k | 2 ( 𝒦 ( x 1 - y 1 , x 2 - y 2 ) - 𝒦 ( x 1 , x 2 - y 2 ) ) h h h h h h h h h h × ϕ ~ ( 2 j y 1 , 2 k y 2 ) d y 1 d y 2 2 | = 2 j + k | 2 x 1 x 1 - y 1 z 1 1 𝒦 ( z 1 , x 2 - y 2 ) d z 1 h h h h h h h h h h h h h × ϕ ~ ( 2 j y 1 , 2 k y 2 ) d y 1 d y 2 x 1 x 1 - y 1 | C 2 j + k | y 1 | 2 - j | y 2 | 2 - k 1 | x 1 | 2 1 | x 2 - y 2 | h h h h h h h h h h × ( | x 1 | | x 2 - y 2 | + | x 2 - y 2 | | x 1 | ) - δ | y 1 | d y 1 d y 2 C 2 - j + k 1 | x 1 | 2 C 2 j + k 1 1 + | 2 j x 1 | 1 + λ 1 1 + | 2 k x 2 | 1 + λ .

The cases (II) and (III) are symmetric, so case (III) follows.

For case (IV), | x 1 | < 2 - j + 1 , | x 2 | < 2 - k + 1 , we let θ C 0 ( ) with 0 θ ( x ) 1 and θ ( x ) = 1 when | x | 4 and θ ( x ) = 0 when | x | 8 . Then (8) | 𝒦 * ϕ ~ j , k ( x ) | = 2 j + k | 2 𝒦 ( y 1 , y 2 ) ϕ ~ ( 2 j ( x 1 - y 1 ) , 2 k ( x 2 - y 2 ) ) h h h h h h h h h × θ ( 2 j y 1 ) θ ( 2 k y 2 ) d y 1 d y 2 2 | 2 j + k | 2 𝒦 ( y 1 , y 2 ) h h h h h h h h h h × ( ϕ ~ ( 2 j ( x 1 - y 1 ) , 2 k ( x 2 - y 2 ) ) - ϕ ~ h h h h h h h h h h h h h × ( 2 j x 1 , 2 k x 2 ) ) θ ( 2 j y 1 ) θ ( 2 k y 2 ) d y 1 d y 2 2 | m m l + 2 j + k | 2 𝒦 ( y 1 , y 2 ) ϕ ~ ( 2 j x 1 , 2 k x 2 ) h h h h h h h h h h h × θ ( 2 j y 1 ) θ ( 2 k y 2 ) d y 1 d y 2 2 | C 2 j + k | y 1 | 2 - j + 3 | y 2 | 2 - k + 3 1 | y 1 | 1 | y 2 | ( | y 1 | | y 2 | + | y 2 | | y 1 | ) - δ h h h h h h h h h h h × ( | 2 j y 1 | + | 2 k y 2 | ) d y 1 d y 2 m m l + C | 2 𝒦 ^ ( ξ 1 , ξ 2 ) θ ^ ( 2 - j ξ 1 ) θ ^ ( 2 - k ξ 2 ) d ξ 1 d ξ 2 | C 2 j + k C 2 j + k 1 1 + | 2 j x 1 | 1 + λ 1 1 + | 2 k x 2 | 1 + λ .

This completes the proof of Lemma 4.

As a consequence of Lemma 4, we have the following.

Lemma 5.

(a) Under hypothesis (a) of Lemma 4, one has (9) | ϕ j * 𝒦 * ϕ j ( x ) | C 2 - | j - j | 2 j j 1 + | 2 j j x 1 | 1 + λ × 2 j j 1 + | 2 j j x 2 | 1 + λ , for all j , j and x = ( x 1 , x 2 ) 2 .

(b) Under hypothesis (b) of Lemma 4, one has (10) | ϕ ~ j , k * 𝒦 * ϕ ~ j , k ( x ) | C 2 - | j - j | 2 - | k - k | 2 j j 1 + | 2 j j x 1 | 1 + λ m m m l × 2 k k 1 + | 2 k k x 2 | 1 + λ , for all j , j , k and k and x = ( x 1 , x 2 ) 2 .

The proof of Lemma 5 is based on the following two observations: ( 1 ) convolution operation is commutative; that is, ϕ j * 𝒦 * ϕ j ( x ) = 𝒦 * ( ϕ j * ϕ j ) ( x ) (or ϕ ~ j , k * 𝒦 * ϕ ~ j , k ( x ) = 𝒦 * ( ϕ ~ j , k * ϕ ~ j , k ) ( x ) ); ( 2 )    ϕ j * ϕ j (or ϕ ~ j , k * ϕ ~ j , k ) satisfies the same estimate as ϕ j j (or ϕ ~ j j , k k ) in Lemma 4 with the bound C 2 - | j - j | (or C 2 - | j - j | 2 - | k - k | ). The details are left to the readers.

The last crucial tool in the proof of Theorem 1 is the following strongly maximal function estimates.

Lemma 6.

Suppose that λ > 0 , 2 / 3 < q 1 , N , j , j , k and k , and F L 2 ( 2 ) . Then consider the following.

( a ) If Q 𝒟 j ( 2 ) , u = ( u 1 , u 2 ) , and v = ( v 1 , v 2 ) Q , then one has (11) Q = I × J 𝒟 j N ( 2 ) 2 j j ( 1 + 2 j j | u 1 - c I | ) 1 + λ 2 j j ( 1 + 2 j j | u 2 - c J | ) 1 + λ hhhhhhhh × | F ( c Q ) | hhhhh C 2 2 ( j j ) ( 1 - 1 / q ) + 2 j / q hhhhhhhh × { M S [ ( Q 𝒟 j N ( 2 ) | F ( c Q ) | 2 χ Q ) q / 2 ] } 1 / q ( v ) , where M S is the strongly maximal operator.

( b ) If R = I × J 𝒟 j , k ( 2 ) , u = ( u 1 , u 2 ) , and v = ( v 1 , v 2 ) R , then one has (12) R = I × J 𝒟 j , k N ( 2 ) 2 j j ( 1 + 2 j j | u 1 - c I | ) 1 + λ 2 k k ( 1 + 2 k k | u 2 - c J | ) 1 + λ h h h h h h h h × | F ( c I , c J ) | hhhhh C 2 ( j j ) ( 1 - 1 / q ) + j / q 2 ( k k ) ( 1 - 1 / q ) + k / q h h h h h h h h × { M S [ ( R 𝒟 j , k N ( 2 ) | F ( c R ) | 2 χ R ) q / 2 ] } 1 / q ( v ) .

For the proof, we refer readers to .

Proof of Theorem <xref ref-type="statement" rid="thm1.1">1</xref>.

Firstly we show that T is bounded on the classical Hardy space H p ( 2 ) . Since L 2 ( 2 ) H p ( 2 ) is dense in H p ( 2 ) , we just need to show that, for all f L 2 ( 2 ) H p ( 2 ) , we have T f H p ( 2 ) C f H p ( 2 ) ; that is, for a fixed ψ 𝒮 1 , (13) g ψ ( T f ) L p ( 2 ) C f H p ( 2 ) .

Note that (14) g ψ ( T f ) ( x ) = { j Q = I × J 𝒟 j ( 2 ) | ψ j * 𝒦 * f ( c Q ) | 2 χ Q ( x ) } 1 / 2 .

For j , Q = I × J 𝒟 j ( 2 ) , and x Q , applying (a) of Lemma 3, we have (15) | ψ j * 𝒦 * f ( c Q ) | = | { j Q 𝒟 j N ( 2 ) | Q | ψ j * 𝒦 * ψ j ( u - c Q ) ( ψ j * h ) ( c Q ) } | .

By (a) of Lemma 5, we have (16) | ψ j * 𝒦 * f ( c Q ) | C j Q = I × J 𝒟 j ( 2 ) 2 - 2 j 2 - | j - j | hhhhhhhhhhhhhhh × 2 j j 1 + | 2 j j ( u 1 - c I ) | 1 + λ hhhhhhhhhhhhhhh × 2 j j 1 + | 2 j j ( u 2 - c J ) | 1 + λ ( ψ j * h ) ( c Q ) .

Applying (a) of Lemma 6 with F = ψ j * h and 2 / 3 < q < p , we have (17) | ψ j * 𝒦 * f ( c Q ) | C j 2 - 2 j 2 - | j - j | 2 2 ( j j ) ( 1 - 1 / q ) + 2 j / q m m l × { M S [ ( Q 𝒟 j N ( 2 ) | ψ j * h ( c Q ) | 2 χ Q ) q / 2 ] } 1 / q ( x ) .

Therefore, (18) | g ψ ( T f ) ( x ) | 2 = j Q 𝒟 j ( 2 ) | ψ j * 𝒦 * f ( c Q ) | 2 χ Q ( x ) C j Q 𝒟 j ( 2 ) | { M S [ ( Q 𝒟 j N ( 2 ) | ψ j * h ( c Q ) | 2 χ Q Q 𝒟 j N ( 2 ) ) q 2 ] } 1 q ( x ) | 2 j 2 - 2 j 2 - | j - j | 2 2 ( j j ) ( 1 - 1 / q ) + 2 j / q h h h h h h h h h h h h h h h h h h × { [ ( Q 𝒟 j N ( 2 ) | ψ j * h ( c Q ) | 2 χ Q ) q / 2 ] } 1 / q ( x ) | { M S [ ( Q 𝒟 j N ( 2 ) | ψ j * h ( c Q ) | 2 χ Q Q 𝒟 j N ( 2 ) ) q 2 ] } 1 q ( x ) | 2 j 2 - 2 j 2 - | j - j | 2 2 ( j j ) ( 1 - 1 / q ) + 2 j / q | 2 χ Q ( x ) C j ( j 2 - 2 j 2 - | j - j | 2 2 ( j j ) ( 1 - 1 / q ) + 2 j / q ) hhhhhh × ( j 2 - 2 j 2 - | j - j | 2 2 ( j j ) ( 1 - 1 / q ) + 2 j / q { M S [ ( Q 𝒟 j N ( 2 ) | ψ j * h ( c Q ) | 2 χ Q ) q / 2 ] } 2 / q ( x ) j ) C j j 2 - 2 j 2 - | j - j | 2 2 ( j j ) ( 1 - 1 / q ) + 2 j / q { M S [ ( Q 𝒟 j N ( 2 ) | ψ j * h ( c Q ) | 2 χ Q ) q / 2 ] } 2 / q ( x ) C j { M S [ ( Q 𝒟 j N ( 2 ) | ψ j * h ( c Q ) | 2 χ Q ) q / 2 ] } 2 / q ( x ) .

Applying Fefferman-Stein’s vector-valued strong maximal inequality (see  for more details) on L p / q ( 2 / q ) , we have (19) T ( f ) H p ( 2 ) = g ψ ( T f ) L p ( 2 ) C { j { M S [ ( Q 𝒟 j N ( 2 ) | ψ j * h ( c Q ) | 2 χ Q ) q / 2 ] } 2 / q ( x ) } 1 / 2 L p ( 2 ) C { j Q 𝒟 j N ( 2 ) | ψ j * h ( c Q ) | 2 χ Q ( x ) } 1 / 2 L p ( 2 ) C h H p ( 2 ) C f H p ( 2 ) .

The proof of T ’s boundedness on H p ( × ) is almost the same as above; that is, we just need to replace (a) of the required lemmas to (b). Here we omit the details.

Conflict of Interests

The author declares that there is no conflict of interests regarding the publication of this paper.

Acknowledgment

Chaoqiang Tan is supported by SRF for the Doctoral Program of Higher Education (Grant no. 20104402120002).

Fefferman C. Stein E. M. H p spaces of several variables Acta Mathematica 1972 129 3-4 137 193 MR0447953 10.1007/BF02392215 ZBL0257.46078 Gundy R. F. Stein E. M. H p theory for the poly-disc Proceedings of the National Academy of Sciences of the United States of America 1979 76 3 1026 1029 MR524328 10.1073/pnas.76.3.1026 ZBL0405.32002 Carleson L. A counterexample for measures bounded on Hp for the bi-disc Mittag Leffler Report 1974 7 Chang S. Y. Fefferman R. A continuous version of duality of H 1 with BMO on the bidisc Annals of Mathematics 1980 112 1 179 201 10.2307/1971324 MR584078 Chang S. Y. Fefferman R. Some recent developments in Fourier analysis and H p -theory on product domains Bulletin of the American Mathematical Society 1985 12 1 1 43 10.1090/S0273-0979-1985-15291-7 MR766959 Tan C. Q. Boundedness of classical Calderón-Zygmund convolution operators on product Hardy space Mathematical Research Letters 2013 20 3 591 599 10.4310/MRL.2013.v20.n3.a15 Fefferman R. Calderón-Zygmund theory for product domains: H p spaces Proceedings of the National Academy of Sciences of the United States of America 1986 83 4 840 843 10.1073/pnas.83.4.840 MR828217 Tan C. Q. Singular integral operators between the classical type and the product type Acta Scientiarum Naturalium Universitatis Sunyatseni 2012 51 5 58 62 MR3014683 ZBL1274.42035 Han Y. S. Lu G. Z. Zhao K. Discrete Calderón’s identity, atomic decomposition and boundedness criterion of operators on multiparameter hardy spaces Journal of Geometric Analysis 2010 20 3 670 689 Han Y.-S. Plancherel-Pólya type inequality on spaces of homogeneous type and its applications Proceedings of the American Mathematical Society 1998 126 11 3315 3327 10.1090/S0002-9939-98-04445-1 MR1459123 ZBL0920.42011 Stein E. M. Note on singular integrals Proceedings of the American Mathematical Society 1957 8 250 254 MR0088606 10.1090/S0002-9939-1957-0088606-8 ZBL0920.42011