## Weighted Composition Operators between different Bloch-type Spaces in Polydisk

### Zehua Zhou Department of Mathematics, Tianjin University Tianjin 300072, P. R. China E-mail: zehuazhou2003@yahoo.com.cn

$1$
Abstract
Let $\phi \left(z\right)=\left({\phi }_{1}\left(z\right),\dots ,{\phi }_{n}\left(z\right)\right)$  be a holomorphic self-map of ${U}^{n}$  and $\psi \left(z\right)$  a holomorphic function on ${U}^{n},$  where ${U}^{n}$  is the unit polydisk of ${\text{C}}^{n}.$  Let $p\ge 0,$  $q\ge 0$  , this paper gives some necessary and sufficient conditions for the weighted composition operator ${W}_{\psi ,\phi }$  induced by $\psi$  and $\phi$  to be bounded and compact between $p$  -Bloch space ${\mathcal{ℬ}}^{p}\left({U}^{n}\right)$  and $q$  -Bloch space ${\mathcal{ℬ}}^{q}\left({U}^{n}\right).$  Keywords Bloch-type space, Weighted composition operator, Composition operator.
2000 Mathematics Subject Classification 47B38, 47B33, 32A37, 32A30

1 Introduction

Let $\Omega$  be a bounded homogeneous domain in ${\text{C}}^{n}.$  The class of all holomorphic functions with domain $\Omega$  will be denoted by $H\left(\Omega \right).$  Let $\phi$  be a holomorphic self-map of $\Omega ,$  the composition operator ${C}_{\phi }$  induced by $\phi$  is defined by $\left({C}_{\phi }f\right)\left(z\right)=f\left(\phi \left(z\right)\right),$  for $z$  in $\Omega$  and $f\in H\left(\Omega \right)$  . If, in addition, $\psi$  is a holomorphic function defined on $\Omega ,$  the multiplication operator induced by $\psi$  is defined by ${M}_{\psi }f\left(z\right)=\psi \left(z\right)f\left(z\right),$  and weighted composition operators ${W}_{\psi ,\phi }$  induced by $\psi$  and $\phi$  is defined by $\left({W}_{\psi ,\phi }f\right)\left(z\right)=\psi \left(z\right)f\left(\phi \left(z\right)\right)$  for $z$  in $\Omega$  and $f\in H\left(\Omega \right).$  If let $\psi \equiv 1,$  then ${W}_{\psi ,\phi }={C}_{\phi },$  if let $\phi =Id$  , then ${W}_{\psi ,\phi }={M}_{\psi }.$  So we can regard weighted composition operator as a generalization of a multiplication operator and a composition operator. Let $K\left(z,z\right)$  be the Bergman kernel function of $\Omega$  , the Bergman metric ${H}_{z}\left(u,u\right)$  in $\Omega$  is defined by ${H}_{z}\left(u,u\right)=\frac{1}{2}{\sum }_{j,k=1}^{n}\frac{{\partial }^{2}logK\left(z,z\right)}{\partial {z}_{j}\partial {\overline{z}}_{k}}{u}_{j}{\overline{u}}_{k},$  where $z\in \Omega$  and $u=\left({u}_{1},\dots ,{u}_{n}\right)\in {\text{C}}^{n}.$  Following Timoney [1], we say that $f\in H\left(\Omega \right)$  is in the Bloch space $\mathcal{ℬ}\left(\Omega \right),$  if $\parallel f{\parallel }_{\mathcal{ℬ}\left(\Omega \right)}={sup}_{z\in \Omega }{Q}_{f}\left(z\right)<\infty ,$  where ${Q}_{f}\left(z\right)=sup\left\{\frac{|▽f\left(z\right)u|}{{H}_{z}^{\frac{1}{2}}\left(u,u\right)}:u\in {\text{C}}^{n}-\left\{0\right\}\right\},$  and $▽f\left(z\right)=\left(\frac{\partial f\left(z\right)}{\partial {z}_{1}},\dots ,\frac{\partial f\left(z\right)}{\partial {z}_{n}}\right),▽f\left(z\right)u={\sum }_{l=1}^{n}\frac{\partial f\left(z\right)}{\partial {z}_{l}}{u}_{l}.$  The little Bloch space ${\mathcal{ℬ}}_{0}\left(\Omega \right)$  is the closure in the Banach space $\mathcal{ℬ}\left(\Omega \right)$  of the polynomial functions.
Let $\partial \Omega$  denote the boundary of $\Omega$  . Following Timoney [2], for $\Omega ={B}_{n}$  the unit ball of ${\text{C}}^{n}$  , ${\mathcal{ℬ}}_{0}\left({B}_{n}\right)=\left\{f\in \mathcal{ℬ}\left({B}_{n}\right):{Q}_{f}\left(z\right)\to 0,\text{as}z\to \partial {B}_{n}\right\};$  for $\Omega =\mathcal{D}$  the bounded symmetric domain other than ${B}_{n}$  , $\left\{f\in \mathcal{ℬ}\left(\mathcal{D}\right):{Q}_{f}\left(z\right)\to 0,\text{as}z\to \partial \mathcal{D}\right\}$  isthe set of constant functions on $\mathcal{D}.$  So if $\mathcal{D}$  is a bounded symmetric domain other than the ball, we denote the ${\mathcal{ℬ}}_{0*}\left(\mathcal{D}\right)=\left\{f\in \mathcal{ℬ}\left(\mathcal{D}\right):{Q}_{f}\left(z\right)\to 0,\text{as}z\to {\partial }^{*}\mathcal{D}\right\}$  and also call it little Bloch space, here ${\partial }^{*}\mathcal{D}$  means the distinguished boundary of $\mathcal{D}$  .
The unit ball is the only bounded symmetric domain $\mathcal{D}$  with the property that ${\partial }^{*}\mathcal{D}=\partial \mathcal{D}.$  Let ${U}^{n}$  be the unit polydisk of ${\text{C}}^{n}.$  Timony [1] shows that $f\in \left({U}^{n}\right)$  if and only if $\parallel f{\parallel }_{=}|f\left(0\right)|+{sup}_{z\in {U}^{n}}{\sum }_{k=1}^{n}|\frac{\partial f}{\partial {z}_{k}}\left(z\right)|\left(1-|{z}_{k}{|}^{2}\right)<+\infty .$  This equality was the starting point for introducing the $p$  -Bloch spaces.
Let $p\ge 0,$  a function $f\in H\left({U}^{n}\right)$  is said to belong to the $p$  -Bloch space ${\mathcal{ℬ}}^{p}\left({U}^{n}\right)$  if $\parallel f{\parallel }_{p}=|f\left(0\right)|+{sup}_{z\in {U}^{n}}{\sum }_{k=1}^{n}|\frac{\partial f}{\partial {z}_{k}}\left(z\right)|{\left(1-|{z}_{k}{|}^{2}\right)}^{p}<+\infty .$  It is easy to show that ${\mathcal{ℬ}}^{p}\left({U}^{n}\right)$  is a Banach space with the norm $\parallel \cdot {\parallel }_{p}.$  It is easy to see that if ${lim}_{z\to \partial {U}^{n}}{\sum }_{k=1}^{n}|\frac{\partial f}{\partial {z}_{k}}\left(z\right)|\left(1-|{z}_{k}{|}^{2}{\right)}^{p}=0,$  then $f$  must be a constant. So, there is no sense to introduce the corresponding little $p$  -Bolch space in this way. We will say that the little $p$  -Bolch space ${\mathcal{ℬ}}_{0}^{p}\left({U}^{n}\right)$  is the closure of the polynomials in the $p$  -Bolch space.
In the recent years, there have been many papers focused on studying the composition operators in function spaces (say, for 1-dimensional case see [3-8], for n-dimensional case see [9-12]). More recently, S.Ohno, K. Stroethoff and R.H.Zhao in [8] discuss the weighted composition operators between-type spaces for 1-dimensional case.
In this paper, we discuss the boundedness and compactness of the weighted composition operators between $p$  -Bloch space and $q$  -Bloch space in polydisk, some new methods and techniques have been used because of the difference between topology boundary $\partial {U}^{n}$  and distinguished boundary ${\partial }^{*}{U}^{n}$  of ${U}^{n}$  , where $n>1$  , especially in the proof of Theorems 2 and 3 (need to be discussed according to the properties of the boundary). The results in this paper will extend corresponding results on the Bloch spaces (see [4-11]).
Our main results are the following:
Theorem 1 Let $\phi =\left({\phi }_{1},\dots ,{\phi }_{n}\right)$  be a holomorphic self-map of ${U}^{n}$  and $\psi \left(z\right)$  a holomorphic function of ${U}^{n},$  $p\ge 0,$  $q\ge 0.$  (i) If $p=1$  , then ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)⟶{\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is bounded if and only if
 $\begin{array}{c}{\sum }_{k,l=1}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left(z\right)|{\left(1-|{z}_{k}{|}^{2}\right)}^{q}ln\frac{4}{1-|{\phi }_{l}\left(z\right){|}^{2}}=O\left(1\right)\left(z\in {U}^{n}\right)\end{array}$ (1)
and
 $\begin{array}{c}|\psi \left(z\right)|{\sum }_{k,l=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{{\left(1-|{z}_{k}{|}^{2}\right)}^{q}}{1-|{\phi }_{l}\left(z\right){|}^{2}}=O\left(1\right)\left(z\in {U}^{n}\right).\end{array}$ (2)
(ii) If $0\le p<1$  , then ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)⟶{\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is bounded if and only if
 $\begin{array}{c}\psi \in {\mathcal{ℬ}}^{q}\left({U}^{n}\right)\end{array}$ (3)
and
 $\begin{array}{c}|\psi \left(z\right)|{\sum }_{k,l=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{{\left(1-|{z}_{k}{|}^{2}\right)}^{q}}{{\left(1-|{\phi }_{l}\left(z\right){|}^{2}\right)}^{p}}=O\left(1\right)\left(z\in {U}^{n}\right).\end{array}$ (4)
(iii) If $p>1$  , then ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)⟶{\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is bounded if and only if
 $\begin{array}{c}{\sum }_{k,l=1}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left(z\right)|\frac{{\left(1-|{z}_{k}{|}^{2}\right)}^{q}}{{\left(1-|{\phi }_{l}\left(z\right){|}^{2}\right)}^{p-1}}=O\left(1\right)\left(z\in {U}^{n}\right)\end{array}$ (5)
and
 $\begin{array}{c}|\psi \left(z\right)|{\sum }_{k,l=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{{\left(1-|{z}_{k}{|}^{2}\right)}^{q}}{{\left(1-|{\phi }_{l}\left(z\right){|}^{2}\right)}^{p}}=O\left(1\right)\left(z\in {U}^{n}\right).\end{array}$ (6)
Theorem 2 Let $\phi =\left({\phi }_{1},\dots ,{\phi }_{n}\right)$  be a holomorphic self-map of ${U}^{n}$  and $\psi \left(z\right)$  a holomorphic function of ${U}^{n}$  , $p\ge 0,q\ge 0.$  (i) If $p=1$  , then ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)⟶{\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is compact if and only if ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)⟶{\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is bounded and
 $\begin{array}{c}{\sum }_{k,l=1}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left(z\right)|{\left(1-|{z}_{k}{|}^{2}\right)}^{q}ln\frac{4}{1-|{\phi }_{l}\left(z\right){|}^{2}}=o\left(1\right)\left(\phi \left(z\right)\to \partial {U}^{n}\right),\end{array}$ (7)
and
 $\begin{array}{c}|\psi \left(z\right)|{\sum }_{k,l=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{{\left(1-|{z}_{k}{|}^{2}\right)}^{q}}{1-|{\phi }_{l}\left(z\right){|}^{2}}=o\left(1\right)\left(\phi \left(z\right)\to \partial {U}^{n}\right).\end{array}$ (8)
(ii) If $p>1$  , then ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)⟶{\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is compact if and only if ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)⟶{\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is bounded and
 $\begin{array}{c}{\sum }_{k,l=1}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left(z\right)|\frac{{\left(1-|{z}_{k}{|}^{2}\right)}^{q}}{{\left(1-|{\phi }_{l}\left(z\right){|}^{2}\right)}^{p-1}}=o\left(1\right)\left(\phi \left(z\right)\to \partial {U}^{n}\right),\end{array}$ (9)
and
 $\begin{array}{c}|\psi \left(z\right)|{\sum }_{k,l=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{{\left(1-|{z}_{k}{|}^{2}\right)}^{q}}{{\left(1-|{\phi }_{l}\left(z\right){|}^{2}\right)}^{p}}=o\left(1\right)\left(\phi \left(z\right)\to \partial {U}^{n}\right).\end{array}$ (10)
Remark 1 It is easy to show that if (1) or (5) or (7) or (9) holds, then $\psi \in {\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  and
 $\begin{array}{c}{\sum }_{k=1}^{n}|\frac{\partial \psi \left(z\right)}{\partial {z}_{k}}\left(z\right)|{\left(1-|{z}_{k}{|}^{2}\right)}^{q}=o\left(1\right)\left(\phi \left(z\right)\to \partial {U}^{n}\right).\end{array}$ (11)
In fact, (1) implies that ${\sum }_{k}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left(z\right)|{\left(1-|{z}_{k}{|}^{2}\right)}^{q}{\sum }_{l=1}^{n}ln\frac{4}{1-|{\phi }_{l}\left(z\right){|}^{2}}\le C$  for all $z\in {\text{C}}^{n}.$  The same reason for others.
Theorem 3 Let $\phi =\left({\phi }_{1},\dots ,{\phi }_{n}\right)$  be a holomorphic self-map of ${U}^{n}$  and $\psi \left(z\right)$  a holomorphic function of ${U}^{n}$  , $0\le p<1,$  $q\ge 0.$  Then ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)⟶{\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is compact if and only if ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)⟶{\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is bounded and
 $\begin{array}{c}{lim}_{|{\phi }_{l}\left(z\right)|\to 1}{\sum }_{k=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{\left(1-|{z}_{k}{|}^{2}{\right)}^{q}}{\left(1-|{\phi }_{l}\left(z\right){|}^{2}{\right)}^{p}}=0\end{array}$ (12)
for each $l\in \left\{1,...,n\right\}$  with $|{\phi }_{l}\left(z\right)|\to 1.$
Corollary 1 Let $\phi =\left({\phi }_{1},...,{\phi }_{n}\right)$  be a holomorphic self-map of ${U}^{n}$  , $p,q\ge 0.$  Then ${C}_{\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)\to {\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is bounded if and only if there exists a constant $C$  such that ${\sum }_{k,l=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{\left(1-|{z}_{k}{|}^{2}{\right)}^{q}}{\left(1-|{\phi }_{l}\left(z\right){|}^{2}{\right)}^{p}}\le C,$  for all $z\in {U}^{n}.$
Proof Let $\psi \left(z\right)\equiv 1,z\in {U}^{n}$  , then $\frac{\partial \psi }{\partial {z}_{k}}\left(z\right)=0$  for all $k\in \left\{1,2,\cdots ,n\right\}.$  It is clear that condition ( 1 ) in Theorem 1, condition ( 7 ) in Theorem 2 and condition ( 9 ) in Theorem 3 hold, note that ${W}_{\psi ,\phi }={C}_{\phi }$  , the Corollary follows by combining the above Theorems. Similar to Corollary 1, the following Corollary follows.
Corollary 2 Let $\phi =\left({\phi }_{1},...,{\phi }_{n}\right)$  be a holomorphic self-map of ${U}^{n}.$  If $p\ge 1$  and $q\ge 0,$  then ${C}_{\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)\to {\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is compact if and only if ${\sum }_{k,l=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{\left(1-|{z}_{k}{|}^{2}{\right)}^{q}}{\left(1-|{\phi }_{l}\left(z\right){|}^{2}{\right)}^{p}}\le C,$  for all $z\in {U}^{n},$  and ${\sum }_{k,l=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{\left(1-|{z}_{k}{|}^{2}{\right)}^{q}}{\left(1-|{\phi }_{l}\left(z\right){|}^{2}{\right)}^{p}}=o\left(1\right)\left(\phi \left(z\right)\to \partial {U}^{n}\right).$  If $0\le p<1\right)$  and $q\ge 0,$  Then ${C}_{\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)\to {\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is compact if and only if ${\sum }_{k,l=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{\left(1-|{z}_{k}{|}^{2}{\right)}^{q}}{\left(1-|{\phi }_{l}\left(z\right){|}^{2}{\right)}^{p}}\le C,$  for all $z\in {U}^{n},$  and ${lim}_{|{\phi }_{l}\left(z\right)|\to 1}{\sum }_{k=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{\left(1-|{z}_{k}{|}^{2}{\right)}^{q}}{\left(1-|{\phi }_{l}\left(z\right){|}^{2}{\right)}^{p}}=0$  for each $l\in \left\{1,...,n\right\}$  with ${\phi }_{l}\left(z\right)|\to 1.$
Remark 2 If let $\phi =id:{U}^{n}⟶{U}^{n},$  then we can obtain the corresponding results about multiplication operator ${M}_{\psi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)\to {\mathcal{ℬ}}^{q}\left({U}^{n}\right).$  Throughout the remainder of this paper $C$  will denote a positive constant, the exact value of which will vary from one appearance to the next.

2 Some Lemmas

Lemma 1 Let $f\in {\mathcal{ℬ}}^{p}\left({U}^{n}\right),$  (i) If $p=1,$  then $|f\left(z\right)|\le \left(\frac{1}{2}+\frac{1}{2nln2}\right)\left({\sum }_{l=1}^{n}ln\frac{4}{1-|{z}_{l}{|}^{2}}\right)\parallel f{\parallel }_{p};$  (ii) If $0\le p<1,$  then $|f\left(z\right)|\le \left(1+\frac{1}{1-p}\right)\parallel f{\parallel }_{p};$  (iii) If $p>1$  , then $|f\left(z\right)|\le \left(\frac{1}{n}+\frac{{2}^{p-1}}{p-1}\right)\left({\sum }_{l=1}^{n}\frac{1}{\left(1-|{z}_{l}{|}^{2}{\right)}^{p-1}}\right)\parallel f{\parallel }_{p}.$
Proof This Lemma can be proved by some integral estimates (if necessary, the proof can be omitted).
By the definition of $\parallel .{\parallel }_{p}$  , $|f\left(0\right)|\le \parallel f{\parallel }_{p},|\frac{\partial f\left(z\right)}{\partial {z}_{l}}|\le \frac{\parallel f{\parallel }_{p}}{\left(1-|{z}_{l}{|}^{2}{\right)}^{p}}\left(l\in \left\{1,2,\cdots ,n\right\}\right)$  and
 $\begin{array}{ccc}& & f\left(z\right)-f\left(0\right)={\int }_{0}^{1}\frac{df\left(tz\right)}{dt}dt={\sum }_{l=1}^{n}{\int }_{0}^{1}{z}_{l}\frac{\partial f}{\partial {\zeta }_{l}}\left(tz\right)dt,\end{array}$
So
 $\begin{array}{ccc}& & |f\left(z\right)|\le |f\left(0\right)|+{\sum }_{l=1}^{n}|{z}_{l}|{\int }_{0}^{1}\frac{\parallel f{\parallel }_{p}}{{\left(1-{t}^{2}|{z}_{l}{|}^{2}\right)}^{p}}dt\end{array}$
 $\begin{array}{ccc}& & \le \parallel f{\parallel }_{p}+\parallel f{\parallel }_{p}{\sum }_{l=1}^{n}{\int }_{0}^{|{z}_{l}|}\frac{1}{{\left(1-{t}^{2}\right)}^{p}}dt.\end{array}$ (13)
If $p=1,$
 $\begin{array}{c}{\int }_{0}^{|{z}_{l}|}\frac{1}{{\left(1-{t}^{2}\right)}^{p}}dt=\frac{1}{2}ln\frac{1+|{z}_{l}|}{1-|{z}_{l}|}\le \frac{1}{2}ln\frac{4}{1-|{z}_{l}{|}^{2}}.\end{array}$ (14)
It is clear that $ln\frac{4}{1-|{z}_{l}{|}^{2}}>ln4=2ln2,$  so
 $\begin{array}{c}1\le \frac{1}{2ln2}ln\frac{4}{1-|{z}_{l}{|}^{2}}\le \frac{1}{2nln2}{\sum }_{l=1}^{n}ln\frac{4}{1-|{z}_{l}{|}^{2}}.\end{array}$ (15)
Combining ( 13 ),( 14 ) and ( 15 ), we get $|f\left(z\right)|\le \left(\frac{1}{2}+\frac{1}{2nln2}\right)\left({\sum }_{l=1}^{n}ln\frac{4}{1-|{z}_{l}{|}^{2}}\right)\parallel f{\parallel }_{p}.$  If $p\ne 1,$
 $\begin{array}{ccc}& & {\int }_{0}^{|{z}_{l}|}\frac{1}{{\left(1-{t}^{2}\right)}^{p}}dt={\int }_{0}^{|{z}_{l}|}\frac{1}{\left(1-t{\right)}^{p}}\cdot \frac{1}{\left(1+t{\right)}^{p}}dt\end{array}$
 $\begin{array}{ccc}& & \le {\int }_{0}^{|{z}_{l}|}\frac{1}{\left(1-t{\right)}^{p}}dt=\frac{1-\left(1-|{z}_{l}|{\right)}^{-p+1}}{1-p}.\end{array}$ (16)
If $0\le p<1,$  ( 16 ) gives that ${\int }_{0}^{|{z}_{l}|}\frac{1}{{\left(1-{t}^{2}\right)}^{p}}dt\le \frac{1}{1-p},$  it follows from ( 13 ) that $|f\left(z\right)|\le \left(1+\frac{n}{1-p}\right)\parallel f{\parallel }_{p}.$  If $p>1,$  ( 16 ) gives that ${\int }_{0}^{|{z}_{l}|}\frac{1}{{\left(1-{t}^{2}\right)}^{p}}dt\le \frac{1-\left(1-|{z}_{l}{|}^{p-1}\right)}{\left(p-1\right)\left(1-|{z}_{l}|{\right)}^{p-1}}\le \frac{{2}^{p-1}}{\left(p-1\right)\left(1-|{z}_{l}^{2}|{\right)}^{p-1}},$  it follows from ( 13 ) that
 $\begin{array}{ccc}|f\left(z\right)|& \le & \parallel f{\parallel }_{p}+\frac{{2}^{p-1}}{p-1}\left({\sum }_{l=1}^{n}\frac{1}{\left(1-|{z}_{l}{|}^{2}{\right)}^{p-1}}\right)\parallel f{\parallel }_{p}\end{array}$
 $\begin{array}{ccc}& \le & \left(\frac{1}{n}+\frac{{2}^{p-1}}{p-1}\right)\left({\sum }_{l=1}^{n}\frac{1}{\left(1-|{z}_{l}{|}^{2}{\right)}^{p-1}}\right)\parallel f{\parallel }_{p}.\end{array}$
Now the Lemma is proved.
Lemma 2 Let $\phi$  be a holomorphic self-map of ${U}^{n}$  and $\psi \left(z\right)$  a holomorphic function on ${U}^{n},$  then ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)\to {\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is compact if and only if for any bounded sequence $\left\{{f}_{j}\right\}$  in ${\mathcal{ℬ}}^{p}\left({U}^{n}\right)$  which converges to zero uniformly on compact subsets of ${U}^{n}$  , we have $\parallel {W}_{\psi ,\phi }{f}_{j}{\parallel }_{q}\to 0,$  as $j\to \infty .$
Proof Assume that ${W}_{\psi ,\phi }$  is compact and suppose $\left\{{f}_{j}\right\}$  is a sequence in ${\mathcal{ℬ}}^{p}\left({U}^{n}\right)$  with ${sup}_{j\in \mathbf{N}}\parallel {f}_{j}{\parallel }_{p}<\infty$  and ${f}_{j}\to 0$  uniformly on compact subsets of ${U}^{n}.$  By the compactness of ${W}_{\psi ,\phi }$  we have that ${W}_{\psi ,\phi }\left({f}_{j}\right)=\psi {f}_{j}\circ \phi$  has a subsequence $\psi {f}_{{j}_{m}}\circ \phi$  which converges in ${\mathcal{ℬ}}^{q},$  say, to $g.$  By Lemma 1 we have that for any compact $K\subset {U}^{n}$  there is a positive constant ${C}_{K}$  independent of $f$  such that $|\psi \left(z\right){f}_{{j}_{m}}\left(\phi \left(z\right)\right)-g\left(z\right)|\le {C}_{K}\parallel \psi {f}_{{j}_{m}}\circ \phi -g{\parallel }_{q}$  for all $z\in K$  . This implies that $\psi \left(z\right){f}_{{j}_{m}}\left(\phi \left(z\right)\right)-g\left(z\right)\to 0$  uniformly on compacts of ${U}^{n}.$  Since $K$  is a compact subset of ${U}^{n},$  $|\psi \left(z\right)|\le C$  for all $z\in K$  and $\phi \left(K\right)$  isalso a compact subset of ${U}^{n},$  by the hypothesis, $|\psi \left(z\right){f}_{{j}_{m}}\left(\phi \left(z\right)\right)|\le C|{f}_{{j}_{m}}\left(\phi \left(z\right)\right)|$  converges to zero uniformly on $K$  . It follows from the arbitrary of $K$  that the limit function $g$  is equal to $0.$  Since it is true for arbitrary subsequence of $\left\{{f}_{j}\right\}$  we see that ${W}_{\psi ,\phi }{f}_{j}\to 0$  in ${\mathcal{ℬ}}^{q}.$  Conversely, $\left\{{g}_{j}\right\}$  be any sequence in the ball ${\mathcal{K}}_{M}={B}_{{\mathcal{ℬ}}^{p}}\left(0,M\right)$  of the space ${\mathcal{ℬ}}^{p}\left({U}^{n}\right).$  Since $\parallel {g}_{j}{\parallel }_{p}\le M<\infty ,$  by Lemma 1, $\left\{{g}_{j}\right\}$  is uniformly bounded on compact subsets of ${U}^{n}$  and hence normal by Montel's theorem. Hence we may extract a subsequence $\left\{{g}_{{j}_{m}}\right\}$  which converges uniformly on compact subsets of ${U}^{n}$  to some $g\in H\left({U}^{n}\right).$  It follows that $\frac{\partial {g}_{{j}_{m}}}{\partial {z}_{l}}\to \frac{\partial g}{\partial {z}_{l}}$  uniformly on compacts, for each $l\in \left\{1,...,n\right\},$  which implies $g\in {\mathcal{ℬ}}^{p}\left({U}^{n}\right)$  and $\parallel g{\parallel }_{p}\le M.$  Hence the sequence $\left\{{g}_{{j}_{m}}-g\right\}$  is such that $\parallel {g}_{{j}_{m}}-g{\parallel }_{p}\le 2M<\infty ,$  and converges to $0$  on compact subsets of ${U}^{n},$  by the hypothesis and Lemma 2, we have that $\psi {g}_{{j}_{m}}\circ \phi \to \psi g\circ \phi$  in ${\mathcal{ℬ}}^{q}.$  Thus the set ${W}_{\psi ,\phi }\left({\mathcal{K}}_{M}\right)$  is relatively compact, finishing the proof.
Lemma 3 Let $0\le p<1.$  If $f\in {\mathcal{ℬ}}^{p}\left({U}^{n}\right)$  , then $|f\left(z\right)-f\left(w\right)|\le \frac{2}{1-p}\parallel f{\parallel }_{p}{\sum }_{k=1}^{n}|{z}_{k}-{w}_{k}{|}^{1-p},z,w\in {U}^{n}.$
Proof For any $z=\left({z}_{1},{z}_{2},\cdots ,{z}_{n}\right),$  $w=\left({w}_{1},{w}_{2},\cdots ,{w}_{n}\right)\in {U}^{n},$  then $tz+\left(1-t\right)w\in {U}^{n}$  for $t\in \left[0,1\right].$  Denote $F\left(t\right)=f\left(tz+\left(1-t\right)w\right),$  then
 $\begin{array}{ccc}& & f\left(z\right)-f\left(w\right)=F\left(1\right)-F\left(0\right)={\int }_{0}^{1}{F}^{\prime }\left(t\right)dt={\int }_{0}^{1}\frac{d}{dt}\left(f\left(tz+\left(1-t\right)w\right)dt\end{array}$
 $\begin{array}{ccc}& & ={\sum }_{k=1}^{n}\left({z}_{k}-{w}_{k}\right){\int }_{0}^{1}\frac{\partial f}{\partial {\zeta }_{k}}\left(tz+\left(1-t\right)w\right)dt.\end{array}$ (17)
$f\in {\mathcal{ℬ}}^{p}\left({U}^{n}\right),$  so for $k\in \left\{1,2,\cdots ,n\right\}$  $\left(1-|{\zeta }_{k}{|}^{2}{\right)}^{p}|\frac{\partial f}{\partial {\zeta }_{k}}\left(\zeta \right)|\le \parallel f{\parallel }_{p}$  $|\frac{\partial f}{\partial {\zeta }_{k}}|\le \parallel f{\parallel }_{p}\left(1-|{\zeta }_{k}{|}^{2}{\right)}^{-p}\le \parallel f{\parallel }_{p}\left(1-|{\zeta }_{k}|{\right)}^{-p}.$  Let ${\zeta }_{k}=t{z}_{k}+\left(1-t\right){w}_{k}={w}_{k}+t\left({z}_{k}-{w}_{k}\right),$  note that if $|a|<1,|b|<1$  and $|a+b|<1$  then $|a|+|b|+|a+b|<2,$  so $1-|{\zeta }_{k}|=1-|{w}_{k}+t\left({z}_{k}-{w}_{k}\right)|\ge |1-|{w}_{k}|-t|{z}_{k}-{w}_{k}||,$
 $\begin{array}{c}|\frac{\partial f}{\partial {\zeta }_{k}}\left(tz+\left(1-t\right)w\right)|\le \parallel f{\parallel }_{p}\left(1-|{\zeta }_{k}{|}^{2}{\right)}^{-p}\le \parallel f{\parallel }_{p}{|1-|{w}_{k}|-t|{z}_{k}-{w}_{k}||}^{-p}.\end{array}$ (18)
It follows from ( 17 ) and ( 18 ) that
 $\begin{array}{ccc}& & |f\left(z\right)-f\left(w\right)|\le \parallel f{\parallel }_{p}{\sum }_{k=1}^{n}|{z}_{k}-{w}_{k}|{\int }_{0}^{1}{|1-|{w}_{k}|-t|{z}_{k}-{w}_{k}||}^{-p}dt\end{array}$
 $\begin{array}{ccc}& & =\parallel f{\parallel }_{p}{\sum }_{k=1}^{n}{\int }_{0}^{|{z}_{k}-{w}_{k}|}{|1-|{w}_{k}|-t|}^{-p}dt.\end{array}$ (19)
If $1-|{w}_{k}|\le |{z}_{k}-{w}_{k}|,$  then
 $\begin{array}{ccc}& & {\int }_{0}^{|{z}_{k}-{w}_{k}|}{|1-|{w}_{k}|-t|}^{-p}dt\end{array}$
 $\begin{array}{ccc}& & ={\int }_{0}^{1-|{w}_{k}|}\left(1-|{w}_{k}|-t{\right)}^{-p}dt+{\int }_{1-|{w}_{k}|}^{|{z}_{k}-{w}_{k}|}{\left(t-\left(1-|{w}_{k}|\right)\right)}^{-p}dt\end{array}$
 $\begin{array}{ccc}& & =\frac{1}{1-p}\left(\left(1-|{w}_{k}|{\right)}^{1-p}+\left(|{z}_{k}-{w}_{k}|-\left(1-|{w}_{k}|\right){\right)}^{1-p}\right)\end{array}$
 $\begin{array}{ccc}& & \le \frac{2}{1-p}\left(|{z}_{k}-{w}_{k}|{\right)}^{1-p}.\end{array}$ (20)
If $1-|{w}_{k}|>|{z}_{k}-{w}_{k}|,$  let $a=1-|{w}_{k}|,b=|{z}_{k}-{w}_{k}|,a>b$  and note that the fundamental inequality ${a}^{1-p}<\left(a-b{\right)}^{1-p}+{b}^{1-p},$  where $0\le p<1,$  so
 $\begin{array}{ccc}& & {\int }_{0}^{|{z}_{k}-{w}_{k}|}{|1-|{w}_{k}|-t|}^{-p}dt={\int }_{0}^{|{z}_{k}-{w}_{k}|}{\left(1-|{w}_{k}|-t\right)}^{-p}dt\end{array}$
 $\begin{array}{ccc}& & =\frac{1}{1-p}\left(\left(1-|{w}_{k}|{\right)}^{1-p}-\left(1-|{w}_{k}|-|{z}_{k}-{w}_{k}|{\right)}^{1-p}\right)\end{array}$
 $\begin{array}{ccc}& & <\frac{1}{1-p}|{z}_{k}-{w}_{k}{|}^{1-p}.\end{array}$ (21)
Combining  20  21  $\left(\right),\left(\right)$  and  19  $\left(\right)$  , this Lemma is proved.
Lemma 4 Let $0\le p<1.$  Every norm-bounded sequence in ${\mathcal{ℬ}}^{p}\left({U}^{n}\right)$  has a subsequence that converges uniformly on $\overline{{U}^{n}}$  .
Proof Let $\left\{{f}_{j}\right\}$  be a sequence in ${\mathcal{ℬ}}^{p}\left({U}^{n}\right),$  and $\parallel {f}_{j}{\parallel }_{p}\le C$  for $j=1,2,\cdots .$  It follows from Lemma 3 that $|{f}_{j}\left(z\right)-{f}_{j}\left(w\right)|\le C{\sum }_{k=1}^{n}|{z}_{k}-{w}_{k}{|}^{1-p}$  for $z,w\in {U}^{n}.$  Thus the family $\left\{{f}_{j}:n=1,2,\cdots \right\}$  is equicontinuous. By Lemma 1, $|{f}_{j}\left(z\right)|\le C\parallel {f}_{j}{\parallel }_{p}\le C,$  the family $\left\{{f}_{j}:n=1,2,\cdots \right\}$  is bounded uniformly on $\overline{{U}^{n}}$  . The statement of the lemma now follows from the Arzela-Ascoli Theorem.
Lemma 5 Let $p\in \left[0,1\right)$  and $\left\{{f}_{j}\right\}$  be a norm bounded sequence in ${\mathcal{ℬ}}^{p}\left({U}^{n}\right)$  which converges to zero uniformly on compacts of ${U}^{n}.$  Then it converges to zero uniformly on $\overline{{U}^{n}}.$
Proof Let $\varepsilon \in \left(0,1\right),$  $r\in \left(1-\varepsilon ,1\right),$  and $w\in r{U}^{n},$  then from Lemma 3 we have
 $\begin{array}{ccc}& & |{f}_{j}\left(z\right)|\le |{f}_{j}\left(w\right)|+C\parallel {f}_{j}{\parallel }_{p}{\sum }_{k=1}^{n}|{z}_{k}-{w}_{k}{|}^{1-p}\end{array}$
 $\begin{array}{ccc}& & \le |{f}_{j}\left(w\right)|+C{\sum }_{k=1}^{n}|{z}_{k}-{w}_{k}{|}^{1-p},z\in {U}^{n}.\end{array}$
Since ${sup}_{w\in r{U}^{n}}|{f}_{j}\left(w\right)|\to 0$  and since for each $z\in \overline{{U}^{n}}\r{U}^{n},$  $1-\varepsilon   there exists $\eta$  such that $1-\varepsilon <\eta   Choosing ${w}_{z}=\left(\eta {z}_{1},\eta {z}_{2},\cdots ,\eta {z}_{n}\right),$  then ${w}_{z}\in r{U}^{n}$  and $|{z}_{k}-\left({w}_{z}{\right)}_{k}{|}^{1-p}=|{z}_{k}-\eta {z}_{k}{|}^{1-p}=\left(1-\eta {\right)}^{1-p}|{z}_{k}|<{\varepsilon }^{1-p}\le \varepsilon .$  So ${lim}_{k\to \infty }{sup}_{z\in \overline{{U}^{n}}}|{f}_{j}\left(z\right)|\le C\varepsilon ,$  from which the result follows.
Combining Lemma 2 and Lemma 5, we can obtain the following Lemma at once.
Lemma 6 Let $\phi$  be a holomorphic self-map of ${U}^{n}$  and $\psi \left(z\right)$  a holomorphic function on ${U}^{n}.$  Let $0\le p<1$  and $q\ge 0,$  then ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)\to {\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is compact if and only if for any bounded sequence $\left\{{f}_{j}\right\}$  in ${\mathcal{ℬ}}^{p}\left({U}^{n}\right)$  which converges to zero uniformly on $\overline{{U}^{n}}$  we have $\parallel {W}_{\psi ,\phi }{f}_{j}{\parallel }_{q}\to 0,$  as $j\to \infty .$

3 The Proof of Theorem 1

First shows the sufficiency. If $f\in {\mathcal{ℬ}}^{p}\left({U}^{n}\right),$  a direct calculation gives
 $\begin{array}{ccc}& & {\sum }_{k=1}^{n}|\frac{\partial \left({W}_{\psi ,\phi }f\left(z\right)\right)}{\partial {z}_{k}}|\left(1-|{z}_{k}{|}^{2}{\right)}^{q}\end{array}$
 $\begin{array}{ccc}& & ={\sum }_{k=1}^{n}|\frac{\partial \psi \left(z\right)}{\partial {z}_{k}}f\left(\phi \left(z\right)\right)+\psi \left(z\right){\sum }_{l=1}^{n}\frac{\partial f}{\partial {w}_{l}}\left(\phi \left(z\right)\right)\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\left(1-|{z}_{k}{|}^{2}{\right)}^{q}\end{array}$ (22)
 $\begin{array}{ccc}& & \le {\sum }_{k=1}^{n}|\frac{\partial \psi \left(z\right)}{\partial {z}_{k}}||f\left(\phi \left(z\right)\right)|\left(1-|{z}_{k}{|}^{2}{\right)}^{q}\end{array}$
 $\begin{array}{ccc}& & +|\psi \left(z\right)|{\sum }_{k,l=1}^{n}|\frac{\partial f}{\partial {w}_{l}}\left(\phi \left(z\right)\right)\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\left(1-|{z}_{k}{|}^{2}{\right)}^{q}.\end{array}$ (23)
By Lemma 1, if $p=1$  , ( 22 ) gives that
 $\begin{array}{ccc}& & {\sum }_{k=1}^{n}|\frac{\partial \left({W}_{\psi ,\phi }f\left(z\right)\right)}{\partial {z}_{k}}|\left(1-|{z}_{k}{|}^{2}{\right)}^{q}\end{array}$
 $\begin{array}{ccc}& & \le C\left({\sum }_{k,l=1}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left(z\right)|{\left(1-|{z}_{k}{|}^{2}\right)}^{q}ln\frac{4}{1-|{\phi }_{l}\left(z\right){|}^{2}}\end{array}$
 $\begin{array}{ccc}& & +|\psi \left(z\right)|{\sum }_{k,l=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{{\left(1-|{z}_{k}{|}^{2}\right)}^{q}}{1-|{\phi }_{l}\left(z\right){|}^{2}}\right)\parallel f{\parallel }_{p}.\end{array}$ (24)
Combining ( 1 ), ( 2 ) and ( 24 ), we know ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)⟶{\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is bounded.
If $0\le p<1$  , by Lemma 1, ( 23 ) gives that
 $\begin{array}{ccc}& & {\sum }_{k=1}^{n}|\frac{\partial \left({W}_{\psi ,\phi }f\left(z\right)\right)}{\partial {z}_{k}}|\left(1-|{z}_{k}{|}^{2}{\right)}^{q}\end{array}$
 $\begin{array}{ccc}& & \le C\left({\sum }_{k,l=1}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left(z\right)|{\left(1-|{z}_{k}{|}^{2}\right)}^{q}\end{array}$
 $\begin{array}{ccc}& & +|\psi \left(z\right)|{\sum }_{k,l=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{{\left(1-|{z}_{k}{|}^{2}\right)}^{q}}{{\left(1-|{\phi }_{l}\left(z\right){|}^{2}\right)}^{p}}\right)\parallel f{\parallel }_{p}.\end{array}$ (25)
Combining ( 3 ), ( 4 ), Lemma 1 and ( 25 ), we know ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)⟶{\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is bounded.
If $p>1,$  by Lemma 1, ( 23 ) gives that
 $\begin{array}{ccc}& & {\sum }_{k=1}^{n}|\frac{\partial \left({W}_{\psi ,\phi }f\left(z\right)\right)}{\partial {z}_{k}}|\left(1-|{z}_{k}{|}^{2}{\right)}^{q}\end{array}$
 $\begin{array}{ccc}& & \le C\left({\sum }_{k,l=1}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left(z\right)|\frac{{\left(1-|{z}_{k}{|}^{2}\right)}^{q}}{{\left(1-|{\phi }_{l}\left(z\right){|}^{2}\right)}^{p-1}}\end{array}$
 $\begin{array}{ccc}& & +|\psi \left(z\right)|{\sum }_{k,l=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{{\left(1-|{z}_{k}{|}^{2}\right)}^{q}}{{\left(1-|{\phi }_{l}\left(z\right){|}^{2}\right)}^{p}}\right)\parallel f{\parallel }_{p}.\end{array}$ (26)
Combining ( 5 ), ( 6 ) and ( 26 ), we know ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)⟶{\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is bounded.
To show the necessity, assume that ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)⟶{\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is bounded, with
 $\begin{array}{c}\parallel {W}_{\psi ,\phi }f{\parallel }_{q}\le C\parallel f{\parallel }_{p}\end{array}$ (27)
for all $f\in {\mathcal{ℬ}}^{p}\left({U}^{n}\right).$  It is clear that $\psi \in {\mathcal{ℬ}}^{q}\left({U}^{n}\right),$  that is, ( 3 ) holds.
If $p=1$  , For fixed $l\left(1\le l\le n\right),$  we will make use of a family of test functions $\left\{{f}_{w}:w\in \text{C},|w|<1\right\}$  in ${\mathcal{ℬ}}^{p}\left({U}^{n}\right)$  defined as follows: let ${f}_{w}\left(z\right)=\frac{{z}_{l}}{1-\overline{w}{z}_{l}},$  then $\frac{\partial {f}_{w}\left(z\right)}{\partial {z}_{l}}=\frac{1}{\left(1-\overline{w}{z}_{l}{\right)}^{2}},\frac{\partial {f}_{w}\left(z\right)}{\partial {z}_{k}}=0\left(k\ne l\right)$  , $\parallel {f}_{w}{\parallel }_{p}\le \frac{4}{1-|w{|}^{2}}.$  It follows from ( 22 ) and ( 27 ) that ${\sum }_{k=1}^{n}|\frac{\partial \psi \left(z\right)}{\partial {z}_{k}}\frac{{\phi }_{l}\left(z\right)}{1-\overline{w}{\phi }_{l}\left(z\right)}+\psi \left(z\right)\frac{1}{\left(1-\overline{w}{\phi }_{l}\left(z\right){\right)}^{2}}\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\left(1-|{z}_{k}{|}^{2}{\right)}^{q}\le C\frac{1}{1-|w{|}^{2}},$  let $w={\phi }_{l}\left(z\right),$  then
 $\begin{array}{c}|\psi \left(z\right)|{\sum }_{k=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{{\left(1-|{z}_{k}{|}^{2}\right)}^{q}}{1-|{\phi }_{l}\left(z\right){|}^{2}}\le C+\parallel \psi {\parallel }_{q}\le C.\end{array}$ (28)
So for any $z\in {U}^{n}$  , $|\psi \left(z\right)|{\sum }_{k,l=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{{\left(1-|{z}_{k}{|}^{2}\right)}^{q}}{1-|{\phi }_{l}\left(z\right){|}^{2}}\le C,$  that is, ( 2 ) is valid.
If we set ${g}_{w}\left(z\right)=ln\frac{4}{1-\overline{w}{z}_{l}},$  then $\frac{\partial {g}_{w}\left(z\right)}{\partial {z}_{l}}=\frac{\overline{w}}{1-\overline{w}{z}_{l}},\frac{\partial {g}_{w}\left(z\right)}{\partial {z}_{k}}=0\left(k\ne l\right)$  , $\parallel {g}_{w}{\parallel }_{p}\le 2+ln4.$  It follows from ( 22 ) and ( 27 ) that ${\sum }_{k=1}^{n}|\frac{\partial \psi \left(z\right)}{\partial {z}_{k}}ln\frac{4}{1-\overline{w}{\phi }_{l}\left(z\right)}+\psi \left(z\right)\frac{\overline{w}}{1-\overline{w}{\phi }_{l}\left(z\right)}\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\left(1-|{z}_{k}{|}^{2}{\right)}^{q}\le C,$  let $w={\phi }_{l}\left(z\right)$  , ( 1 ) is valid. If $0\le p<1$  , $\psi ={W}_{\psi ,\phi }1\in {\mathcal{ℬ}}^{q}\left({U}^{n}\right),$  (3) is valid. For fixed $l\left(1\le l\le n\right),$  we will make use of a family of test functions $\left\{{f}_{w}:w\in \text{C}-\left\{0\right\},|w|<1\right\}$  in ${\mathcal{ℬ}}^{p}\left({U}^{n}$  defined as follows: set ${f}_{w}\left(z\right)={\int }_{0}^{{z}_{l}}{\left(1-\frac{{\overline{w}}^{2}}{|w{|}^{2}}{z}_{l}^{2}\right)}^{-p}d{z}_{l}$  with ${f}_{w}\left(0\right)=0.$  Since for $w\ne 0,$  $\frac{\partial {f}_{w}}{\partial {z}_{l}}={\left(1-\frac{{\overline{w}}^{2}}{|w{|}^{2}}{z}_{l}^{2}\right)}^{-p},\frac{\partial {f}_{w}}{\partial {z}_{i}}=0,\left(i\ne l\right),$  it is easy to show $\parallel {f}_{w}{\parallel }_{\alpha }=1.$  Lemma 1 gives $|{f}_{w}\left(z\right)|\le 1+\frac{n}{1-p},$  so by ( 3 ),
 $\begin{array}{c}{\sum }_{k=1}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left(z\right){f}_{w}\left(\phi \left(z\right)\right)\left(1-|{z}_{k}{|}^{2}{\right)}^{q}|\le C{\sum }_{k=1}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left(z\right)|{\left(1-|{z}_{k}{|}^{2}\right)}^{q}\le C.\end{array}$ (29)
For $z\in {U}^{n},$  it follows from ( 22 ) and ( 27 ) that ${\sum }_{k=1}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left(z\right){f}_{w}\left(\phi \left(z\right)\right)+\psi \left(z\right){\sum }_{l=1}^{n}\frac{\partial {f}_{w}\left(\phi \left(z\right)\right)}{\partial {\phi }_{l}}\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\left(1-|{z}_{k}{|}^{2}{\right)}^{q}\le C,$  combining ( 29 ), we get ${\sum }_{k=1}^{n}|\psi \left(z\right)\frac{\partial {f}_{w}\left(\phi \left(z\right)\right)}{\partial {\phi }_{l}}\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\left(1-|{z}_{k}{|}^{2}{\right)}^{q}\le C.$  For an arbitrary $z\in {U}^{n}$  with ${\phi }_{l}\left(z\right)\ne 0$  and set $w={\phi }_{l}\left(z\right)$  in the above inequality to obtain $|\psi \left(z\right)|{\sum }_{k=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{{\left(1-|{z}_{k}{|}^{2}\right)}^{q}}{{\left(1-|{\phi }_{l}\left(z\right){|}^{2}\right)}^{p}}\le C.$  If ${\phi }_{l}\left(z\right)=0,$  let $f\left(z\right)={z}_{l},$  it follows from ( 22 ) and ( 27 ) that $|\psi \left(z\right)||\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{{\left(1-|{z}_{k}{|}^{2}\right)}^{q}}{{\left(1-|{\phi }_{l}\left(z\right){|}^{2}\right)}^{p}}=|\frac{\partial \psi }{\partial {z}_{k}}\left(z\right){\phi }_{l}\left(z\right)+\psi \left(z\right)\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\left(1-|{z}_{k}{|}^{2}{\right)}^{q}\le C.$  So for all $z\in {U}^{n},$  $|\psi \left(z\right)|{\sum }_{k,l=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{{\left(1-|{z}_{k}{|}^{2}\right)}^{q}}{{\left(1-|{\phi }_{l}\left(z\right){|}^{2}\right)}^{p}}\le C,$  that is, ( 4 ) is valid.
If $p>1$  , similarly, the test functions would be ${f}_{w}\left(z\right)=\frac{1}{\overline{w}}\left(\frac{1-|w{|}^{2}}{{\left(1-\overline{w}{z}_{l}\right)}^{p}}-\frac{1}{{\left(1-\overline{w}{z}_{l}\right)}^{p-1}}\right),\left(|w|<1,w\ne 0,1\le l\le n\right)$  for proving ( 6 ), and ${g}_{w}\left(z\right)=\left(\frac{p}{{\left(1-\overline{w}{z}_{l}\right)}^{p-1}}-\frac{\left(p-1\right)\left(1-|w{|}^{2}\right)}{{\left(1-\overline{w}{z}_{l}\right)}^{p}}\right),\left(|w|<1,1\le l\le n\right)$  for proving ( 5 ), we omit the details. Now the proof of Theorem 1 is completed.

4 The Proof of Theorem 2

To show the sufficiency.
If $p=1$  . First assume conditions ( 7 ) and ( 8 ) hold, we need to prove ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)⟶{\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is compact. According to Lemma 2, assume that ${\parallel {f}_{j}\parallel }_{p}\le C,j=1,2,\dots ,$  and $\left\{{f}_{j}\right\}$  converges to zero uniformly on compact subsets of ${U}^{n},$  we need only prove that ${\parallel {W}_{\psi ,\phi }{f}_{j}\parallel }_{q}\to 0,$  as $j\to \infty .$  For every $\varepsilon >0,$  ( 7 ) and ( 8 ) imply that there exists a $r$  , $0  , such that
 $\begin{array}{c}{\sum }_{k,l=1}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left(z\right)|{\left(1-|{z}_{k}{|}^{2}\right)}^{q}ln\frac{4}{1-|{\phi }_{l}\left(z\right){|}^{2}}<\varepsilon \end{array}$ (30)
and
 $\begin{array}{c}{\sum }_{k,l=1}^{n}|\psi \left(z\right)\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{{\left(1-|{z}_{k}{|}^{2}\right)}^{q}}{1-|{\phi }_{l}\left(z\right){|}^{2}}<\varepsilon ,\end{array}$ (31)
whenever $dist\left(\phi \left(z\right),\partial {U}^{n}\right)  Since ${f}_{j}\left(\phi \left(0\right)\right)$  converges to zero, by ( 30 ) and ( 31 ), it follows from ( 24 ) that for large enough $j$
 $\begin{array}{c}{\sum }_{k=1}^{n}|\frac{\partial \left({W}_{\psi ,\phi }{f}_{j}\right)}{\partial {z}_{k}}\left(z\right)|\left(1-|{z}_{k}{|}^{2}{\right)}^{q}\le C\varepsilon ,\end{array}$ (32)
whenever $dist\left(\phi \left(z\right),\partial {U}^{n}\right)  If we set $\varepsilon =1,$  by ( 30 ) and ( 31 ),it is easy to show that ( 1 ) and ( 2 ) holds, from Remark 1, we know $\psi \in {\mathcal{ℬ}}^{q}\left({U}^{n}\right).$  On the other hand, if we write $E=\left\{w\in {U}^{n}:dist\left(w,\partial {U}^{n}\right)\ge r\right\},$  which is a closed subset of ${U}^{n}$  , then ${f}_{j}\left(w\right)$  and $\frac{\partial {f}_{j}}{\partial {w}_{l}}\left(w\right)\to 0$  uniformly on $E$  . So by ( 22 ) and the boundedness condition (2) in Theorem 1, we obtain for $dist\left(\phi \left(z\right),\partial {U}^{n}\right)\ge r,$
 $\begin{array}{ccc}& & {\sum }_{k=1}^{n}|\frac{\partial \left({W}_{\psi ,\phi }{f}_{j}\right)}{\partial {z}_{k}}\left(z\right)|\left(1-|{z}_{k}{|}^{2}{\right)}^{q}\le C\left(|{f}_{j}\left(\phi \left(z\right)\right)|+{\sum }_{l=1}^{n}|\frac{\partial {f}_{j}}{\partial {w}_{l}}\left(\phi \left(z\right)\right)|\right)\end{array}$
 $\begin{array}{ccc}& & =C\left(|{f}_{j}\left(w\right)|+{\sum }_{l=1}^{n}|\frac{\partial {f}_{j}}{\partial {w}_{l}}\left(w\right)|\right)\le C\varepsilon .\end{array}$ (33)
Since $\psi \left(0\right){f}_{j}\left(\phi \left(0\right)\right)\to 0$  for large enough $j,$  combining ( 32 ) and ( 33 ) we know ${\parallel {W}_{\psi ,\phi }{f}_{j}\parallel }_{q}\to 0,$  as $j\to \infty .$  If $p>1,$  note that ( 26 ), in a similar manner to the case $p=1$  , we can show the sufficiency, omit the details.
To show the necessity.
First we will prove the following Lemma:
Lemma 7 Let $p\ge 1$  and $q\ge 0$  . If ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)⟶{\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is compact, then
 $\begin{array}{c}|\psi \left(z\right)|{\sum }_{k,l=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{{\left(1-|{z}_{k}{|}^{2}\right)}^{q}}{{\left(1-|{\phi }_{l}\left(z\right){|}^{2}\right)}^{p}}=o\left(1\right)\left(\phi \left(z\right)\to \partial {U}^{n}\right).\end{array}$ (34)
Proof For any $\left\{{z}^{j}\right\}$  in ${U}^{n}$  with $\phi \left({z}^{j}\right)\to \partial {U}^{n}$  as $j\to \infty .$  Let ${A}_{j}=\phi \left({z}^{j}\right)=\left({a}_{1j},\dots ,{a}_{nj}\right),$  where ${a}_{kj}={\phi }_{k}\left({z}^{j}\right),1\le k\le n.$  Since $\phi \left({z}^{j}\right)\to \partial {U}^{n},$  as $j\to \infty$  , there exists some $s,\left(1\le s\le n\right),$  with $|{a}_{sj}|\to 1$  as $j\to \infty .$  Without loss of generality, we may assume that $s=1$  .
Case 1: If for some $l\left(1\le l\le n\right),$  $|{a}_{lj}|\to 1$  as $j\to \infty ,$  then set
 $\begin{array}{c}{f}_{j}\left(z\right)=\frac{{\left(1-|{a}_{lj}{|}^{2}\right)}^{2}}{{\left(1-{z}_{l}\overline{{a}_{lj}}\right)}^{p+1}}-\frac{1-|{a}_{lj}{|}^{2}}{{\left(1-{z}_{l}\overline{{a}_{lj}}\right)}^{p}},j=1,2,\cdots .\end{array}$ (35)
$\frac{\partial {f}_{j}}{\partial {z}_{k}}=0,\left(k\ne l\right),$  $\frac{\partial {f}_{j}}{\partial {z}_{l}}=\left(p+1\right)\overline{{a}_{lj}}\frac{{\left(1-|{a}_{lj}{|}^{2}\right)}^{2}}{{\left(1-{z}_{l}\overline{{a}_{lj}}\right)}^{p+2}}-p\overline{{a}_{lj}}\frac{1-|{a}_{lj}{|}^{2}}{{\left(1-{z}_{l}\overline{{a}_{lj}}\right)}^{p+1}},$  so $\parallel {f}_{j}{\parallel }_{p}\le C$  and $\left\{{f}_{j}\left(z\right)\right\}$  tends to zero uniformly on compact subsets of ${U}^{n}.$  So by Lemma 2, we know ${\parallel {W}_{\psi ,\phi }{f}_{j}\parallel }_{q}\to 0$  as $j\to \infty .$  It follows from ( 22 ) that
 $\begin{array}{ccc}& & {\sum }_{k=1}^{n}{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}|\frac{\partial \psi }{\partial {z}_{k}}\left({z}^{j}\right){f}_{j}\left(\phi \left({z}^{j}\right)\right)+\psi \left({z}^{j}\right)\frac{\partial {f}_{j}}{\partial {w}_{l}}\left(\phi \left({z}^{j}\right)\right)\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left({z}^{j}\right)|\end{array}$
 $\begin{array}{ccc}& & \le {\parallel {W}_{\psi ,\phi }{f}_{j}\parallel }_{q}\to 0.\end{array}$ (36)
Note that ${f}_{j}\left(\phi \left({z}^{j}\right)\right)=0$  and $\frac{\partial {f}_{j}}{\partial {z}_{l}}\left(\phi \left({z}^{j}\right)\right)=\overline{{a}_{lj}}\frac{1}{{\left(1-|{a}_{lj}{|}^{2}\right)}^{p}}$  , ( 36 ) gives that
 $\begin{array}{ccc}& & |\psi \left({z}^{j}\right)|{\sum }_{k=1}^{n}\frac{{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}}{{\left(1-|{a}_{lj}{|}^{2}\right)}^{p}}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left({z}^{j}\right)|\end{array}$
 $\begin{array}{ccc}& & \le C\frac{1}{|{a}_{lj}|}{\parallel {W}_{\psi ,\phi }{f}_{j}\parallel }_{q}\to 0\text{as}j\to \infty .\text{}\end{array}$ (37)
Case 2: If for some $l\left(2\le l\le n\right),$  $|{a}_{lj}|\to ̸1$  as $j\to \infty ,$  then we assume $|{a}_{lj}|\le \rho <1.$  If $p\ge 1$  , The compactness of ${W}_{\psi ,\phi }$  implies ${W}_{\psi ,\phi }$  is bounded, by Theorem 1 and Remark 1, ( 11 ) is valid. Set
 $\begin{array}{c}{f}_{j}\left(z\right)={z}_{l}{\left(\frac{1-|{a}_{1j}{|}^{2}}{1-{z}_{1}\overline{{a}_{1j}}}\right)}^{p},j=1,2,\cdots .\end{array}$ (38)
It is easy to show that $\parallel {f}_{j}{\parallel }_{p}\le C$  , and $\left\{{f}_{j}\right\}$  tends to zero uniformly on compact subsets of ${U}^{n}.$  It follows from ( 22 ) that
 $\begin{array}{ccc}& & |\psi \left({z}^{j}\right)|{\sum }_{k=1}^{n}{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left({z}^{j}\right)|\end{array}$
 $\begin{array}{ccc}& & \le {\parallel {W}_{\psi ,\phi }{f}_{j}\parallel }_{q}+C{\sum }_{k=1}^{n}{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}|\frac{\partial \psi }{\partial {z}_{k}}\left({z}^{j}\right)|\end{array}$
 $\begin{array}{ccc}& & +C|\psi \left({z}^{j}\right)|{\sum }_{k=1}^{n}\frac{{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}}{1-|{a}_{1j}{|}^{2}}|\frac{\partial {\phi }_{1}}{\partial {z}_{k}}\left({z}^{j}\right)|\end{array}$
 $\begin{array}{ccc}& & \le {\parallel {W}_{\psi ,\phi }{f}_{j}\parallel }_{q}+C{\sum }_{k=1}^{n}{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}|\frac{\partial \psi }{\partial {z}_{k}}\left({z}^{j}\right)|\end{array}$
 $\begin{array}{ccc}& & +C|\psi \left({z}^{j}\right)|{\sum }_{k=1}^{n}\frac{{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}}{\left(1-|{a}_{1j}{|}^{2}{\right)}^{p}}|\frac{\partial {\phi }_{1}}{\partial {z}_{k}}\left({z}^{j}\right)|\to 0.\end{array}$
So
 $\begin{array}{ccc}& & |\psi \left({z}^{j}\right)|{\sum }_{k=1}^{n}\frac{{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}}{\left(1-|{a}_{lj}{|}^{2}{\right)}^{p}}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left({z}^{j}\right)|\end{array}$
 $\begin{array}{ccc}& & \le \frac{1}{\left(1-{\rho }^{2}{\right)}^{p}}|\psi \left({z}^{j}\right)|{\sum }_{k=1}^{n}{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left({z}^{j}\right)|\to 0.\end{array}$ (39)
If $p=1$  , by Lemma 7, ( 8 ) is necessary, now we prove that ( 7 ) is also necessary.
In case 1, set
 $\begin{array}{c}{f}_{j}\left(z\right)={\left(ln\frac{4}{1-|{a}_{lj}{|}^{2}}\right)}^{-1}{\left(ln\frac{4}{1-{z}_{l}\overline{{a}_{lj}}}\right)}^{2},\end{array}$ (40)
It is easy to show that $\parallel {f}_{j}{\parallel }_{p}\le C$  , and $\left\{{f}_{j}\right\}$  tends to zero uniformly on compact subsets of ${U}^{n}.$  It follows from ( 22 ) that ${\sum }_{k=1}^{n}{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}|\frac{\partial \psi }{\partial {z}_{k}}\left({z}^{j}\right){f}_{j}\left(\phi \left({z}^{j}\right)\right)+\psi \left({z}^{j}\right)\frac{\partial {f}_{j}}{\partial {w}_{l}}\left(\phi \left({z}^{j}\right)\right)\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left({z}^{j}\right)|\le {\parallel {W}_{\psi ,\phi }{f}_{j}\parallel }_{q}\to 0.$  By a direct calculation and by ( 34 ), we obtain
 $\begin{array}{ccc}& & {\sum }_{k=1}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left(z\right)|{\left(1-|{z}_{k}{|}^{2}\right)}^{q}ln\frac{4}{1-|{\phi }_{l}\left(z\right){|}^{2}}\end{array}$
 $\begin{array}{ccc}& & \le {\parallel {W}_{\psi ,\phi }{f}_{j}\parallel }_{q}+2|\psi \left({z}^{j}\right)|{\sum }_{k=1}^{n}\frac{{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}}{{\left(1-|{a}_{lj}{|}^{2}\right)}^{p}}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left({z}^{j}\right)|\to 0.\end{array}$ (41)
In case 2, by remark 1, ${\sum }_{k=1}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left({z}^{j}\right)|{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}ln\frac{4}{1-|{\phi }_{l}\left({z}^{j}\right){|}^{2}}\le ln\frac{4}{1-{\rho }^{2}}{\sum }_{k=1}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left({z}^{j}\right)|{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}\to 0.$  Combining case 1 and case 2, we know ( 7 ) is necessary.
If $p>1.$  by Lemma 7, we know ( 10 ) is necessary, now we prove ( 9 ) is also necessary.
In case 1, set ${f}_{j}\left(z\right)=\left(p+1\right)\frac{1-|{a}_{lj}{|}^{2}}{{\left(1-{z}_{l}\overline{{a}_{lj}}\right)}^{p}}-p\frac{{\left(1-|{a}_{lj}{|}^{2}\right)}^{2}}{{\left(1-{z}_{l}\overline{{a}_{lj}}\right)}^{p+1}},j=1,2,\cdots .$  It is easy to show that $\parallel {f}_{j}{\parallel }_{p}\le C$  , and $\left\{{f}_{j}\right\}$  tends to zero uniformly on compact subsets of ${U}^{n}.$  Note that ${f}_{j}\left(\phi \left({z}^{j}\right)\right)=\frac{1}{{\left(1-|{a}_{lj}{|}^{2}\right)}^{p-1}}$  and $\frac{\partial {f}_{j}}{\partial {z}_{l}}\left(\phi \left({z}^{j}\right)\right)=0,$  it follows from ( 22 ) that
 $\begin{array}{ccc}& & {\sum }_{k=1}^{n}{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}|\frac{\partial \psi }{\partial {z}_{k}}\left({z}^{j}\right){f}_{j}\left(\phi \left({z}^{j}\right)\right)+\psi \left({z}^{j}\right)\frac{\partial {f}_{j}}{\partial {w}_{l}}\left(\phi \left({z}^{j}\right)\right)\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left({z}^{j}\right)|\end{array}$
 $\begin{array}{ccc}& & ={\sum }_{k=1}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left({z}^{j}\right)|\frac{{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}}{\left(1-|{a}_{lj}{|}^{2}{\right)}^{p-1}}\le {\parallel {W}_{\psi ,\phi }{f}_{j}\parallel }_{q}\to 0.\end{array}$ (42)
In case 2, by remark 1,
 $\begin{array}{ccc}& & {\sum }_{k=1}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left({z}^{j}\right)|\frac{{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}}{{\left(1-|{\phi }_{l}\left({z}^{j}\right){|}^{2}\right)}^{p-1}}\end{array}$
 $\begin{array}{ccc}& & \le \frac{1}{{\left(1-{\rho }^{2}\right)}^{p-1}}{\sum }_{k=1}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left({z}^{j}\right)|{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}\to 0.\end{array}$
Combining cas1 and case 2, we know ( 9 ) is necessary.
Now the proof of Theorem 2 is completed.

5 Proof of Theorem 3

Proof of Theorem 3. First assume ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)⟶{\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is bounded and ( 12 ) holds, we prove that ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)\to {\mathcal{ℬ}}^{q}\left({U}^{n}\right)$  is compact. By Lemma 6, assume that ${\parallel {f}_{j}\parallel }_{p}\le C,j=1,2,...,$  and $\left\{{f}_{j}\right\}$  converges to zero uniformly on $\overline{{U}^{n}},$  we need to prove that ${\parallel {W}_{\psi ,\phi }{f}_{j}\parallel }_{q}\to 0,$  as $j\to \infty .$  Note that the boundedness of ${W}_{\psi ,\phi }$  and Theorem 1, ( 1 ) and ( 2 ) hold. So
 $\begin{array}{c}|\psi \left(z\right)|\le C\parallel \psi {\parallel }_{q}\le C,|\psi \left(z\right)|{\sum }_{k=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{\left(1-|{z}_{k}{|}^{2}{\right)}^{q}}{\left(1-|{\phi }_{l}\left(z\right){|}^{2}{\right)}^{p}}\le C\end{array}$ (43)
for $l\in \left\{1,2,\cdots ,n\right\}.$  By the assumption as $j\to \infty$
 $\begin{array}{c}{sup}_{z\in \overline{{U}^{n}}}|{f}_{j}\left(z\right)|\to 0\end{array}$ (44)
 $\begin{array}{ccc}& & {\sum }_{k=1}^{n}|\frac{\partial \left({W}_{\psi ,\phi }{f}_{j}\left(z\right)\right)}{\partial {z}_{k}}|\left(1-|{z}_{k}{|}^{2}{\right)}^{q}\end{array}$
 $\begin{array}{ccc}& & ={\sum }_{k=1}^{n}|\frac{\partial \psi \left(z\right)}{\partial {z}_{k}}{f}_{j}\left(\phi \left(z\right)\right)+\psi \left(z\right){\sum }_{l=1}^{n}\frac{\partial {f}_{j}}{\partial {w}_{l}}\left(\phi \left(z\right)\right)\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\left(1-|{z}_{k}{|}^{2}{\right)}^{q}\end{array}$
 $\begin{array}{ccc}& & \le |{f}_{j}\left(\phi \left(z\right)\right)|{\sum }_{k=1}^{n}|\frac{\partial \psi }{\partial {z}_{k}}\left(z\right)|{\left(1-|{z}_{k}{|}^{2}\right)}^{q}\end{array}$
 $\begin{array}{ccc}& & +{\sum }_{k=1}^{n}{\sum }_{l=1}^{n}|\frac{\partial {f}_{j}}{\partial {w}_{l}}\left(\phi \left(z\right)\right)||\psi \left(z\right)||\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{\left(1-|{z}_{k}{|}^{2}{\right)}^{q}}{\left(1-|{\phi }_{l}\left(z\right){|}^{2}{\right)}^{p}}\end{array}$
 $\begin{array}{ccc}& & \le C|{f}_{j}\left(\phi \left(z\right)\right)|+{\sum }_{k=1}^{n}{\sum }_{k=1}^{n}|\frac{\partial {f}_{j}}{\partial {w}_{l}}\left(\phi \left(z\right)\right)||\psi \left(z\right)||\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{\left(1-|{z}_{k}{|}^{2}{\right)}^{q}}{\left(1-|{\phi }_{l}\left(z\right){|}^{2}{\right)}^{p}}\end{array}$
 $\begin{array}{ccc}& & \le C\varepsilon +{\sum }_{k=1}^{n}{\sum }_{l=1}^{n}|\frac{\partial {f}_{j}}{\partial {w}_{l}}\left(\phi \left(z\right)\right)||\psi \left(z\right)||\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{\left(1-|{z}_{k}{|}^{2}{\right)}^{q}}{\left(1-|{\phi }_{l}\left(z\right){|}^{2}{\right)}^{p}}.\end{array}$ (45)
For every $\varepsilon >0$  , $l\in \left\{1,...,n\right\},$  ( 12 ) implies that there exists an $r$  , $0  , such that
 $\begin{array}{c}|\psi \left(z\right)|{\sum }_{k=1}^{n}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{\left(1-|{z}_{k}{|}^{2}{\right)}^{q}}{\left(1-|{\phi }_{l}\left(z\right){|}^{2}{\right)}^{p}}<\varepsilon ,\end{array}$ (46)
whenever $|{\phi }_{l}\left(z\right)|>r.$  Note that ${U}^{n}=\left\{z\in {U}^{n}:|{\phi }_{l}\left(z\right)|>r\right\}\bigcup \left\{z\in {U}^{n}:|{\phi }_{l}\left(z\right)|\le r\right\}.$  For each $l\in \left\{1,2,\cdots ,n\right\}.$  If $|{\phi }_{l}\left(z\right)|>r,$  from ( 46 ) and $\parallel {f}_{j}{\parallel }_{p}\le C$  , we have
 $\begin{array}{c}|\frac{\partial {f}_{j}}{\partial {w}_{l}}\left(\phi \left(z\right)\right)||\psi \left(z\right)||\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{\left(1-|{z}_{k}{|}^{2}{\right)}^{q}}{\left(1-|{\phi }_{l}\left(z\right){|}^{2}{\right)}^{p}} (47)
If $|{\phi }_{l}\left(z\right)|\le r,$  by the Cauchy's estimate applied to the function $g\left({w}_{l}\right)=f\left({w}_{1},...,{w}_{l-1},{w}_{l},{w}_{l+1}...,{w}_{n}\right),$  of one variable, if $|{w}_{l}|\le r,$  we have $|\frac{\partial {f}_{j}}{\partial {w}_{l}}\left(w\right)|\le C{sup}_{|{w}_{l}|\le \frac{1+r}{2}}|{f}_{j}\left(w\right)|\le C{sup}_{z\in \overline{{U}^{n}}}|{f}_{j}\left(z\right)|,$  for some $C>0$  independent of $f.$  From this, ( 43 ) and ( 44 ) we have
 $\begin{array}{ccc}& & |\frac{\partial {f}_{j}}{\partial {w}_{l}}\left(\phi \left(z\right)\right)||\psi \left(z\right)||\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left(z\right)|\frac{\left(1-|{z}_{k}{|}^{2}{\right)}^{q}}{\left(1-|{\phi }_{l}\left(z\right){|}^{2}{\right)}^{p}}\end{array}$
 $\begin{array}{ccc}& & \le C{sup}_{z\in \overline{{U}^{n}}}|{f}_{j}\left(z\right)|\to 0\end{array}$ (48)
as $j\to \infty .$  So From ( 45 ),( 47 ) and ( 48 ), and since ${lim}_{j\to \infty }{f}_{j}\left(\phi \left(0\right)\right)=0,$  we obtain ${\parallel {C}_{\phi }{f}_{j}\parallel }_{q}\to 0,$  as $j\to \infty ,$  from which the result follows.
Now suppose that ${W}_{\psi ,\phi }:{\mathcal{ℬ}}^{p}\left({U}^{n}\right)\to {\mathcal{ℬ}}^{q}\left({\text{C}}^{n}\right)$  is compact. Then ${W}_{\psi ,\phi }$  is bounded, and by Theorem 1 we know that ( 4 ) holds. Now we prove that condition ( 12 ) holds. Let $\left\{{z}^{j}\right\}$  be a sequence in ${U}^{n}$  and ${w}^{j}=\phi \left({z}^{j}\right)=\left({w}_{1}^{j},...,{w}_{n}^{j}\right).$  For each $l\in \left\{1,2,\cdots ,n\right\}$  with ${lim}_{j\to \infty }|{w}_{l}^{j}|\to 1.$  Let ${f}_{j}\left(z\right)=\frac{1-|{w}_{l}^{j}{|}^{2}}{\left(1-{z}_{l}\overline{{w}_{l}^{j}}{\right)}^{p}},j=1,2,...,$  then $\frac{\partial {f}_{j}}{\partial {z}_{l}}\left(z\right)=p\overline{{w}_{l}^{j}}\frac{1-|{w}_{l}^{j}{|}^{2}}{\left(1-{z}_{l}\overline{{w}_{l}^{j}{|}^{2}}{\right)}^{p+1}}.$  As in the proof of Theorem 2 we can prove that $\parallel {f}_{j}{\parallel }_{p}\le 1+{2}^{p+1}p,$  for all $j\in \mathbf{N},$  i.e. $\left\{{f}_{j}\right\}$  is bounded on ${\mathcal{ℬ}}^{p}\left({U}^{n}\right).$  It is easy to see that ${f}_{j}$  tends to zero uniformly on compact subsets of ${\text{C}}^{n}.$  So by Lemma 2, we know ${\parallel {W}_{\phi }{f}_{j}\parallel }_{q}\to 0$  as $j\to \infty$  and notice that ${f}_{j}\left(\phi \left({z}^{j}\right)\right)=\left(1-|{w}_{l}^{j}{|}^{2}{\right)}^{1-p},\frac{\partial {f}_{j}}{\partial {w}_{l}}\left(\phi \left({z}^{j}\right)\right)=p\overline{{w}_{l}^{j}}\frac{1}{\left(1-|{w}_{l}^{j}{|}^{2}{\right)}^{p}},$  consequently, by ( 22 )
 $\begin{array}{ccc}& & {\sum }_{k=1}^{n}{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}|\frac{\partial \psi }{\partial {z}_{k}}\left({z}^{j}\right){f}_{j}\left(\phi \left({z}^{j}\right)\right)+\psi \left({z}^{j}\right)\frac{\partial {f}_{j}}{\partial {w}_{l}}\left(\phi \left({z}^{j}\right)\right)\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left({z}^{j}\right)|\end{array}$
 $\begin{array}{ccc}& & \le {\parallel {W}_{\psi ,\phi }{f}_{j}\parallel }_{q}.\end{array}$ (49)
So ${\sum }_{k=1}^{n}p|{w}_{l}^{j}||\psi \left({z}^{j}\right)|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left({z}^{j}\right)|\frac{{\left(1-|{z}_{k}^{j}{|}^{2}\right)}^{q}}{\left(1-|{w}_{l}^{j}{|}^{2}{\right)}^{p}}\le \parallel \psi {\parallel }_{q}{\left(1-|{w}_{l}^{j}{|}^{2}\right)}^{1-p}+{\parallel {W}_{\psi ,\phi }{f}_{j}\parallel }_{q}.$  Hence $|\psi \left({z}^{j}\right)|{\sum }_{k=1}^{n}\frac{\left(1-|{z}_{k}^{j}{|}^{2}{\right)}^{q}}{\left(1-|{w}_{l}^{j}{|}^{2}{\right)}^{p}}|\frac{\partial {\phi }_{l}}{\partial {z}_{k}}\left({z}^{j}\right)|\le \frac{1}{p|{w}_{l}^{j}|}\left(C\left(1-|{w}_{l}^{j}{|}^{2}{\right)}^{1-p}+{\parallel {W}_{\psi ,\phi }{f}_{j}\parallel }_{q}\right)\to 0$  as $j\to \infty .$  Now the proof of Theorem 3 is finished.
References

1. R.Timoney, Bloch Function in Several Complex Variables, I, Bull.London Math.Soc. 12(1980),241-267.
2. R.Timoney, Bloch Function in Several Complex Variables, II, J. Reine Angew Math., 319(1980):1-22.
3. C. C. Cowen and B. D. MacCluer, Composition operators on spaces of analytic functions, CRC Press, Boca Raton, FL,1995.
4. K. Madigan and A. Matheson, Compact composition operators on the Bloch space, Trans. Amer. Math. Soc. 347(1995): 2679-2687.
5. K. Madigan, Compact composition operators on analytic Lipschitz space, Proc. Amer. Math. Soc. 119(1993): 465-473.
6. M. D. Contreras and A. G. Hernandez-Diaz, Weighted composition operators in weighted Banach spaces of ananytic functions, J. Austral. Math. Soc. (Series A), 69(2000):41-46.
7. Alfonso Montes-Rodriguez, Weighted composition operators on weighted banach spaces of analytic functions, J. London Math. Soc. (2)61(2000):872-884.
8. S.Ohno, K. Stroethoff and R.H.Zhao, Weighted composition operators between-type spaces, Rocky Mountain Journal of Mathematics,Vol33, No.1(2003):191-215.
9. J. H. Shi and L. Luo, Composition operators on the Bloch space of several complex variables, Acta Math. Sinica, English Series, 16(2000):85-98.
10. Z. H. Zhou and J. H. Shi, Compact composition operators on the Bloch space in polydisk, Science in China (Series A), 44(2001): 286-291.
11. Z. H. Zhou, Composition operators on the Lipschitz spaces in polydisk. Science in China (Series A), 46(2003): 33-38.
12. Z. H. Zhou and J. H. Shi, Compactness of composition operators on the Bloch space in classical bounded symmetric domains, The Michigan Mathematical Journal, 50(2002): 381-405.
13. J. P. D'Angelo, Several complex variables and geometry of real hypersurfaces, CRC Press, 1993.

$1$  Supported in part by the National Natural Science Foundation of China (Grand No.10371091), and LiuHui Center for Applied Mathematics, Nankai University and Tianjin University