数学物理学报, 2023, 43(1): 27-34

调和Hardy空间上的Toeplitz算子的酉等价性

丁宣浩,1,2, 黄雨浩,1, 李永宁,1,2,*

1重庆工商大学数学与统计学院 重庆 400067

2经济社会应用统计重庆市重点实验室 重庆 400067

The Unitary Equivalence of the Toeplitz Operators on the Harmonic Hardy Space

Ding Xuanhao,1,2, Huang Yuhao,1, Li Yongning,1,2,*

1School of Mathematics and Statistics, Chongqing Technology and Business University, Chongqing 400067

2Chongqing Key Laboratory of Social Economy and Applied Statistics, Chongqing 400067

通讯作者: *李永宁, E-mail: yongningli@ctbu.edu.cn

收稿日期: 2021-12-22   修回日期: 2022-08-17  

基金资助: 国家自然科学基金(11871122)
国家自然科学基金(12101092)
重庆市自然科学基金(cstc2020jcyj-msxmX0318)
重庆工商大学基金(2056008)
重庆工商大学校级科研项目(yjscxx2022-112-73)

Received: 2021-12-22   Revised: 2022-08-17  

Fund supported: The NSFC(11871122)
The NSFC(12101092)
Natural Science Foundation of Chongqing(cstc2020jcyj-msxmX0318)
Chongqing Technology and Business University(2056008)
Scientific research project in Chongqing Technology and Business University-level(yjscxx2022-112-73)

作者简介 About authors

丁宣浩,E-mail:dingxuanhao@ctbu.edu.cn

黄雨浩,E-mail:huangyuhaoctbu@163.com

摘要

$H^{2}$是单位圆盘${\Bbb D}=\{\xi\in{\Bbb C}:|\xi|<1\}$上的经典Hardy空间. 设$u$$v$是内函数且至少其中一个是非常值的, 调和Hardy空间$H_{u,v}^{2}$定义为$H_{u,v}^{2}=uH^{2}\oplus\overline{v}(H^{2})^{\perp}=uH^{2}\oplus\overline{vzH^{2}}$. 对任意的$x\in H_{u,v}^{2},$ 定义$H_{u,v}^{2}$上的调和Toeplitz算子 $\widehat{T}_{\varphi}x=Q_{u,v}(\varphi x),$ 其中, $Q_{u,v}:L^{2}\rightarrow H_{u,v}^{2}$为正交投影. 该文刻画了调和Toeplitz算子和对偶截断Toeplitz算子的酉等价性, 并给出了两个调和Toeplitz算子可交换的充要条件, 调和Toeplitz代数的性质以及$\widehat{T}_{z}$的换位子的刻画. 最后, 该文还得到了有限多个连续符号的调和 Toeplitz算子乘积的本质谱.

关键词: 调和Hardy空间; 调和Toeplitz算子; 酉等价; 对偶截断Toeplitz算子; 本质谱

Abstract

Let $H^{2}$ be the Hardy space on the unit disk ${\Bbb D}=\{\xi\in{\Bbb C}:|\xi|<1\}$. Suppose $u$ and $v$ are inner functions and at least one of them is nonconstant, the harmonic Hardy space $H_{u,v}^{2}$ is defined by $H_{u,v}^{2}=uH^{2}\oplus\overline{v}(H^{2})^{\bot}=uH^{2}\oplus\overline{vzH^{2}}.$ For any $x\in H_{u,v}^{2},$ define the Toeplitz operator on the $H_{u,v}^{2}$ by $\widehat{T}_{\varphi}x=Q_{u,v}(\varphi x),$ where $Q_{u,v}$ is the orthogonal projection from $L^{2}\rightarrow H_{u,v}^{2}.$ In this paper, the unitary equivalence of the harmonic Toeplitz operator and the dual truncated Toeplitz operator is obtained, moreover, the sufficient and necessary conditions for when two Toeplitz operators commute is given, and the properties of the harmonic Toeplitz algebra and the commutant of $\widehat{T}_{z}$ are described. Finally, the essential spectrum for the product of finitely many harmonic Toeplitz operators with continuous symbols is obtained in this paper.

Keywords: Harmonic Hardy space; Harmonic Toeplitz operator; Unitary equivalence; Dual truncated Toeplitz operator; Essential spectrum

PDF (348KB) 元数据 多维度评价 相关文章 导出 EndNote| Ris| Bibtex  收藏本文

本文引用格式

丁宣浩, 黄雨浩, 李永宁. 调和Hardy空间上的Toeplitz算子的酉等价性[J]. 数学物理学报, 2023, 43(1): 27-34

Ding Xuanhao, Huang Yuhao, Li Yongning. The Unitary Equivalence of the Toeplitz Operators on the Harmonic Hardy Space[J]. Acta Mathematica Scientia, 2023, 43(1): 27-34

1 引言

${\Bbb D}=\{\xi\in{\Bbb C}:|\xi|<1\}$是复平面${\Bbb C}$上的单位圆盘, $\partial{\Bbb D}=\{\xi\in{\Bbb C}:|\xi|=1\}$ 是单位圆周. 记$H^{2}$是经典的Hardy空间, $(H^{2})^{\perp}=\overline{zH^{2}}$是其正交补空间. 记$L^{2}=L^{2}(\partial{\Bbb D})$ 是单位圆周 $\partial{\Bbb D}$上的Lebesgue平方可积函数构成的空间, $L^{\infty}$是单位圆周 $\partial{\Bbb D}$上的本性有界的可测函数全体构成的空间, $H^{\infty}$为单位开圆盘${\Bbb D}$上的有界解析函数全体构成的空间[1]. 如果$u \in H^{\infty}$$|u({\rm e}^{{\rm i}t})|=1$$\partial{\Bbb D}$上几乎处处成立, 则称$u$ 是内函数.

$P:L^{2}\rightarrow H^{2}$ 为正交投影, 则对于 $\varphi\in L^{\infty}$, 符号为$\varphi$的Toeplitz算子$T_{\varphi}$定义为

$T_{\varphi}f=P(\varphi f), \forall f\in H^{2}.$

Hankel算子$H_{\varphi}$定义为

$H_{\varphi}g=(I-P)(\varphi g), \forall g\in H^{2}. $

对偶Toeplitz算子$S_{\varphi}$定义为

$S_{\varphi}h=(I-P)(\varphi h), \forall \in (H^{2})^{\perp}. $

容易得到

$H_{\varphi}^{\ast} h=P(\overline{\varphi} h), \forall h\in (H^{2})^{\perp}. $

$M_{\varphi}$$L^{2}$上的乘法算子, 且$M_{\varphi} x=\varphi x.$$M_{\varphi}$在空间分解$L^{2}=H^{2}\oplus (H^{2})^{\bot}$ 下可表示为如下$2\times 2$算子矩阵形式

$M_{\varphi}=\left[\begin{array}{cccccc}T_\varphi& & H_{\overline{\varphi}}^{*} \\ H_\varphi && S_\varphi \\ \end{array}\right]. $

由于$M_{\varphi}M_{\psi}=M_{\varphi\psi},$ 则有

$T_{\varphi\psi}=T_{\varphi}T_{\psi}+H_{\overline{\varphi}}^{*}H_{\psi}; H_{\varphi\psi}=H_{\varphi}T_{\psi}+S_{\varphi}H_{\psi}; S_{\varphi\psi}=S_{\varphi}S_{\psi}+H_{\varphi}H_{\overline{\psi}}^{*}. $

$\psi\in H^{\infty}$, 则$H_{\psi}=0$, 故有

$H_{\varphi\psi}=H_{\varphi}T_{\psi}. $

Toeplitz算子理论发展至今, 已形成一个庞大的知识体系[2-4], 在物理学, 概率论, 通信理论及控制论等领域中均有重要的应用, 吸引着相关领域的学者们的关注和兴趣. 更多关于Toeplitz算子的相关问题研究可参考文献[5-9].

$u$是一个非常值内函数, 称$K_{u}^{2}=H^{2}\ominus uH^{2}$为模空间. 显然地, $H^{2}=K_{u}^{2}\oplus uH^{2}, L^{2}=K_{u}^{2}\oplus (K_{u}^{2})^{\bot}=K_{u}^{2}\oplus(uH^{2}\oplus\overline{zH^{2}}).$ 2007 年, Sarason在文献[10] 中引入了模空间$K_{u}^{2}$上的Toeplitz算子, 即截断Toeplitz算子, 引起了众多学者的关注和研究[11-14].

$P_{u}:L^{2}\rightarrow K_{u}^{2}$为正交投影, 则对于$\varphi\in L^{2}$, 符号为$\varphi$的截断Toeplitz算子$A_{\varphi}$定义为

$A_{\varphi}h=P_{u}(\varphi h), \forall h\in K_{u}^{2}\cap L^{\infty}. $

截断Hankel算子$B_{\varphi}$定义为

$ B_{\varphi}h=(I-P_{u})(\varphi h), \forall h\in K_{u}^{2}\cap L^{\infty}. $

2018年, Sang等[15]引入了$(K_{u}^{2})^{\bot}$ 上的Toeplitz算子, 即对偶截断 Toeplitz算子, 给出了两个对偶截断Toeplitz算子的乘积为一个有限秩算子的充要条件. 对偶截断Toeplitz算子$D_{\varphi}$定义为

$ D_{\varphi}x=(I-P_{u})(\varphi x), \forall x\in (K_{u}^{2})^{\bot}\cap L^{\infty}. $

容易得到

$ B_{\varphi}^{\ast} x=P_{u}(\overline{\varphi} x), \forall x\in (K_{u}^{2})^{\bot}\cap L^{\infty}. $

与经典的Hardy空间类似, $L^{2}$上的乘法算子$M_{\varphi}$ 在空间分解$L^{2}=K_{u}^{2}\oplus (K_{u}^{2})^{\bot}$下可表示为如下$2\times 2$算子矩阵形式

$ M_{\varphi}=\left[\begin{array}{cccccc} A_\varphi && B_{\overline{\varphi}}^{*} \\ B_\varphi && D_\varphi \\ \end{array}\right]. $

由于$M_{\varphi}M_{\psi}=M_{\varphi\psi},$ 则有

$ A_{\varphi\psi}=A_{\varphi}A_{\psi}+B_{\overline{\varphi}}^{*}B_{\psi}; B_{\varphi\psi}=B_{\varphi}A_{\psi}+D_{\varphi}B_{\psi}; D_{\varphi\psi}=D_{\varphi}D_{\psi}+B_{\varphi}B_{\overline{\psi}}^{*}. $

在2020年, Qin等[16]引入了调和函数Hardy空间及其上的Toeplitz 算子. 设$u$$v$是内函数且至少有一个是非常值的, 定义调和Hardy空间$H_{u,v}^{2}$

$ H_{u,v}^{2}=uH^{2}\oplus\overline{v}(H^{2})^{\bot}=uH^{2}\oplus\overline{vzH^{2}}. $

特别地, $(H_{u,1}^{2})^{\bot}=K_{u}^{2},$ 于是有 $ K_{u}^{2}\subseteq H^{2}\subseteq H_{1,v}^{2}\subseteq L^{2}. $$M_{u}$$M_{\overline{u}}$$L^{2}$上由$u$$\overline{u}$诱导的乘法算子, 则$Q_{u,v}:=M_{u}PM_{\overline{u}}+M_{\overline{v}}P_{-}M_{v}$$L^{2}\rightarrow H_{u,v}^{2}$上的正交投影. 对于 $\varphi\in L^{2}$, 调和Hardy空间$H_{u,v}^{2}$上符号为$\varphi$的调和Toeplitz算子$\widehat{T}_{\varphi}$定义为

$ \widehat{T}_{\varphi} h=Q_{u,v}(\varphi h), \forall h\in H_{u,v}^{2}. $

调和Hankel算子$\widehat{H}_{\varphi}$定义为

$ \widehat{H}_{\varphi}h=(I-Q_{u,v})(\varphi h), \forall h\in H_{u,v}^{2}. $ 对偶调和Toeplitz算子$\widehat{S}_{\varphi}$定义为 $ \widehat{S}_{\varphi}x=(I-Q_{u,v})(\varphi x),\forall x\in (H_{u,v}^{2})^{\bot}. $

容易得到

$ \widehat{H}_{\varphi}^{\ast} x=Q_{u,v}(\overline{\varphi} x), \forall x\in (H_{u,v}^{2})^{\bot}. $

乘法算子$M_{\varphi}$在空间分解$L^{2}=H_{u,v}^{2}\oplus (H_{u,v}^{2})^{\bot}$下可表示为如下$2\times 2$算子矩阵形式

$ M_{\varphi}=\left[\begin{array}{cccccc} \widehat{T}_\varphi && \widehat{H}_{\overline{\varphi}}^{*} \\ \widehat{H}_\varphi & &\widehat{S}_\varphi \\ \end{array}\right]. $

因为$M_{\varphi}M_{\psi}=M_{\varphi\psi}$, 则有

$ \widehat{T}_{\varphi\psi}=\widehat{T}_{\varphi}\widehat{T}_{\psi}+\widehat{H}_{\overline{\varphi}}^{*}\widehat{H}_{\psi}; \widehat{H}_{\varphi\psi}=\widehat{S}_{\varphi}\widehat{H}_{\psi}+\widehat{H}_{\varphi}\widehat{T}_{\psi}; \widehat{S}_{\varphi\psi}=\widehat{S}_{\varphi}\widehat{S}_{\psi}+\widehat{H}_{\varphi}\widehat{H}_{\overline{\psi}}^{*}. $

本文的内容安排如下: 第二节主要给出了调和Hardy空间上的Toeplitz算子与对偶截断Toeplitz算子的酉等价性, 根据该结论并结合对偶截断Toeplitz算子的相关性质; 本文的第三节给出了两个调和Toeplitz算子的可交换性的充要刻画, 调和Toeplitz代数的性质, 以及调和Toeplitz 算子$\widehat{T}_{z}$的换位子的刻画; 在本文的第四节中, 我们研究并得到了一些特殊符号的有限多个调和Toeplitz算子乘积的谱与紧性的刻画.

2 调和Toeplitz算子的酉等价性

算子的酉等价性在研究算子结构中起着至关重要的作用. 一直以来众多的学者想要在Hilbert空间上研究有界线性算子的酉等价性, 但想要找到两个算子在什么条件下是酉等价是相当困难的. 1985年, 孙顺华[17]证明了Cowen问题之一的答案是否定的; Cima 等[18]证明了所有秩一算子, $2\times2$矩阵和正规算子与截断Toeplitz算子是酉等价的.

本节主要证明了$H_{u,v}^2$上的调和Toeplitz算子与$(K_{uv}^{2})^{\bot}$上的对偶截断Toeplitz算子是酉等价的.

$u$$v$是内函数且至少其中之一是非常值的, 则

$ (K_{uv}^{2})^{\bot}=uvH^{2}+(H^{2})^{\bot}=uvH^{2}+\overline{zH^{2}}=H_{uv,1}^{2}. $

对任意的$h\in H_{u,v}^{2}$, 则$h$可唯一分解为: $h=ux+\bar{v}\overline{zy}$, 其中, $x, y\in H^{2}$. 由于对偶Toeplitz算子的性质已有很多相关结果, 而调和Toeplitz算子是最近由Qin 等[16]引入的, 其性质还有待研究, 故下面我们所给出的$H_{u,v}^2$上的调和Toeplitz算子与$(K_{uv}^{2})^{\bot}$上的对偶截断Toeplitz算子是酉等价的结论是十分有意义的.

定理2.1$f \in L^{\infty}$, $u, v$是内函数且至少其中一个是非常值的, 令$\theta=uv$, 则$H_{u,v}^{2}$上的调和 Toeplitz算子$\widehat{T}_{f}$$(K_{\theta}^{2})^{\bot}$上的对偶截断Toeplitz算子$D_{f}$是酉等价的.

定义算子${\cal A}:H_{u,v}^{2}\rightarrow(K_{uv}^{2})^{\bot}$

$ {\cal A}(ux+\bar{v}\overline{zy})=uvx+\overline{zy}, \forall x, y\in H^{2}. $

容易验证${\cal A}$满足

$ {\cal A}(h_{1}+h_{2})={\cal A}(h_{1})+{\cal A}(h_{2}), {\cal A}(\alpha h)=\alpha {\cal A}(h); \|{\cal A}h\|=\|h\|. $

因此${\cal A}$是有界线性算子. 对于任意的$g=uvx+\overline{zy}\in (K_{uv}^{2})^{\bot}$, 这里$x, y \in H^2$, 则有$h=ux+\bar{v}\overline{zy} \in H_{u,v}^2$满足${\cal A}h=g$, 故${\cal A}$是酉算子且${\cal A}^{*}(uvx+\overline{zy})=ux+\overline{vzy}.$

对于$\forall x, y\in L^{2}$, 有

$\begin{eqnarray*} \widehat{T}_{f}{\cal A}^{*}(uvx+\overline{zy})&=&\widehat{T}_{f}(ux+\bar{v}\overline{zy})\\ &=&Q_{u,v}f(ux+\bar{v}\bar{z}\bar{y})\\ &=&(uP\bar{u}+\bar{v}P_{-}v)f(ux+\bar{v}\bar{z}\bar{y})\\ &=&uPfx+uPf\overline{uvzy}+\bar{v}P_{-}uvfx+\bar{v}P_{-}f\overline{zy}\\ &=&u(Pfx+Pf\overline{uvzy})+\bar{v}P_{-}(uvfx+f\overline{zy}). \\ D_{f}(uvx+\overline{zy})&=&Qf(uvx+\overline{zy})\\ &=&(uvP\overline{uv}+P_{-})f(uvx+\overline{zy})\\ &=&uvPfx+uvP\overline{uv}f\overline{zy}+P_{-}fuvx+P_{-}f\overline{zy}\\ &=&uv(Pfx+Pf\overline{uvzy})+P_{-}(fuvx+f\overline{zy}). \end{eqnarray*}$

于是

$\begin{eqnarray*} {\cal A}^{*}D_{f}(uvx+\overline{zy})&=&{\cal A}^{*}\{uv(Pfx+Pf\overline{uvzy})+P_{-}(fuvx+f\overline{zy})\}\\ &=&u(Pfx+Pf\overline{uvzy})+\bar{v}P_{-}(fuvx+f\overline{zy})\\ &=&\widehat{T}_{f}{\cal A}^{*}(uvx+\overline{zy}), \end{eqnarray*}$

所以$\widehat{T}_{f}{\cal A}^{*}={\cal A}^{*}D_{f},$$\widehat{T}_{f}={\cal A}^{*}D_{f}{\cal A},$$\widehat{T}_{f}$$D_{f}$是酉等价的.证毕.

由上述结论,则显然有以下推论.

推论2.1$f \in L^{\infty}$, $u, v$是内函数且至少其中一个是非常值的,

$H^2_{u,v}$上的调和 Toeplitz 算子$\widehat{T}_{f}$$H^2_{v,u}$上的调和Toeplitz算子$\widehat{\mathsf{T}}_{f}$是酉等价的.

由于$(K_{uv}^2)^{\bot}=(K_{vu}^2)^{\bot}$, 根据定理2.1可知$H^2_{u,v}$上的调和Toeplitz算子$\widehat{T}_{f}$$H^2_{v,u}$上的调和Toeplitz算子$\widehat{\mathsf{T}}_{f}$均与$(K_{uv}^2)^{\bot}$上的对偶截断Toeplitz算子$D_{f}$是酉等价的, 从而$\widehat{T}_{f}$$\widehat{\mathsf{T}}_{f}$是酉等价的. 证毕.

特别地, 当内函数$v=1$时, 结合定理2.1和上述推论则有如下推论.

推论2.2$f \in L^{\infty}$, $u$是非常值的内函数, 则$H_{1,u}^2(=H^2 \oplus \overline{uzH^2})$上的调和 Toeplitz算子$\widehat{\mathsf{T}}_{f}$$(K_u^2)^{\bot}(=uH^2 \oplus \overline{zH^2})$上的对偶截断Toeplitz算子$D_{f}$是酉等价的.

3 若干推论

根据上一节的主要结果以及Sang, Qin和本文的第一作者、第三作者所给出的关于对偶截断Toeplitz算子的一些已有结果[15-16,19-20], 本节我们得到了调和Hardy空间上的调和Toeplitz算子在交换性, 调和Toeplitz代数, 换位子等方面的一些结论. 本节证明全部省略.

根据Sang等[19]关于两个对偶截断Toeplitz算子可交换的充要刻画, 则可得

定理3.1$f, g \in L^{\infty}$, $u, v$是内函数且至少其中一个是非常值的, 则$\widehat{T}_{f}\widehat{T}_{g}=\widehat{T}_{g}\widehat{T}_{f}$当且仅当下述条件之一成立

(1) 存在$\lambda\in {\Bbb C}$, 使得$f, g, \bar{f}(uv-\lambda)$$\bar{g}(uv-\lambda)$均在$H^2$中;

(2) 存在$\lambda\in {\Bbb C}$, 使得$\bar{f}, \bar{g}, f(uv-\lambda)$$g(uv-\lambda)$均在$H^2$中;

(3) 存在$a, b, c \in {\Bbb C}$$|a|+|b| \neq 0$, 使得$af+bg=c$.

$u, v$是内函数且至少其中一个是非常值的, 记$B(H_{u,v}^2)$$H_{u,v}^2$上的所有有界线性算子的全体. 设$\mathfrak{X}$$L^{\infty}$的闭自伴子代数, 记

$ {\cal T}_{\mathfrak{X}}=clos\bigg\{\sum_{i=1}^{n}\prod_{j=1}^{j=n}\widehat{T}_{\varphi_{ij}}: \varphi_{ij} \in \mathfrak{X}\bigg\} $

$B(H_{u,v}^2)$的包含$\{\widehat{T}_{\varphi}: \varphi \in \mathfrak{X}\}$且在范数意义下封闭的最小子代数, 从而${\cal T}_{\mathfrak{X}}$是由$\{\widehat{T}_{\varphi}: \varphi \in \mathfrak{X}\}$所生成的$C^{*}$-代数.

${\cal T}_{\mathfrak{X}}$中由所有的半换位子

$ [\widehat{T}_{\varphi},\widehat{T}_{\psi}):= \widehat{T}_{\varphi}\widehat{T}_{\psi}-\widehat{T}_{\varphi\psi}, \varphi, \psi \in \mathfrak{X} $

所生成的${\cal T}_{\mathfrak{X}}$的一个闭理想称为半换位子理想, 并记为$\mathfrak{S}{\cal T}_{\mathfrak{X}}$.

${\cal T}_{\mathfrak{X}}$中由所有的换位子

$ [\widehat{T}_{\varphi},\widehat{T}_{\psi}]:= \widehat{T}_{\varphi}\widehat{T}_{\psi}-\widehat{T}_{\psi}\widehat{T}_{\varphi}, \varphi, \psi \in \mathfrak{X} $

所生成的${\cal T}_{\mathfrak{X}}$的一个闭理想称为换位子理想, 并记为$\mathfrak{C}{\cal T}_{\mathfrak{X}}$.

根据Sang等[9]所给出的对偶截断Toeplitz $C^{*}$ -代数的刻画以及调和Toeplitz算子和对偶截断Toeplitz算子的酉等价性, 我们得到了如下关于调和Toeplitz代数${\cal T}_{L^{\infty}}$${\cal T}_{C({\Bbb T})}$的结构定理.

定理3.2 序列

$ 0 \rightarrow \mathfrak{S}{\cal T}_{L^{\infty}} \rightarrow {\cal T}_{L^{\infty}} \rightarrow L^{\infty} \rightarrow 0 $

是一个短正合列, 即, 商代数${\cal T}_{L^{\infty}}/\mathfrak{S}{\cal T}_{L^{\infty}}$$L^{\infty}$$*$-等距同构的.

定理3.3 序列

$ 0 \rightarrow {\cal K} \rightarrow {\cal T}_{C({\Bbb T})} \rightarrow C({\Bbb T}) \rightarrow 0 $

是一个短正合列, 其中${\cal K}$$H_{u,v}^2$上的所有紧算子全体.

记Hilbert空间$H$上所有与算子$T$可交换的算子的集合为$\{T\}^{'}$, 即

$\{T\}^{'}=\{S \in B(H): TS=ST\}, $

其中$B(H)$为Hilbert空间$H$上的所有有界线性算子全体.

下面根据Sang等[20]给出的关于对偶截断Toeplitz 算子$D_z$的换位子的刻画, 我们可以得到$H_{u,v}^2$ 上的调和移位Toeplitz算子$\widehat{T}_z$ 的换位子的刻画如下, 其中${\cal A}$为定理2.1的证明中所定义的酉算子.

定理3.4$u, v$是内函数且至少其中一个是非常值的. 令$\theta=uv$.

(1) 若$\theta(0)=0$, 则

$ \{\widehat{T}_z\}^{'}=\left\{ {\cal A}^{*}\left[\begin{array}{ccccc} t_{\varphi} & &0 \\ b_{\bar{\theta}f} && S_{\psi} \end{array}\right]{\cal A}: \varphi, \psi \in H^{\infty}, \, f \in L^{\infty} \right\}, $

这里上述矩阵是在分解$H_{u,v}^2=uH^2\oplus \bar{v}\overline{zH^2}$下给出的, 其中, 对任意的$x \in \theta H^2$, $t_{\varphi}x=M_{\theta}PM_{\bar{\theta}}(\varphi x)$称为定义在$\theta H^2$上的小Toeplitz算子, $b_{f\bar{\theta}}x=P_{-}(f\bar{\theta}x)$ 称为定义在$\theta H^2$上的小Hankel算子;

(2)若$\theta(0)\neq 0$, 则

$ \{\widehat{T}_z\}^{'}=\left\{ {\cal A}^{*}\left[\begin{array}{ccccc} t_{\varphi} && \bar{\theta}(0)b_{\overline{\theta\varphi}}^{*}\\ \frac{1}{\bar{\theta}(0)}b_{\bar{\theta}\varphi} && S_{\varphi} \end{array}\right]{\cal A}: \varphi \in L^{\infty} \right\}. $

而且, 进一步地, 可以得到如下结论.

定理3.5$u, v$为非常值的内函数, 对$H_{u,v}^2$上的调和Toeplitz算子则有

$ \{\widehat{T}_{z^n}: n \in {\Bbb N}\}^{'}=\left\{ cI: \, c \in {\Bbb C} \right\}, $

更一般地, 则有

$ \{\widehat{T}_{f}: f \in H^{\infty}\}^{'}=\left\{ cI: \, c \in {\Bbb C} \right\}. $

4 调和Toeplitz算子的谱

本节我们主要给出了连续符号的调和Toeplitz算子与乘法算子的关系以及一些特殊符号的有限个调和Toeplitz算子的乘积$\widehat{T}_{\varphi_{1}}\widehat{T}_{\varphi_{2}}\dots\widehat{T}_{\varphi_{n}}$ 的本质谱刻画.

定义$L^{2}=H^{2}\oplus\overline{zH^{2}}\rightarrow [K_{u}^{2}]^{\bot}=uH^{2}\oplus\overline{zH^{2}}$上的算子

$U_{u}=\left[\begin{array}{cccccc} M_{u}& & 0 \\ 0 & &I_{\overline{zH^{2}}} \\ \end{array}\right]. $

其中, $M_{u}$$H^{2}$上的乘法算子, $I_{\overline{zH^{2}}}$$\overline{zH^{2}}$上的恒等算子. 易知, $U_{u}$是一个酉算子. 下述的$2\times2$矩阵表示表明了对偶截断Toeplitz算子与Toeplitz算子, Hankel 算子以及截断 Toeplitz算子有着密不可分的联系.

引理4.1[19]$\varphi\in L^{\infty}$, 则在空间分解$L^{2}=H^{2}\oplus\overline{zH^{2}}$

$ \widetilde{D}_{\varphi}=U_{u}^{\ast}D_{\varphi}U_{u}=\left[\begin{array}{cccccc} T_{\varphi}& & H_{u\bar{\varphi}}^{\ast} \\ H_{u\varphi}& & S_{\varphi} \\ \end{array}\right]. $

$\varphi\in H^{2}$, 则有

$ \widetilde{D}_{\varphi}=U_{u}^{\ast}D_{\varphi}U_{u}=\left[\begin{array}{cccccc} T_{\varphi}& & H_{u\bar{\varphi}}^{\ast} \\ 0 && S_{\varphi} \\ \end{array}\right]. $

推论4.1$u, v$为内函数且至少其中一个是非常值的, $\varphi$$\partial{\Bbb D}$上的连续函数, 则$H_{u, v}^2$上的调和Toeplitz 算子$\widehat{T}_{\varphi}$模掉紧算子后与$M_{\varphi}$是酉等价的.

$\theta=uv$, 显然$\theta$为内函数. 因为$\varphi$$\partial{\Bbb D}$上连续, 由文献[3] 可知$H_{\varphi}$$H_{\overline{\varphi}}^{*}$ 是紧的, 则有$H_{\theta\varphi}(=H_{\varphi}T_{\theta})$$H_{\theta\bar{\varphi}}^{\ast} (=T_{\bar{\theta}}H_{\bar{\varphi}}^{\ast})$ 也是紧的. 由于

$ M_{\varphi}=\left[\begin{array}{cccccc} T_\varphi& & H_{\overline{\varphi}}^{*} \\ H_\varphi && S_\varphi \\ \end{array}\right], $

则对于$(K_{\theta}^{2})^{\bot}$上的对偶截断Toeplitz算子$D_{\varphi}$, 结合引理4.1可得

$ U_{\theta}^{\ast}D_{\varphi}U_{\theta}-M_{\varphi}=\left[\begin{array}{cccccc} 0 && H_{\theta\bar{\varphi}-\bar{\varphi}}^{\ast} \\ H_{\theta\varphi-\varphi}& & 0 \\ \end{array}\right] $

是一个紧算子, 故$D_{\varphi}$模掉紧算子后与$M_{\varphi}$是酉等价的, 即$U_{\theta}^{*}D_{\varphi}U_{\theta}=M_{\varphi}{\rm mod} ({\cal K})$, 这里${\cal K}$表示$L^2$上的紧算子全体. 由定理2.1知$D_{\varphi}={\cal A}\widehat{T}_{\varphi}{\cal A}^{*}$, 因此

$ ({\cal A}^{*}U_{\theta})^{*}\widehat{T}_{\varphi}({\cal A}^{*}U_{\theta})=M_{\varphi}{\rm mod} ({\cal K}), $

显然${\cal A}^{*}U_{\theta}$是酉算子, 故$\widehat{T}_{\varphi}$模掉紧算子后与$M_{\varphi}$是酉等价的.证毕.

下述推论可直接由定理3.3得到. 为了文章的完整性, 这里我们给出其另一种证明.

推论4.2$u, v$为内函数且至少其中一个是非常值的, $\varphi_{i} (i=1, 2,\cdots, n)$$\partial{\Bbb D}$上的连续函数, 对于$H_{u, v}^2$上的调和 Toeplitz算子$\widehat{T}_{\varphi_{i}} (i=1, 2,\cdots, n)$, 则有

$ \sigma_{e}(\widehat{T}_{\varphi_{1}}\widehat{T}_{\varphi_{2}}\cdots \widehat{T}_{\varphi_{n}})={\rm R}(\varphi_{1}\varphi_{2}\cdots \varphi_{n}). $

其中, $\sigma_{e}(T)$表示算子$T$的本质谱, ${\rm R}(\varphi)$表示$\varphi$的本质值域.

$\theta=uv$$U={\cal A}^{*}U_{\theta}$, 记$\widetilde{T}_{\varphi}=U^{*}\widehat{T}_{\varphi}U$. 由推论4.1知: 若$\varphi$$\partial{\Bbb D}$上连续, 则 $\widetilde{T}_{\varphi}=M_{\varphi}+K,$ 这里$K$$L^2$上的紧算子. 故可设$\widetilde{T}_{\varphi_{i}}=M_{\varphi_{i}}+K_{i}$, 其中$K_{i} (i=1, 2,\cdots, n)$是紧算子, 则

$\begin{eqnarray*} \widetilde{T}_{\varphi_{1}}\widetilde{T}_{\varphi_{2}}\cdots \widetilde{T}_{\varphi_{n}} =(M_{\varphi_{1}}+K_{1})(M_{\varphi_{2}}+K_{2})\cdots (M_{\varphi_{n}}+K_{n}) =M_{\varphi_{1}\varphi_{2}\cdots \varphi_{n}}+\widetilde{K}, \end{eqnarray*}$ 这里$\widetilde{K}$是紧算子, 所以 $ \sigma_{e}(\widetilde{T}_{\varphi_{1}}\widetilde{T}_{\varphi_{2}}\cdots \widetilde{T}_{\varphi_{n}}) =\sigma_{e}( M_{\varphi_{1}\varphi_{2}\cdots \varphi_{n}}) ={\rm R}(\varphi_{1}\varphi_{2}\cdots \varphi_{n}). $

$\begin{eqnarray*} \widetilde{T}_{\varphi_{1}}\widetilde{T}_{\varphi_{2}}\cdots \widetilde{T}_{\varphi_{n}} = (U^{*}\widehat{T}_{\varphi_{1}}U)(U^{*}\widehat{T}_{\varphi_{2}}U)\cdots (U^{*}\widehat{T}_{\varphi_{n}}U) =U^{*}(\widehat{T}_{\varphi_{1}}\widehat{T}_{\varphi_{2}}\cdots \widehat{T}_{\varphi_{n}})U. \end{eqnarray*}$ 于是 $\sigma_{e}(\widehat{T}_{\varphi_{1}}\widehat{T}_{\varphi_{2}}\cdots \widehat{T}_{\varphi_{n}})={\rm R}(\varphi_{1}\varphi_{2}\cdots \varphi_{n}).$ 证毕.

参考文献

Sarason D.

Algebras of functions on the unit circle

Bulletin of the American Mathematical Society, 1973, 79(2): 286-299

DOI:10.1090/S0002-9904-1973-13144-1      URL     [本文引用: 1]

Böttcher A, Silbermann B.

Analysis of Toeplitz Operators

Heidelberg: Springer-Verlag, 2005

[本文引用: 1]

Peller V. Hankel Operators and Their Applications. New York: Springer, 2002

[本文引用: 2]

Upmeier H. Toeplitz Operators and Index Theory in Several Complex Variables. Boston: Birkhäuser, 1996

[本文引用: 1]

刘元, 丁宣浩.

重调和Hardy空间上的Toeplitz算子

中国科学: 数学, 2013, 43(7): 675-690

DOI:10.1360/012012-292      URL     [本文引用: 1]

Liu Y, Ding X.

Toeplitz operators on the reharmonic Hardy spaces

Scientia Sinica Mathematica, 2013, 43(7): 675-690 (in Chinese)

DOI:10.1360/012012-292      URL     [本文引用: 1]

Guediri H.

Dual Toeplitz operators on the sphere

Acta Mathematica Sinica, 2013, 29: 1791-1808

[本文引用: 1]

陈泳, 于涛.

单位球上本质交换的对偶Toeplitz 算子

数学进展, 2009, 38(4): 453-464

[本文引用: 1]

Chen Y, Yu T.

Dual Toeplitz operators with essential exchange on the unit ball

Advances In Mathematics, 2009, 38(4): 453-464 (in Chinese)

[本文引用: 1]

Sarason D.

A remark on the Volterra operator

Journal of Mathematical Analysis and Applications, 1965, 12: 244-246

DOI:10.1016/0022-247X(65)90035-1      URL     [本文引用: 1]

Sang Y Q, Qin Y S, Ding X H.

Dual truncated Toeplitz $C^{*}$-algebras

Banach Journal of Mathematical Analysis, 2019, 13(2): 275-292

DOI:10.1215/17358787-2018-0030      URL     [本文引用: 2]

Sarason D.

Algebraic properties of truncated Toeplitz operators

Operators and Matrices, 2007, 1: 491-526

[本文引用: 1]

Baranov A, Bessonov R, Kapustin V.

Symbols of truncated Toeplitz operators

Journal of Functional Analysis, 2011, 261: 3437-3456

DOI:10.1016/j.jfa.2011.08.005      URL     [本文引用: 1]

Nagy B, Foias C, Bercovici H, Kérchy L. Harmonic Analysis of Operators on Hilbert Space. New York: Springer, 2010

[本文引用: 1]

Yang J Y, Lu Y F.

Commuting dual Toeplitz operators on the harmonic Bergman space

Science China Mathematics, 2015, 58(7): 1461-1472

DOI:10.1007/s11425-014-4940-x      URL     [本文引用: 1]

Lakhdar B, Hocine G.

Properties of dual Toeplitz operators with applications to haplitz products on the Hardy space of the polydisk

Taiwanese Journal of Mathematics, 2015, 19(1): 31-49

[本文引用: 1]

Ding X H, Sang Y Q.

Dual truncated Toeplitz operators

Journal of Mathematical Analysis and Applications, 2018, 461(1) : 929-946

DOI:10.1016/j.jmaa.2017.12.032      URL     [本文引用: 2]

Ding X H, Qin Y S, Sang Y Q.

Harmonic Hardy space and their operators

Operators and Matrices, 2020, 14(4): 837-855

[本文引用: 3]

孙顺华.

Toeplitz算子酉等价的某些问题

科学通报, 1985, 8: 570-572

[本文引用: 1]

Sun S.

Some problems on unitary equivalence of Toeplitz operators

Chinese Science Bulletin, 1985, 8: 570-572 (in Chinese)

[本文引用: 1]

Cima J, Matheson A, Ross W, Wogen W.

Truncated Toeplitz operators: spatial isomorphism, unitary equivalence, and similarity

Indiana University Mathematics Journal, 2010, 59(2): 595-620

DOI:10.1512/iumj.2010.59.4097      URL     [本文引用: 1]

Sang Y Q, Qin Y S, Ding X H.

A theorem of Brown-Halmos type for dual truncated Toeplitz operators

Annals of Function Analysis, 2020, 11: 271-284

DOI:10.1007/s43034-019-00002-7      URL     [本文引用: 3]

Li Y N, Ding X H, Sang Y Q.

The commutant and invariant subspaces for dual truncated Toeplitz operators

Banach Journal of Mathematical Analysis, 2021, 15: 224-250

[本文引用: 2]

/