Operator-valued dyadic shifts and the \(T(1)\) theorem (Q293064): Difference between revisions

Let \(p\in(1,\infty)\) and \(N\in\mathbb{N}\). A Banach space \((E,|\cdot|_E)\) is said to be a UMD (unconditional martingale difference) space, if there exists a positive constant \(\beta_p(E)\) such that, for all \(E\)-valued \(L^p\)-martingale difference sequences \(\{d_n\}_{n=1}^N\) and choices of signs \(\{\epsilon_n\}_{n=1}^N\in\{-1,1\}^N\), \[ \left\|\sum_{n=1}^N\epsilon_n d_n\right\|_{L^p(E)}\leq\beta_p(E) \left\|\sum_{n=1}^Nd_n\right\|_{L^p(E)}. \] Let \(\mathcal{D}\) be a collection of dyadic cubes in \(\mathbb{R}^d\). Then the Haar projection \(D_I\) associated with a dyadic cube \(I\in\mathcal{D}\) is defined by \[ D_If:=\sum_{I'\in\mathrm{child}(I)}\langle f\rangle_{I'}\chi_{I'}- \langle f\rangle_{I}\chi_{I}, \quad\quad \langle f\rangle_{I}:=\frac{1}{|I|}\int_{I}f\,dx, \] where \(\mathrm{child}(I)\) and \(\chi_I\) denote the dyadic children of \(I\) and the characteristic function of \(I\), respectively. Moreover, the shifted Haar projection \(D^i_K\) associated with a dyadic cube \(K\in\mathcal{D}\) is defined by \[ D^i_Kf:=\sum_{I\in\mathcal{D}: I\subset K,\,\ell(I)=2^{-i}\ell(K)}D_I f, \] where \(\ell(I)\) denotes the sidelength of \(I\). Assume that \(E\) is a UMD space. Then an operator-valued dyadic shift \(S^{ji}\), associated with parameters \(j\) and \(i\) and operator-valued kernels \(a_K\), is defined, for every locally integrable function \(f: \mathbb{R}^d\rightarrow E\), by \[ S^{ji}f:=\sum_{K}D^j_K A_K D^i_K f, \] where, for any \(K\in\mathcal{D}\), the averaging operator \(A_K\) associated with the operator-valued kernel \(a_K:\;\mathbb{R}^d\times\mathbb{R}^d\rightarrow L(E)\) is defined by \[ A_K f(x):=\frac{\chi_K(x)}{|K|}\int_{K}a_K(x,x')f(x')\,dx. \] In this article, the authors proved that the operator-valued dyadic shift defined as above is bounded on the space \(L^p(\mathbb{R}^d;E)\). Moreover, the authors extended the dyadic representation theorem, which states that every scalar-valued Calderón-Zygmund operator can be represented as a series of dyadic shifts and paraproducts averaged over randomized dyadic systems, to operator-valued Calderón-Zygmund operators. As a corollary, the authors further gave another proof of the operator-valued global \(T1\) theorem. Furthermore, the authors obtained new proofs for the following well-known theorems: the boundedness of the dyadic (operator-valued) paraproduct, a variant of Pythagoras' theorem for (operator-valued) functions adapted to a sparse collection of dyadic cubes, and a decoupling inequality for (UMD-valued) martingale differences.