Fourier multiplier theorems involving type and cotype (Q1746607)

This paper deals with the theory of vector-valued Fourier multiplier operators $T_m:\ L^p(\mathbb{R}^d; X)\to L^q(\mathbb{R}^d; Y)$ with $1\le p\le q\le \infty$ and $m:\ \mathbb{R}^d\to \mathcal{L}(X,Y)$ being operator symbols, where $X$ and $Y$ are Banach spaces, $\mathcal{L}(X,Y)$ denotes the set of all linear bounded operators from $X$ to $Y$. These vector-valued Fourier multipliers have been extensively studied in the case $p=q$, but far less is known for $p <q$. In the scalar setting one can deduce results for $p < q$ from the case $p = q$; but to do this in the vector-valued setting, one needs some additional restrictions such as the smoothness of the symbols $m$ and the UMD assumption of $X$ and $Y$. \par In this paper the authors study the boundedness of $T_m$ from $L^p(\mathbb{R}^d; X)$ to $L^q(\mathbb{R}^d; Y)$ with $p < q$, and their main result reads as follows: let $X$ be a Banach space with type $p_0\in(1,2]$ and $Y$ a Banach space with cotype $q_0\in[2,\infty)$, $p\in (1,p_0)$, $q\in (q_0,\infty)$ and $r=\frac{pq}{q-p}$. If the symbol $m:\ \mathbb{R}^d\to \mathcal{L}(X,Y)$ is $X$-strongly measurable and the set \[ \{|\xi|^{\frac dr}m(\xi)|\ \xi\in \mathbb{R}^d\setminus\{0\}\}\subset \mathcal{L}(X,Y) \] is $\gamma$-bounded, then $T_m$ is bounded from $L^p(\mathbb{R}^d; X)$ to $L^q(\mathbb{R}^d; Y)$. Recall that a collection $\mathcal{T}\subset \mathcal{L}(X, Y )$ is $\gamma$-bounded if there exists a constant $C>0$ such that \[ \left(\mathbb{E}\left\|\sum_{k=1}^n \gamma_k T_kx_k\right\|^2_Y\right)^{\frac12}\le C\left(\mathbb{E}\left\|\sum_{k=1}^n \gamma_kx_k\right\|^2_X\right)^{\frac12} \] for any $n\in\mathbb{N}$, $T_1,\ldots T_n\in \mathcal{T}$ and $x_1,\ldots,x_n\subset X$. The $\gamma$-boundedness condition is weaker than the classical $R$-boundedness condition. Another highlight of the above result is that no smoothness condition on $m$ is required. A variant of this result for $p$-convex and $q$-concave Banach lattices is also obtained. \par Under some smoothness conditions on $m$, the authors further prove that the boundedness result of $T_m$ can be extrapolated to other values of $p$ and $q$ as long as $\frac1p-\frac1q$ remains constant.