Approximation of differential polynomials on classes of vector functions (Q1177809)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Approximation of differential polynomials on classes of vector functions
scientific article

    Statements

    Approximation of differential polynomials on classes of vector functions (English)
    0 references
    0 references
    26 June 1992
    0 references
    Let \(H\) be a separable Hilbert space, \(f: \mathbb{R}^ m\to H\) and \[ E(N):=\inf_{T\in{\mathcal L}_ p(N)}\sup_{f\in BW^{\mathcal P}_ p}\| Q(D)f-Tf\|_ p,\tag{1} \] where \(\|\cdot\|_ p\) \((1\leq p\leq\infty)\) is the norm of the space \(L_ p:=L_ p(\mathbb{R}^ m,H)\), \({\mathcal L}_ p(N)\) is the set of the bounded linear operators \(T: L_ p\to L_ p\), \(\| T\|\leq N\in R^*_ +:=(0,\infty)\) and \(BW^{\mathcal P}_ p:=\left\{f\in L_ p: \sum^ n_{j=1}\| P_ j(D)f\|_ p<1\right\}\) with \(P(D)f:=\sum{\alpha\in A_ P}b_ \alpha D^ \alpha f\) where \[ D^ \alpha:=((\partial_ 1/i)^{\alpha_ 1},\dots,(\partial_ m/i)^{\alpha_ m}) \quad A_ P:=\{\alpha: \alpha=(\alpha_ 1,\dots,\alpha_ m)\in\mathbb{Z}^ m_ +\} \] and, for every \(\alpha\in A_ p\), \(0\neq b_ \alpha: H\to H\) is a bounded linear operator. The main result is the following theorem. Theorem 1. If \(1<p<\infty\), \({\mathcal P}:=\{P_ 1,\dots,P_ n\}\) where \(P_ i\), \(1\leq i\leq n\), are differentiable polynomial functions whose coefficients are operators of Hilbert-Schmidt type and \(Q(D)\) is a differentiable polynomial function, where the coefficients are operators then the following assertions are equivalent: I) There is \(C>0\) such that \(1+p(\xi)\geq C\sum_{\alpha\in\overline A_ p}|\xi^ \alpha|\), \(\xi\in\mathbb{R}^ m\), where \(p:=Tr.\sum^ n_{j=1}(P^*_ jP_ j)\). II) The inclusion \(A_ Q\subset\overline A_{\mathcal P}\) is equivalent to the following assertion: ``There exists \(N>0\) such that the best possible approximation (1) is a real number''. III) The inclusion II) is equivalent to the following inequality \(\| Q(D)f\|_ p\leq C\left(\| f\|_ p+\sum^ n_{j=1}\| P_ j(D)f\|_ p\right)\), where \(C>0\) is independent of \(f\in W^{\mathcal P}_ P:=\{f\in L_ p; P_ j(D)f\in L_ p, j\in 1,\dots,n\}\).
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references