Realization and smoothness related to the Laplacian (Q1848152): Difference between revisions

For \(f\in L_p(T^d)\), where \(T=[-\pi,\pi]\), the authors define a modulus of smoothness by \[ \omega_{\Delta}(f,t)_p=\sup_{|h|\leq t}\Bigl\|2df(\cdot)-\sum_{i=1}^d\left(f(\cdot+he_i)+f(\cdot-he_i)\right)\Bigr\|_p, \] where \(\{e_1,\dots,e_d\}\) is the canonical basis of \(\mathbb R^d\), and they show that \[ \omega_{\Delta}(f,\frac 1n)_p\approx\|f-T_n\|_p+n^{-2}\|\Delta T_n\|_p, \] where \(T_n\) is the best trigonometric polynomial approximant of degree \(n\) (with respect to each variable) to \(f\) in \(L_p\) and \(\Delta\) is the Laplacian. In the paper, an important role is played by the inequality \[ \|D^{\alpha}t_n\|_{L_p(T^d)}\leq C(d,m,p)n^{m-2}\|\Delta t_n\|_p \] for \(|\alpha|=m>2\), \(\frac{d}{d-2+m}<p\leq\infty\), where \(t_n\) is any trigonometric polynomial of degree \(n\), which is interesting in itself. The authors also show that the last inequality fails for \(0<p\leq\frac{d}{d-2+m}\).