Introduction to \(L_p\) Sobolev spaces via Muramatu's integral formula (Q2401903)

The author gives a survey of the use of Muramatu's integral formula for the study of Sobolev spaces on \(\mathbb R^d\) or on special Lipschitz domains. He notes several advantages of this formula: it is simple and easy to use since it reduces most questions to things that can be handled by Hölder- or Minkowski-type inequalities, it suggests alternative results once you understand a given result and it works for Sobolev spaces of negative order. In this paper, he mainly limits himself to Sobolev spaces of integral order, but the results apply more generally and to Besov spaces as well. The formula itself is very straightforward. For the case of \(\mathbb R^d\), there are functions \(K_{\alpha}(x), \phi(x) \in C_{0}^{\infty}(\mathbb R^d)\) such that \[ f(x) = \int_0^1 \sum_{|\alpha| = m} \left[ \int_{\mathbb R^d} t^{-d} \partial^{\alpha} K_{\alpha}\left( \frac{x - y}{t} \right) f(y) \, dy \right] \frac{dt}{t} + \phi * f(x). \] The formula is only slightly more complicated (the support of the \(K_{\alpha}\) changes) when the domain is a special Lipschitz domain. The author discusses its use for the proof of embedding theorems (Sobolev spaces into \(L_p\) spaces), extension and restriction theorems, and real and complex interpolation theorems for Sobolev spaces. The method elucidated in this paper deserves to be better known. While it has many potential applications, the main examples elucidated are well-known results and those results take their classical forms. It would be interesting to see if this approach gives new results for the real and complex interpolation of Besov spaces. The term \(\phi*f\) in Muramatu's formula makes it seem most useful for inhomogeneous spaces, and an analogous formula for homogeneous spaces could be useful.