next up previous contents
Next: Case 5 Up: Homogeneity of distributions Previous: Case 3   Contents

Case 4

Consider the function $u(\xi) := \vert\xi\vert^{-n}$ for $\xi \neq 0$. (By averaging $v(\om)$ over $\bS^{n-1}$, one can see that any smooth $(-n)$-homogeneous function on $\bR^n \setminus \{0\}$ is a linear combination of $\vert\xi\vert^{-n}$ and a function in Case 3.

We can try the cutoff regularization, anyway. Let $R_f u$ be given by the recipe of ([*]):

\begin{displaymath}
\dst{R_f u}{\phi} :=
\int_{\bR^n} u(\xi)\bigl(\phi(\xi) - \phi(0) f(\vert\xi\vert)\bigr)  d^n\xi.
\end{displaymath} (22)

However, in the present case, $R_f u$ is not homogeneous!


\begin{lem}
If $\dl\: \phi \mapsto \phi(0)$ is the Dirac delta, and if
$u(\xi) ...
..._f u)_t - t^{-n}  R_f u = (\Om_n  t^{-n} \log t) \dl.
\end{equation}\end{lem}


\begin{proof}
We compute $\dst{(R_f u)_t - t^{-n}R_f u}{\phi}$ for
$\phi \in \...
...m measures the failure of homogeneity of the
regularization $R_fu$.
\end{proof}



Pawel Witkowski 2006-03-14