Spectral Theories, IV, p. 339
Let n∈N. Endow Rn with Lebesgue measure μ and identify Rn with its dual by the mapping (x,y)↦ exp(2iπx⋅y).
a) Show that for every function f∈S(Rn), the mapping g↦f∗g is an automorphism of locally convex spaces of S(Rn).
b) Let f∈S′(Rn) and g∈S(Rn). Show that the mapping φ↦⟨f,gˇ∗φ⟩ is a tempered distribution, called the convolution of f and g. If f∈S(Rn), it coincides with the distribution associated with f∗g. This tempered distribution will again be denoted by f∗g.
In the questions which follow, fix f∈S′(Rn) and g∈S(Rn)
c) Show that f∗g∈C∞(R) and
∂α(f∗g)=∂α(f)∗g=f∗∂α(g)
for every α∈Nn.
d) For every y∈Rn, let gy denote the function x↦g(y−x) on Rn. Show that the function f∗g satisfies
(f∗g)(y)=⟨f,gy⟩
for every y∈Rn.
e) Let h∈S(Rn). Show that (f∗g)∗h=f∗(g∗h).
f ) Show that the function f∗g has polynomial growth, as do all its derivatives. (Use exercise 25.)
g) Show that F(f∗g)=F(f)F(g).
h) Show that the Fourier transform of fg is an infinitely differentiable function on Rn and has polynomial growth, as do all its derivatives. Moreover, for every y∈Rn, one has
F(fg)(y)=⟨f,ge−y⟩
where ey denotes the character x↦ exp(2iπx⋅y) of Rn.
Nothing.
Nothing.