Bourbaki

Lemma 4

Tag 02ID, permanent. Cite it as [Bourbaki, Tag 02ID].

Théories spectrales, chapter II, § 1. Transformation de Fourier

There exists a unique measure ν\nu on G^\widehat{G} such that

μf=F(f)⋅ν \mu_f=\mathscr{F}(f)\cdot \nu

for every f∈A(G)f\in A(G). For f∈A(G)f\in A(G), one has F(f)∈L1(G^,ν)∩L2(G^,ν)\mathscr{F}(f)\in L^1(\widehat{G}, \nu )\cap L^2(\widehat{G}, \nu ).

Let f∈A(G)f\in A(G). Let Ωf\Omega_f denote the open set in G^\widehat{G} formed by the χ∈G^\chi \in \widehat{G} such that F(f)(χ)≠0\mathscr{F}(f)(\chi )\not= 0. Let φ\varphi be the characteristic function of G^−Ωf\widehat{G}-\Omega_f. Then for every g∈A(G)g\in A(G), one has

∫G^F(g)d(φ⋅μf)=∫G^φF(f)dμg=0 \int_{\widehat{G}}\mathscr{F}(g)d(\varphi \cdot \mu_f) =\int_{\widehat{G}}\varphi \mathscr{F}(f)d\mu_g= 0

in view of formula (18).

By corollary 1 of II, p. 212, the image A^(G)\widehat{A}(G) of A(G) under the Fourier transform is dense in C0(G^)\mathscr{C}_0(\widehat{G}). One then deduces from the preceding formula that φ⋅μf=0\varphi \cdot \mu_f= 0, hence that μf\mu_f is concentrated on Ωf\Omega_f (INT, IV, §4, no7^o7, def. 4). Let νf\nu_f be the measure on Ωf\Omega_f with density F(f)−1\mathscr{F}(f)^{-1} with respect to μf∣Ωf\mu_f|\Omega_f.

The sets Ωf\Omega_f, for f∈A(G)f\in A(G), form an open covering of G^\widehat{G} (cor. 2). For all ff and gg in A(G), formula (18) shows that νf∣(Ωf∩Ωg)=νg∣(Ωf∩Ωg)\nu_f|(\Omega_f\cap \Omega_g) =\nu_g|(\Omega_f\cap \Omega_g). Consequently, there exists a unique measure ν\nu on G^\widehat{G} such that one has νf=ν∣Ωf\nu_f=\nu |\Omega_f for every f∈A(G)f\in A(G) (INT, III, §2, No.o1^o1, prop. 1).

If f∈A(G)f\in A(G), the measures μf\mu_f and F(f)⋅ν\mathscr{F}(f)\cdot \nu are concentrated on Ωf\Omega_f, and their restrictions to Ωf\Omega_f are equal to F(f)⋅νf\mathscr{F}(f)\cdot \nu_f; these measures are therefore equal.

Since μf\mu_f is a bounded measure, the Fourier transform F(f)\mathscr{F}(f) belongs to the space L1(G^,ν)L^1(\widehat{G}, \nu ). Since moreover F(f)\mathscr{F}(f) belongs to C0(G^)\mathscr{C}_0(\widehat{G}), one has also F(f)∈L2(G^,ν)\mathscr{F}(f)\in L^2(\widehat{G}, \nu ).

Formula (17) now takes the form, for φ∈\varphi \in Stell(G) and f∈A(G)f\in A(G): (20) (φ∗f)(e)=∫G^F(φ)F(f)dν(\varphi *f)(e) =\int_{\widehat{G}}\mathscr{F}(\varphi )\mathscr{F}(f)d\nu.

In particular, for ff and gg in A(G), one has

∫G^F(f)F(g)dν=(f∗g)(e)=∫Gf(x)g(x−1)dx(21) \int_{\widehat{G}}\mathscr{F}(f)\mathscr{F}(g)d\nu = (f*g)(e) =\int_Gf(x)g(x^{-1})dx \tag{21}

Bourbaki sets the proof directly after the statement, and the corpus does not separate them, so what is above is the statement and its proof together.

Cited by

Nothing.

Cites

Nothing.

Machine translation of the English, by gpt-5.4, gpt-5-6-mini, against glossary version 34. Not checked by a person. That model is a cut down version of the one the rest of this language was written by.