Integration, IV, p. 139
Bourbaki marks this one as harder than the rest.
Let X be a locally compact space, μ a positive measure on X, and f a μ-measurable numerical function.
a) Let (an)n∈Z be a sequence of real numbers, λ a finite real number; for every n∈Z, set
un(x)=an+⌊λlog∣f(x)∣⌋f(x)if f(x)=0 and f(x)=±∞un(x)=0for the other values of x∈X (where
[t] denotes the integral part of the finite real number
t). Show that the
un are
μ-measurable and that, for every finite real number
c such that
cλ<1,
n=−∞∑+∞∫∗∣un(x)∣ecxdμ(x)≤e∣c∣∫∗∣f(x)∣1−cλdμ(x)⋅(n=−∞∑+∞ecn∣an∣).
b) Let p′,p′′ be two finite numbers >0, t a real number such that 0<t<1, and let p be the real number defined by 1/p=(1−t)/p′+t/p′′. For every number α>0, set
1/Kα=inf(n=−∞∑+∞∣eαtnan∣p′)(1−t)/p′(n=−∞∑+∞∣e−α(1−t)nan∣p′′)t/p′′, the infimum being taken over all the absolutely convergent series
(an)n∈Z of finite real numbers such that
∑n=−∞+∞an=1. Show that (
) Np(f)≤Kα⋅infF(u) (*)
Np(f)≥Kαe−α⋅infF(u), where
u=(un)n∈Z runs over the set of sequence of functions belonging to
Lp′∩Lp′′ such that the series
(un(x))n∈Z is almost everywhere absolutely convergent with sum
f(x); for every sequence
u having these properties, one sets
F(u)=(n=−∞∑+∞(Np′(eαtnun))p′)(1−t)/p′(n=−∞∑+∞(Np′′(e−α(1−t)nun))p′′)t/p′′ and in the formulas (
) and (*) the infimum is taken over the set of sequences
u=(un) having the preceding properties. (To prove (
), use Hölder’s inequality; to prove (*), use a) twice with suitable choices of
c and
λ.) In particular,
Kα is finite for all
α>0.
c) Let p′,p′′,q′,q′′ be finite numbers ≥1, t a number such that 0<t<1, and let
p1=p′1−t+p′′t,q1=q′1−t+q′′t. Let
Y be a second locally compact space,
ν a positive measure on
Y, and let
w be a linear mapping of
H(X;R) into the vector space (not topological)
S(Y,ν;R) of
ν-measurable finite numerical functions on
Y. Suppose that:
1∘ w maps
H(X;R) into
Lq′(Y;R)∩Lq′′(Y;R);
2∘ for every function
f∈H(X;R),
Nq′(w(f))≤M′Np′(f)andNq′′(w(f))≤M′′Np′′(f). From this, conclude that
w also maps
H(X;R) into
Lq(Y;R) and that
Nq(w(f))≤M⋅Np(f) with
M≤M′1−tM′′t (M. Riesz’s inequality). (Write
f in the form
∑nun as in b), and consider the series with general term
w(un); use the inequalities (
) and (*) of b).)