0% found this document useful (0 votes)
20 views83 pages

Thy Distr HC Notes 2025jan 1

The lecture notes by Hasse Carlsson provide a comprehensive overview of distributions, including their definitions, properties, and applications in analysis. Key topics include the extension of differential calculus and Fourier analysis to distributions, the concept of test functions, and the convergence and operations on distributions. The notes also cover fundamental solutions to differential equations and various applications in mathematical analysis.

Uploaded by

homologicalito
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
20 views83 pages

Thy Distr HC Notes 2025jan 1

The lecture notes by Hasse Carlsson provide a comprehensive overview of distributions, including their definitions, properties, and applications in analysis. Key topics include the extension of differential calculus and Fourier analysis to distributions, the concept of test functions, and the convergence and operations on distributions. The notes also cover fundamental solutions to differential equations and various applications in mathematical analysis.

Uploaded by

homologicalito
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 83

Lecture notes

on
Distributions

Hasse Carlsson

2024
Contents

Preface 5
Chapter 1. A primer on C0∞ -functions 7
1.1. Notation 7
1.2. Approximate identities 8
1.3. Regularization by convolution 9
Chapter 2. Definition of distributions 11
2.1. The support of a distribution 13
Chapter 3. Operations on distributions 17
3.1. The derivative of distributions 17
3.2. Multiplication by functions 18
Chapter 4. Finite parts 21
Chapter 5. Fundamental solutions of the Laplace and heat equations 29
Chapter 6. Distributions with compact support 33
Chapter 7. Convergence of distributions 35
Chapter 8. Convolution of distributions 39
Chapter 9. Fundamental solutions 45
Chapter 10. The Fourier transform 49
Chapter 11. The Fourier transform on L2 57
Chapter 12. The Fourier transform and convolutions 59
Chapter 13. The Paley-Wiener theorem 65
Chapter 14. Existence of fundamental solutions 67
Chapter 15. Fundamental solutions of elliptic differential operators 69
Chapter 16. Fourier series 71
Chapter 17. Some applications 75
3
4 CONTENTS

17.1. The central limit theorem 75


17.2. The mean value property for harmonic functions 76
17.3. The Heisenberg uncertainty principle 77
17.4. A primer on Sobolev inequalities 78
17.5. Minkowski’s theorem 81
Index 83
Preface

Two important methods in analysis is differentiation and Fourier transforma-


tion. Unfortunally not all functions are differentiable or has a Fourier transform.
The theory of distribution tries to remedy this by imbedding classical functions
in a larger class of objects, the so called distributions (or general functions). The
basic idea is not to think of functions as pointwise defined but rather as a ”mean
value”. A locally integrable function f is identified with the map
Z
φ 7→ f φ,

where φ belongs to a space of ”nice” test functions, for instance C0∞ .


As an extension of this we let a distribution be a linear functional on the space
of test functions. When extending operations such as differentiation and Fourier
transformation, we do this by transfering the operations to the test functions,
where they are well defined.
Let us for instance see how to define the derivative of a locally integrable
function f on R. If f is continuously differentiable, an integration by parts implies
that Z Z
f φ = − f φ′ .
Now we use this formula to define the differential of f , when f is not classically
differentiable. f ′ is the map Z
φ 7→ − f φ′ .
In these lectures we will study how differential calculus and Fourier analysis
can be extended to distributions and study some applications mainly in the theory
of partial differential equations.
The presentation is rather short and for a deeper study I recommend the fol-
lowing books:
Laurent Schwartz. Théorie des Distributions I, II. Hermann, Paris, 1950–51.
Lars Hörmander. The Analysis of Linear Partial Differential Operators I, 2nd
ed. Springer, Berlin, 1990.

5
CHAPTER 1

A primer on C0∞ -functions

When we shall extend differential calculus to distributions, it is suitable to use


infintely differentiable functions with compact support as test functions. In this
chapter we will show that there is ”a lot of” C0∞ -functions.

1.1. Notation
Let Ω be a domain in Rn . C k (Ω) denotes the k times continuously differentiable
functions on Ω. (k may be +∞.) C0k (Ω) are those functions in C k (Ω) with compact
support. We denote points in Rn with x = (x1 , . . . , xn ) and dx = dx1 . . . dxn
denotes the Lebesgue measure. For a vector α = (α1 , . . . αn ) ∈ Nn we let
|α| = α1 + . . . + αn , α! = α1 ! . . . αn !, xα = xα1 1 . . . xαnn
and
∂ αf ∂ α1 ∂ αn
∂ αf = = . . . f.
∂xα ∂xα1 1 ∂xαnn
Exampel 1.1. With these notations the Taylorpolynomial of f of degree N
can be written as
X ∂ α f (a)
xα .
α!
|α|≤N
2
As described in the preface, to a function f ∈ L1loc , we will associate the map
Λf , given by Z
φ 7→ f φ dx, φ ∈ C0∞ .
Rn

Problem 1.2. Does the map Λf determine f ?


More precisely, if f, g ∈ L1loc and
Z Z
f φ dx = gφ dx, φ ∈ C0∞ ,
Rn Rn
does this imply that f = g a.e.? 2

To be able to solve this problem we need to construct functions φ ∈ C0∞ . We


start with
7
8 1. A PRIMER ON C0∞ -FUNCTIONS

Exampel 1.3. There are functions f ∈ C ∞ (R) with f (x) = 0 when x ≤ 0 and
f (x) > 0 when x > 0.
Remark 1.4. There is no such real analytic function. 2
Proof. Such a function must satisfy f (n) (0) = 0 for all n. Thus f (x) =
n
0(x ), x → 0, for all n. Guided by this, we put

 e−1/x , x > 0
f (x) =
 0, x≤0.
We have to prove that f ∈ C ∞ . By induction we have

 P 1  e− x1 , x > 0
n x
f (n) (x) =
 0, x≤0
for some polynomials Pn . This is clear when x ̸= 0. But at the origin we have if
h > 0,
f (n) (h) − f (n) (0)
 
1 1 −1
= Pn e h → 0, h → 0.
h h h

Exampel 1.5. There are non-trivial functions in C0∞ (Rn ).
Proof. Let f be the function in Example 2 and put φ(x) = f (1 − |x|2 ). □

1.2. Approximate identities


Pick a function φ ∈ C0∞ (Rn ) with φ = R1 and φ ≥ 0. For δ > 0 we let
R

φδ (x) = δ −n φ(x/δ). Then φδ ∈ C0∞ (Rn ) and φδ = 1. {φδ ; δ > 0} is called an


approximate identity.

”φδ → δ”
1.3. REGULARIZATION BY CONVOLUTION 9

1.3. Regularization by convolution


The convolution of two functions f and φ is defined by
Z
f ∗ φ(x) = f (x − y)φ(y)dy.
Rn
The convolution is defined for instance if f ∈ L1loc and φ ∈ C0∞ . Then f ∗ φ =
φ ∗ f, f ∗ φ ∈ C ∞ and ∂ α (f ∗ φ) = f ∗ ∂ α φ.
Exercise 1.1. Verify this.

Theorem 1.6.
a) If f ∈ C0 , then f ∗ φδ → f, δ → 0, uniformly.
b) If f is continuous in x, then f ∗ φδ (x) → f (x), δ → 0.
c) If f ∈ Lp , 1 ≤ p < +∞, f ∗ φδ → f i Lp (and a.e.).
Remark 1.7. a) implies that C0∞ (Ω) is dense in C0 (Ω) (in the supremum
norm).
Exercise 1.2. Verify this.

Proof.
a) Take R so that supp φ ⊂ {x; |x| ≤ R}. We have
Z
|f ∗ φδ (x) − f (x)| ≤ |f (x − y) − f (x)|φδ (y)dy ≤ uniform continuity
|y|≤δR
Z
≤ϵ φδ (y)dy = ϵ, if δ is small enough.
Rn
b) This is left as
Exercise 1.3.

c) Jensen’s inequality implies


Z p
p
|f ∗ φδ (x) − f (x)| ≤ |f (x − y) − f (x)|φδ (y)dy
Rn
Z Z
p
≤ |f (x − y) − f (x)| φδ (y)dy = |f (x − δt) − f (x)|p φ(t)dt.
Rn Rn
δt
Using Fubini’s theorem and the notation f (x) = f (x − δt), we get
Z Z
p
∥f ∗ φδ − f ∥p ≤ φ(t)dt |f (x − δt) − f (x)|p dx
n Rn
ZR
= ∥f δt − f ∥pp φ(t)dt → 0,
Rn
That the limit is zero follows by dominated convergence and that translation is
continuous on Lp . This in turn follows since C0 is dense in Lp , 1 ≤ p < +∞:
10 1. A PRIMER ON C0∞ -FUNCTIONS

If g ∈ C0 , then
Z
δ
∥g − g∥pp = |g(x − δ) − g(x)|p dx → 0, δ → 0,
K
by dominated convergence. Now approximate f ∈ Lp with g ∈ C0 , ∥f − g∥p < ϵ.
Minkowski’s inequality (the triangle inequality) implies
∥f δ − f ∥p ≤ ∥f δ − g δ ∥p + ∥g δ − g∥p + ∥g − f ∥p ≤ 2ϵ + ∥g δ − g∥p ≤ 3ϵ,
if δ is small enough. □
Exercise 1.4. a) Let Br = {x; |x| < r}. Construct a function ψδ ∈ C0∞ (Rn )
such that 0 ≤ ψδ ≤
1, ψδ = 1 on Br and supp ψδ ⊂ Br+δ . How big must ∥δ α ψδ ∥∞ be?
b) Let K ⊂ Ω where K is compact and Ω is open in Rn . Construct ψ ∈ C0∞ (Ω) with ψ = 1
on a neighborhood of K and 0 ≤ ψ ≤ 1. How big must ∥∂ α ψ∥∞ be?
Now we are able to answer yes to the problem on page 7.
Theorem 1.8. A locally integrable function that is zero as a distribution is
zero a.e.

R
Proof.R We assume that f φ = 0 for all φ ∈ C0 . According to Theorem 1 a),
we have f Φ = 0 for all Φ ∈ C0 , and thus f = 0 a.e. (for instance by the Riesz
representation theorem.)
Alternatively we can argue as follows: Take ψn ∈ C0∞ with ψn (x) = 1 when
|x| ≤ n. Then f ψn ∈ L1 and
Z
f ψn ∗ φδ (x) = f (y)ψn (y)φδ (x − y)dy = 0,
Rn
since y 7→ ψn (y)φδ (x−y) is C0∞ .
But f ψn ∗φδ → f ψn in L1 according to Theorem 1
c). Hence f ψn = 0 a.e., and thus f = 0 a.e. □
CHAPTER 2

Definition of distributions

Definition 2.1. Let Ω be an open domain in Rn . A distribution u in Ω is


a linear functional on C0∞ (Ω), such that for every compact set K ⊂ Ω there are
constants C and k such that
X
|u(φ)| ≤ C ∥∂ α φ∥∞ , (2.1)
|α|≤k

for all φ ∈ C0∞ with supp φ ⊂ K. 2


We denote the distributions on Ω by D ′ (Ω). If the same k can be used for all
K, we say that u has order ≤ k. These distributions are denoted Dk′ (Ω). The
smallest k that can be used is called the order of the distribution. DF′ = ∪k Dk′ are
the distributions of finite order.
Exampel 2.2.
(a) A function f ∈ L1loc is a distribution of order 0.
(b) A measure is a distribution of order 0.
(c) u(φ) = ∂ α φ(x0 ) defines a distribution of order |α|.
(d) Let xj be a sequence without limit point in Ω and let
X
u(φ) = ∂ αj φ(xj ).
Then u is a distribution. u has finite order if and only if
sup |αj | < ∞ and then the order is sup |αj |. 2
We will use the notation D(Ω) to denote the set C0∞ (Ω), in particular when we
consider D(Ω) with a topology that corresponds to the the following convergence
of test functions.
Definition 2.3. φj → 0 in D(Ω) if, for all j, supp φj are contained in a fix
compact set and ∥∂ α φj ∥∞ → 0, j → ∞, for all α. 2
Theorem 2.4. A linear functional u on D(Ω) is a distribution if and only if
u(φj ) → 0 when φj → 0 in D(Ω).
Proof. ⇒): Trivial.
⇐): Assume that (1) doesn’t hold. We have to prove that u(φj ) ̸→ 0, although
φj → 0 in D(Ω). That (1) doesn’t hold implies that there is a compact set K and
11
12 2. DEFINITION OF DISTRIBUTIONS

a function φj ∈ D(Ω), with φj ⊂ K, u(φj ) = 1 and


X
|u(φj )| > j ∥∂ α φj ∥∞ .
|α|≤j

This implies ∥∂ α φj ∥∞ ≤ 1
j
if j ≥ |α|. Thus φj → 0 in D(Ω). □
Theorem 2.5. A distribution u ∈ Dk′ (Ω) can uniquely be extended to a linear
functional on C0k (Ω). For every compact set K ⊂ Ω there is a constant C = CK
such that X
|u(φ)| ≤ C ∥∂ α φ∥∞ , (2.2)
|α|≤k

for all φ ∈ C0k (Ω) with support in K.


Corollary 2.6. Measures and distributions of order 0 coincides.
Proof of Theorem 5. Let φ be a fix function in C0k (Ω). Let Φδ ∈ C0∞ be
an approximate identity and put φn = φ ∗ Φ 1 , n ≥ N . Then all φn are supported
n
in a fix compact set K in Ω and if |α| ≤ k then
∥∂ α (φ − φn )∥∞ = ∥∂ α φ − (∂ α φ) ∗ Φ 1 ∥∞ → 0, n → ∞ . (2.3)
n

Hence, if u has an extension satisfying (2), then u(φ) = limn→∞ u(φn ). This proves
the uniqueness of the extension and makes it natural to define
u(φ) = lim u(φn ) .
n→∞

The limit exists since u(φn ) is a Cauchy sequence:


X
|u(φn ) − u(φm )| = |u(φn − φm )| ≤ C ∥∂ α (φn − φm )∥ → 0,
|α|≤k
as n, m → ∞.
It is easy to see, by taking limits in (1), that u satisfies (2). □
Exercise 2.1. Verify this.
Theorem 2.7. A positive distribution is a positive measure.
Definition 2.8. A distribution u is positive if φ ≥ 0 implies u(φ) ≥ 0 2
Proof. By Corollary 6 it is enough to show that u ∈ D0′ .
Assume first that φ is real valued. Let K ⊂⊂ Ω and take χ ∈ C0∞ (Ω), 0 ≤ χ ≤
1 with χ = 1 on K. If supp φ ⊂ K, then χ∥φ∥∞ ±φ ≥ 0. Hence u(χ∥φ∥∞ ±φ) ≥ 0,
or
|u(φ)| ≤ u(χ∥φ∥∞ ) = u(χ)∥φ∥∞ .
So (1) holds with k = 0, C = u(χ).
If φ = f + ig is complex valued,we get
|u(φ)| ≤ |u(f )| + |u(g)| ≤ u(χ)(∥f ∥∞ + ∥g∥∞ ) ≤ 2u(χ)∥φ∥∞ .

2.1. THE SUPPORT OF A DISTRIBUTION 13

Theorem 2.9. A distribution is determined by its local behavior.


More precisely: Assume that Ω = ∪Ωi and that ui ∈ D ′ (Ωi ). Furthermore we
assume that ui = uj on Ωi ∩ Ωj , i.e. if φ ∈ C0∞ (Ωi ∩ Ωj ) then ui (φ) = uj (φ). Then
there is a unique distribution u on Ω with u = ui on Ωi .
To prove this we need a C0∞ partition of unity.
Proposition 2.10. Let K be a compact set with K ⊂ ∪N
1 Ωi . Then there are

φi ∈ C0 (Ωi ), 0 ≤ φi ≤ 1 and Σφi = 1 on K.
Proof of Theorem 9. Assume that u = ui on Ωi .PLet supp φ = K and φi
be a partition of unity as above. By linearity, since φ = i φφi ,
X X
u(φ) = u(φφi ) = ui (φφi ) (2.4)
i i

This shows the uniqueness.


To prove the existence, we need to show that (4) gives aP well defined distribution
u. But if φ̃k isPanother
P P partition ofPunity,
P then φ̃k = P i φi φ̃k on K and thus
u
k k (φ φ̃ k ) = k u
i k (φ φ̃ φ
k i ) = i k ui (φφ̃k φi ) = i ui (φφi ), so (4) defines
u uniquely.
It is easy to show that u satisfies (1), and the theorem is proved. □
Exercise 2.2. Do it!

Proof of Proposition 10. We shall show the following


Claim 2.11. There are open sets Vi with V i ⊂ Ωi and K ⊂ ∪N
1 Vi .

Assuming this take φ̃i ∈ C0∞ (Ωi ), 0 ≤ φ̃i ≤ 1 with φ̃i = 1 on V i . Then Σφ̃i > 0
on a neighborhood U of K. Take χ with χ = 1 on K and supp χ ⊂ U . Put
φ̃i
φi = χ .
Σφ̃i
It is clear that φi satisfy the conditions in the proposition.
To prove the claim, take to x ∈ K a neighborhood Vx with x ∈ Vx ⊂ V x ⊂ Ωj
for[some j. Then K ⊂ Vx . By compactness we get K ⊂ N
S S
1 Vxk . Let Vi =
Vxk . □
Vxk ⊂Ωi

2.1. The support of a distribution


R
If f ∈ C then supp f = {x; f (x) ̸= 0}. This implies that f φ = 0 for all
φ ∈ C0∞ whose support doesn’t intersect the support of f .
Definition 2.12. If u ∈ D ′ (Ω) then supp u = {x ∈ Ω; There is no neighbor-
hood of x with u = 0 in this neighborhood.}
Exercise 2.3. Show that supp u is closed.
14 2. DEFINITION OF DISTRIBUTIONS

Theorem 2.13. If supp u ∩ supp φ = ∅, then u(φ) = 0.


Proof. This follows directly from Theorem 9, since u = 0 locally on
Ω \ supp u. □
An important extension of Theorem 12 is the following theorem and its corol-
lary.
Theorem 2.14. Assume that u ∈ Dk′ (Ω) and φ ∈ C0k (Ω) with ∂ α φ(x) = 0 if
|α| ≤ k and x ∈ supp u. Then u(φ) = 0.
Corollary 2.15. If u ∈ D ′ (Ω) and supp u = {x0 } ⊂ Ω, then u is of the form
X
u(φ) = aα ∂ α φ(x0 ).
|α|≤k

Proof of Theorem 13. Let K = supp u ∩ supp φ. If K = ∅, the result


follow from Theorem 5. But K can be non-empty. Then, let Kϵ = {x; d(x, K) < ϵ}
and take χϵ ∈ C0∞ (Kϵ ) with χϵ = 1 in a neighborhood of K. Then, by Theorem 5,
u(φ) = u(χϵ φ + (1 − χϵ )φ) = u(χϵ φ).
If k = 0 this implies
|u(φ)| ≤ C∥χϵ φ∥∞ → 0, ϵ → 0.
If k > 0 we get
X X
|u(φ)| ≤ C ∥∂ α (χϵ φ)∥∞ ≤ C ∥∂ α χϵ ∂ β φ∥∞ .
|α|≤k |α|+|β|≤k

We can choose χϵ such that ∥∂ α χϵ ∥∞ ≤ Cϵ−|α| . To estimate ∥∂ β φ∥∞ we consider


the Taylor expansion of φ at a point x ∈ K. Let y ∈ Kϵ and take x ∈ K with
|x − y| ≤ ϵ. Put
g(t) = ∂ β φ(x + t(y − x)).
By the Taylorexpansion of g at t = 0 of order k − |β| − 1, we get
X g (i) (0)
|∂ β φ(y)| = |g(1)| = + R(y) .
i!
i≤k−|β|−1

Now g (i) (0) = 0 and


X
|R(y)| ≤ C sup |∂ k−|β| g(s)| ≤ Cϵk−|β| ∥∂ β φ∥Kϵ .
0≤s≤1
|β|=k

This implies
X X
|u(φ)| ≤ C ϵk−|α|−|β| ∥∂ β φ∥Kϵ → 0, ϵ → 0.
|α|+|β|≤k |β|=k


2.1. THE SUPPORT OF A DISTRIBUTION 15

Proof of the corollary. u is of finite order k for some k. Fix χ ∈ C0∞ (Ω)
with χ = 1 near x0 and put
X ∂ α φ(x0 )
ψ(x) = φ(x) − χ(x) (x − x0 )α .
α!
|α|≤k
α
Then ∂ ψ(x0 ) = 0 if |α| ≤ k. By Theorem 13, u(ψ) = 0 or
X  (x − x )α  X
α 0
u(φ) = ∂ φ(x0 )u χ(x) = aα ∂ α φ(x0 ).
α!
|α|≤k |α|≤k


Exercise 2.4. H 2.2
Exercise 2.5. H 3.1.7.
P∞
Exercise 2.6. Show that u(φ) = 1 nα (φ( n1 ) − φ(− n1 )) is a distribution of order ≤ 1 if α < 0.
Also show that supp u = {0, ±1, ± 12 , ± 13 , . . .}, but if K is a closed set with
k
X
|u(φ)| ≤ C sup |∂ i φ|, φ ∈ C0∞ (R) ,
i=0 K
then either α < −1 or else K contains a neighborhood of the origin. (In particular we can not
choose K = supp u.)
Exercise 2.7. Assume that u ∈ Dk′ (R) and supp u ⊂ I where I is a compact interval. Show
that X
|u(φ)| ≤ C sup |∂ α φ|, φ ∈ C0∞ (R).
I
|α|≤k
(Hint. Theorem 13.)
Exercise 2.8. Is there a linear functional u on C0∞ that isn’t a distribution?
CHAPTER 3

Operations on distributions

3.1. The derivative of distributions


If u is a continuously differentiable function in Rn , an integration by parts gives
Z Z
∂k u · φ dx = − u · ∂k φ dx, φ ∈ D,
Rn Rn
as φ has compact support. This motivates the following definition.
Definition 3.1. If u ∈ D ′ (Ω), we define ∂k u ∈ D ′ (Ω) by
∂k u(φ) = −u(∂k φ).
That ∂k u defines a distribution follows since
X X
|∂k u(φ)| = |u(∂k φ)| ≤ C ∥∂ α (∂k φ)∥∞ ≤ C ∥∂ α φ∥∞ .
|α|≤k |α|≤k+1

If u ∈ C 1 the distribution derivative coincides with the classsical derivative.


Exampel 3.2. Let the Heaviside function H be defined by

 1, x ≥ 0
H(x) =
 0, x < 0.

Then Z ∞
′ ′
H (φ) = −H(φ ) = − φ′ (x)dx = φ(0).
0
n
The Dirac measure δx0 at x0 ∈ R is given by δx0 (φ) = φ(x0 ). So we have
showed that H ′ = δ0 . The derivatives of the Dirac measure are given by ∂ α δx0 (φ) =
(−1)|α| δx0 (∂ α φ) = (−1)|α| ∂ α φ(x0 ). With this notation, by Corollary 2.14, a distri-
bution supported at x0 can be written as
X
u= Cα δx(α)
0
.
|α|≤k

A generalization of Exampel 2 is given by


Proposition 3.3. Let u be a function in Ω ⊂ R, which is continuously differ-
entiable for x ̸= x0 . Assume that the derivative v is integrable near x0 . Then
u′ = v + (u(x0 + 0) − u(x0 − 0))δx0 .
17
18 3. OPERATIONS ON DISTRIBUTIONS

Proof. We start by showing that the limits exist. Let x0 < x < y. Then
Z y
u(x) = u(y) − v(t)dt.
x

Since v is integrable we obtain as x ↓ x0


Z y
u(x0 + 0) = u(y) − v(t)dt.
x0

By the same argument also u(x0 − 0) exists. We get


Z Z
′ ′ ′
u (φ) = −u(φ ) = − uφ dx = lim − u(x)φ′ (x)dx
R ϵ→0 |x−x0 |>ϵ
n h i∞ h ix0 −ϵ Z o
= lim − u(x)φ(x) − u(x)φ(x) + v(x)φ(x)dx
ϵ→0 x0 +ϵ −∞ |x−x0 |>ϵ
Z
= (u(x0 + 0) − u(x0 − 0))φ(x0 ) + v(x)φ(x)dx.
R


Theorem 3.4. Let u be a distribution on an interval I ⊂ R. If u′ = 0, then u
is constant.
Proof. That u′ = 0 as a distribution means that u′ (φ) = 0 or u(φ′ ) = 0 for
all φ ∈ D. To compute u(ϕ), we want to decide if ϕR = ψ ′ for some ψ ∈ RD. This
R x
is the case exactly when ϕ = 0 and then ψ(x) = −∞ ϕ(t)dt. Thus, if ϕ = 0,
R case to this special case. Fix ψ0 ∈R D
then Ru(ϕ) = 0. We shall reduceRthe general
with ψ0 = 1. Put R ϕ̃ = ϕ − ψ0 ϕ. Then ϕ̃ = 0 so 0 = u(ϕ̃) = u(ϕ) − u(ψ0 ) ϕ
or u(ϕ) = u(ψ0 ) ϕ. Thus u is the constant u(ψ0 ). □

3.2. Multiplication by functions


D(Ω) is a linear space, since we can add distributions and multiply a distribu-
tion with a scalar in a natural way. We also want to multiply a distribution with
a function f . If u is a locally integrable function, then
Z Z
f u(φ) = (f u)φ dx = u(f φ) dx = u(f φ) .
Rn Rn

To be able to use this to define f u when u is a distribution we need that f φ ∈ C ∞ .


Definition 3.5. If f ∈ C ∞ we define f u by
f u(φ) = u(f φ).
Exercise 3.1. Show that f u ∈ D ′ (Ω).

Remark 3.6. If u is of order k, it is enough to demand that f ∈ C k . 2


3.2. MULTIPLICATION BY FUNCTIONS 19

Proposition 3.7.
(a) ∂j ∂k u = ∂k ∂j u
(b) ∂k (f u) = (∂k f )u + f ∂k u.
Exercise 3.2. Prove Proposition 3.

Remark 3.8. By (a), the distributional derivatives commutes and we can use
the notation ∂ α u, ∂ α u(φ) = (−1)|α| u(∂ α φ) where α is a multiindex. 2
Theorem 3.9. Assume that u ∈ D ′ (Ω), Ω ⊂ R, and u′ + au = f where f ∈ C
and a ∈ C ∞ . Then u ∈ C 1 and the equation holds classically.
Proof. Assume first that a ≡ 0. Let F be a (classsical) primitive function of
f . Then F ∈ C 1 and (u − F )′ = u′ − F ′ = f − f = 0 as a distribution. Theorem 1
implies that u = F + C, and thus u ∈ C 1 and u′ = F ′ = f classically.
If a ̸≡ 0, we multiply the equation with its integrating factor. Let A be a
primitive function of a. Then A and eA are C ∞ functions. Furthermore, we have
(eA u)′ = eA u′ + eA au = eA (u′ + au)
in the distributional sense. Therefore, the equation is equivalent to
(eA u)′ = eA f,
and we can use the case a ≡ 0. □
Exercise 3.3. H 3.1.1
Exercise 3.4. H 3.1.5
Exercise 3.5. H 3.1.14
Exercise 3.6. H 3.1.21
Exercise 3.7. H 3.1.22.
Exercise 3.8. Assume that u ∈ D ′ (Ω), Ω ⊂ R, satisfies u(m) + am−1 u(m−1) + . . . + a0 u = f ,
where f ∈ C and aj ∈ C ∞ . Show that u ∈ C m , and that the equation holds classically.
CHAPTER 4

Finite parts

In this chapter we will extend Proposition 3.3 to the case where the derivative
is not locally integrable.
 
d 1
Exampel 4.1. What is √ ?
dx x+
We have
1 1
⟨( √ )′ , φ⟩ = −⟨ √ , φ′ ⟩ =
x+ x+
Z ∞ i∞ 1 Z ∞ φ(x) 
1 ′  h 1
= lim − √ φ (x)dx = lim − √ φ(x) − dx
ϵ→0 ϵ x ϵ→0 x ϵ 2 ϵ x3/2
1 Z ∞ 1 2φ(0) 
= − lim φ(x)dx − √
2 ϵ→0 ϵ x3/2 ϵ
2
Definition 4.2.
Z ∞ 
1 φ(x) 2φ(0)
⟨fp 3/2
, φ⟩ = lim 3/2
dx − √ .
x+ ϵ→0 ϵ x ϵ
2
Thus we have shown that
d 1 1 1
( √ ) = − fp 3/2 .
dx x+ 2 x+

A version of the definition that is easier to remember is


Z ∞
1 φ(x) − φ(0)
⟨fp 3/2 , φ⟩ = dx.
x+ 0 x3/2

1
Exampel 4.3. We define fp by
|x|5/2
φ(x) − φ(0) − xφ′ (0)
Z
1
⟨fp 5/2 , φ⟩ = dx .
|x| R |x|5/2
21
22 4. FINITE PARTS

1
The order of fp is 2. To show this we split the integral into two pieces,
|x|5/2
φ(x) − φ(0) − xφ′ (0)
Z Z
1
⟨fp 5/2 , φ⟩ = + dx = I+II.
|x| |x|≤1 |x|>1 |x|5/2
To estimate the first integral we use that
1
|φ(x) − φ(0) − xφ′ (0)| ≤ x2 ∥φ′′ ∥∞ .
2
This implies Z
1 ′′ 1
|I| ≤ ∥φ ∥∞ 1/2
dx ≤ C∥φ′′ ∥∞ .
2 |x|≤1 |x|
For the second integral we have
2∥φ∥∞ + |x|∥φ′ ∥∞
Z
|II| ≤ 5/2
dx ≤ C(∥φ∥∞ + ∥φ′ ∥∞ ) .
|x|>1 |x|
Thus the order is at most two.
To show that the order can not be smaller, we let φ ∈ C0∞ , 0 ≤ φ ≤ 1,
supp φ ⊂ (0, 3) and φ = 1 on [1, 2] and put φϵ (x) = φ(x/ϵ). Then
Z Z 2ϵ
1 φϵ (x) 1 1
|⟨fp 5/2 , φϵ ⟩| = 5/2
dx ≥ 5/2
dx ≥ c 3/2 .
|x| R x ϵ x ϵ
Furthermore ∥φϵ ∥∞ + ∥φ′ϵ ∥∞ ≤ C/ϵ. Thus if the order were less than 2, we would
1
have c/ϵ3/2 ≤ |⟨fp 5/2 , φ⟩| ≤ C/ϵ, a contradiction. 2
|x|
1 1
Since the order of fp 5/2 is 2 and |x|5/2 ∈ C 2 , |x|5/2 fp 5/2 is well defined
|x| |x|
and
1 1
⟨|x|5/2 fp 5/2 , φ⟩ = ⟨fp 5/2 , |x|5/2 φ⟩
|x| |x|
5/2
|x| φ(x)
Z Z
= dx = φ(x)dx = ⟨1, φ⟩.
R |x|5/2 R

Here we have used that |x|5/2 φ(x) and its derivative vanishes at x = 0.
Thus fp |x|15/2 solves the division problem |x|5/2 u = 1. 2
1 3 1
Exercise 4.1. Show that (fp 3/2
)′ = − fp 5/2 .
x+ 2 x+

The above examples can be generalized to to define fp x−a


+ , fp |x|
−a
and (for
−a
certain a) fp x etc. when a is not an integer, for instance
Z ∞
1 φ(x) − P (x)
⟨fp a , φ⟩ = dx,
|x| −∞ |x|a
4. FINITE PARTS 23

where P is the Taylorpolynomial of φ at the origin of order [a] - 1. Then


 
1 ′ sgn x 1 1
(fp a ) = −a fp a+1 = a fp a+1 − fp a+1
|x| |x| x+ x−
and
1
|x|a fp
= 1.
|x|a
1
Another important property of fp a , a ̸= −1, −2, . . . , is that it is homoge-
|x|
neous of degree −a. As we shall see later this fact simplifies the computation of
its Fourier transform.
1
To show that fp a is homogeneous we first must define what it means. If u(x)
|x|
is a function on Rn , then u is homogeneous of degree α if u(tx) = tα u(x), t > 0.
This can be reformulated in a way that is meaningful for distributions. For a
function u, we have
Z Z
1 y
⟨u(tx), φ⟩ = u(tx)φ(x)dx = [y = tx] = u(y) n φ( )dy = ⟨u, φt ⟩.
R R t t
But if u is homogeneous of degree α, we also have
Z Z
⟨u(tx), φ⟩ = u(tx)φ(x)dx = tα u(x)φ(x)dx = tα ⟨u, φ⟩.
R R

Therefore we make the following definition.


Definition 4.4. u ∈ D ′ (Rn ) is homogeneous of degree α if
⟨u, φt ⟩ = tα ⟨u, φ⟩, t > 0.
2
1 1
Proposition 4.5. fp a
och fp a are homogeneous of degree −a if a ̸=
|x| x+
1, 2, 3, . . .
5 1 1
Proof when a = . We have φt (0) = φ(0) and φ′t (0) = 2 φ′ (0). Thus
2 t t
Z
1 1  1 x 1 x  h xi
⟨fp 5/2 , φt ⟩ = φ( ) − φ(0) − φ(0) dx = y =
|x| |x|5/2 t t t t2 t
ZR
1 ′ 1 1
= 5/2 |x|5/2
(φ(x) − φ(0) − xφ (0))dx = ⟨fp , φ⟩.
R t t5/2 |x|5/2

Exampel 4.6. Compute (log |x|)′ .
24 4. FINITE PARTS

We have Z
⟨(log |x|) , φ)⟩ = −⟨log |x|, φ ⟩ = − φ′ (x) log |x|dx
′ ′
R
Z nh i−ϵ

= − lim φ (x) log |x|dx = lim − φ(x) log |x|
ϵ→0 |x|>ϵ ϵ→0 −∞
i∞ Z
h dx o
+ φ(x) log |x| − φ(x)
ϵ |x|>ϵ x
nZ dx o
= lim φ(x) + (φ(ϵ) − φ(−ϵ)) log ϵ
ϵ→0 |x|>ϵ x
Z Z
dx φ(x)
= lim φ(x) = pv dx ,
ϵ→0 |x|>ϵ x R x
where the last equality is a definition. 2

Z
1 φ(x)
Definition 4.7. ⟨pv , φ⟩ = lim dx.
x ϵ→0 |x|>ϵ x
If we instead differentiate log x+ , we get
Z ∞
⟨log x+ , φ⟩ = lim − φ′ (x) log xdx =
ϵ→0 ϵ
 h i∞ Z ∞ φ(x) 
= lim − φ(x) log x + dx =
ϵ→0 ϵ ϵ x
Z ∞  Z ∞ Z 1 
φ(x) φ(x) φ(0)
= lim dx + φ(0) log ϵ = lim dx − dx =
ϵ→0 ϵ x ϵ→0 0 x ϵ x
Z ∞
φ(x) − χ(x)φ(0)
= dx,
0 x
where χ = χ[−1,1] , as in the rest of this chapter.
Thus with the following
Z ∞
1 φ(x) − χ(x)φ(0)
Definition 4.8. ⟨fp , φ⟩ = dx 2
x+ 0 x
1
we have proved that (log x+ )′ = fp .
x+
1
Exercise 4.2. Show that fp solves the division problem xu = H.
x+
The above examples can be generalized to the following
Definition 4.9.
xn−1
1
Z ∞ φ(x) − P (x) − (n−1)!
φ(n−1) (0)χ(x)
⟨fp n , φ⟩ = dx
|x| −∞ |x|n
4. FINITE PARTS 25

if n = 1, 2, 3, . . . and P is the Taylorpolynomial of φ of degree n − 2. 2

1 ′ 1 1 (2)
Exampel 4.10. (fp ) = −2fp + δ . 2
x2+ x3+ 2

Proof.
Z ∞ ′
1 ′ 1 ′ φ (x) − φ′ (0) − xφ′′ (0)χ(x)
⟨(fp 2 ) , φ⟩ = −⟨fp 2 , φ ⟩ = − dx =
x+ x+ 0 x2
n Z ∞ φ′ (x) − φ′ (0) Z 1
xφ′′ (0) o
− lim dx − dx .
ϵ→0 ϵ x2 ϵ x2

As φ(x) − φ(0) − xφ′ (0) is a primitive function of φ′ (x) − φ′ (0), an integration by


parts in the first integral implies that

1 ′ nh φ(x) − φ(0) − xφ′ (0) i∞


⟨(fp ) , φ⟩ = − lim +
x2+ ϵ→0 x2 ϵ
Z ∞ Z 1
φ(x) − φ(0) − xφ′ (0) xφ′′ (0) o
2 dx − dx .
ϵ x3 ϵ x2

Now
h φ(x) − φ(0) − xφ′ (0) i∞ 1
lim = − φ′′ (0)
ϵ→0 x2 ϵ 2
and
n Z ∞ φ(x) − φ(0) − xφ′ (0) Z 1
xφ′′ (0) o
lim 2 dx − dx =
ϵ→0 ϵ x3 ϵ x2
Z ∞
φ(x) − φ(0) − xφ′ (0) − 21 x2 φ′′ (0)χ(x)
2 dx .
0 x3

Hence
Z ∞
1 ′ 1 ′′ φ(x) − φ(0) − xφ′ (0) − 12 x2 φ′′ (0)χ(x)
⟨(fp 2 ) , φ⟩ = φ (0) − 2 dx
x+ 2 0 x3
1 1
= ⟨−2fp 3 + δ (2) , φ⟩.
x+ 2


26 4. FINITE PARTS

1
Exampel 4.11. fp is not homogeneous of degree −3 since
|x|3
1 x ′ 1 x2 ′′
φ( xt ) − 1t φ(0) − φ (0) − φ (0)χ(x)
Z
1 t t2 2 t3
⟨fp 3 , φt ⟩ = dx
|x| R |x|3
φ(x) − φ(0) − xφ′ (0) − 21 x2 φ′′ (0)χ(xt)
Z
h xi 1
= y= = 3 dx
t t R |x|3
Z
1 1 1 ′′ dx
= (If we assume that t > 1) = 3 ⟨fp 3 , φ⟩ + 3 φ (0)
t |x| 2t 1
t
<|x|<1 |x|
1 1 log t
= 3
⟨fp 3 , φ⟩ + φ′′ (0) 3 .
t |x| t
2
Exercise 4.3. What happens if t < 1?
Exercise 4.4. Is fp x13 homogeneous of degree −3?
N
X −1
Exercise 4.5. Show that the equation xN u = 0 has the solution u = cn δ (n) .
n=0

Since xN fp x1N = 1, Exercise 4.5 implies that the equation


N −1
N 1 X
x u = 1 has the general solution u = fp N + cn δ (n) .
x n=0

In the same way the equation


N −1
1 X
(x − a)N u = 1 has the solution u = fp + cn δa(n) .
(x − a)N n=0

1 1
where fp N
is defined in the same way as fp N .
(x − a) x
Now we can solve the division problem P u = 1, where P is a polynomial of
one variable. In a neighborhood where P ̸= 0, u = 1/P is a nice function. So the
the only problem is to understand 1/P near a real zero a of P . But there we have
P (x) = (x−a)n Q(x) where Q(a) ̸= 0. Hence, near x = a, we have (x−a)n Q(x)u =
1 1 1
1. This is satisfied if Qu = fp n
. Hence u = fp solves P u = 1
(x − a) Q(x) (x − a)n
near x = a. By Theorem 2.9, u is a well defined distribution on R that solves
P u = 1.
Exercise 4.6. H 3.1.14
Exercise 4.7. H 3.1.20
4. FINITE PARTS 27

Exercise 4.8. Let u be a continuous function on Rn \ {0} that is homogeneous of degree −n.
Show that we can define a distribution pv u by
Z
⟨pv u, φ⟩ = lim u(x)φ(x)dx,
ϵ→0 |x|>ϵ
R
if and only if |x|=1
u(x)dσ(x) = 0.

+ is by analytic continuation. If φ ∈ D and


Exercise 4.9. An alternative method to define fp xα
Re α > −1 the map Z ∞
Fφ (α) = xα φ(x)dx
0
is analytic. Show that this map can be continued to a meromophic function in C, whose only
singularities are simple poles in −1, −2, −3, . . .. Compute the residues R−k of Fφ and show that
if we for k = 1, 2, 3, . . . extend the definition of Fφ by
 
R−k
Fφ (−k) = lim Fφ (α) − ,
α→−k α+k
we have
Fφ (α) = ⟨fp xα
+ , φ⟩.
This approach gives an alternative proof that
−α

+ fp x+ = H, α∈C
and
′ α−1
(fp xα
+ ) = αfp x+ , α ̸= 0, −1, −2, −3, . . .
CHAPTER 5

Fundamental solutions of the Laplace and heat equations

Definition 5.1. Let P (D) be a differential operator. A distribution E with


P (D)E = δ, is called a fundamental solution of P .
2
In Example 3.2, we saw that the Heaviside function H is a fundamental solu-
tion of d/dx. A little more general, H(x1 ) · · · H(xn ) is a fundamental solution of
∂1 . . . ∂n .
In this chapter we will treat the Laplace operator
n
X ∂ 2u
∆u = 2
,
i=1
∂x i

and the heat operator


n
∂u X ∂ 2 u
 

− ∆x u = − .
∂t ∂t i=1
∂x2i
To accomplish this we need to be able to integrate by parts in Rn , and we remind
the reader about
Green’s identity.
Z Z  
∂v ∂u
(u∆v − v∆u) dx = u −v dσ
Ω ∂Ω ∂n ∂n
where ∂/∂n is the exterior normal derivative.
Theorem 5.2.
1


 log |x|, n = 2,
E(x) = 2π
1
 −
 , n ≥ 3,
ωn (n − 2)|x|n−2
is a fundamental solution of the Laplace operator in Rn .
(ωn is the surface measure of the unit sphere in Rn .)
Exercise 5.1. Compute ωn in terms of the Γ function,
Z ∞
Γ(s) = ts−1 e−t dt, Res > −1.
0

Exercise 5.2. Visa att ∆E(x) = 0 om x ̸= 0.


29
30 5. FUNDAMENTAL SOLUTIONS OF THE LAPLACE AND HEAT EQUATIONS

Proof.
Z
⟨∆E, φ⟩ = ⟨E, ∆φ⟩ = lim E∆φdx
ϵ→0 |x|>ϵ
Z
= Exercise 2 = lim (E∆φ − φ∆E)dx = Green’s identity =
ϵ→0 |x|>ϵ
Z  
∂φ ∂E
= lim E −φ dσ = lim(Iϵ + IIϵ ).
ϵ→0 |x|=ϵ ∂n ∂n ϵ→0

We only consider the case n ≥ 3, and leave the case n = 2 as


Exercise 5.3.

We have
∂φ 1
|Iϵ | ≤ C ωn ϵn−1 −→ 0, ϵ → 0 ,
∂n ∞ ϵn−2
and as ∂/∂n = −∂/∂r,
−(n − 2)
Z Z
∂E 1
IIϵ = φ dσ = − φ(x) dσ(x)
|x|=ϵ ∂r (n − 2)ωn |x|=ϵ |x|n−1
Z Z
φ(0) dσ(x) 1 dσ(x)
= + (φ(x) − φ(0))
ωn |x|=ϵ ϵn−1 ωn |x|=ϵ ϵn−1
−→ φ(0), ϵ → 0.

Theorem 5.3.

1 |x|2
exp(− ), t > 0,


E(x, t) = (4πt)n/2 4t
 0,

t < 0,

is a fundamental solution of the heat equation in Rn+1 .



Exercise 5.4. Show that ( ∂t − ∆x )E(t, x) = 0 if t ̸= 0.

Proof. Let ϕ(x) = E(x, 12 ). When n = 1, this is the density of a N (0, 1)


distributed stochastic variable and, when nR> 1 the product of n such densities.
Furthermore, E(x, t) = ϕ√2t (x) and thus Rn E(x, t)dx = 1 for all t > 0 and
E ∈ L1loc (Rn+1 ). Now
    Z Z
∂E ∂φ ∂φ
, φ = − E, = lim − dx E dt
∂t ∂t ϵ→0 Rn t>ϵ ∂t
Z Z Z 
∂E
= lim E(x, ϵ)φ(x, ϵ)dx + φ dxdt ,
ϵ→0 Rn Rn t>ϵ ∂t
5. FUNDAMENTAL SOLUTIONS OF THE LAPLACE AND HEAT EQUATIONS 31

and Z Z Z Z
⟨∆x E, φ⟩ = ⟨E, ∆x φ⟩ = lim E∆x φdxdt = lim ∆x Eφdxdt .
ϵ→0 Rn t>ϵ ϵ→0 Rn t>ϵ
Thus
  

− ∆x E, φ
∂t
nZ Z Z  
∂E o
= lim E(x, ϵ)φ(x, ϵ)dx + φ − ∆x E dxdt =
ϵ→0 Rn Rn t>ϵ ∂t
Z
= Exercise 4 = lim E(x, ϵ)φ(x, ϵ)dx.
ϵ→0 Rn
Since E(x, t) = ϕ√2t (x) is an approximate identity, we ought to have
Z
Iϵ = E(x, ϵ)φ(x, ϵ)dx → φ(0), ϵ → 0.
Rn
This does not follow directly from Theorem 1.3.1 since the support of√E(x, ϵ) is
not compact and φ(x, ϵ) depend on ϵ. But the change of variables x = 2ϵ y gives
Z √
Iϵ = ϕ(x)φ( 2ϵ x, ϵ)dx .
Rn
1

Since ϕ ∈ L and |φ( 2ϵ x, ϵ)| ≤ ∥φ∥∞ , we get by dominated convergence
Z √ Z
lim Iϵ = ϕ(x) lim φ( 2ϵ x, ϵ)dx = ϕ(x)φ(0, 0)dx = φ(0) .
ϵ→0 Rn ϵ→0 Rn

1 ∂
Exercise 5.5. Show that is a fundamental solution to in C.
πz ∂z

Exercise 5.6. Compute log |z| and ∆ log |z| in C.
∂z
Exercise 5.7. H 3.3.9
Exercise 5.8. H 3.3.11
Exercise 5.9. H 3.3.12
CHAPTER 6

Distributions with compact support

Theorem 6.1. Assume that u ∈ D ′ (Ω) has compact support. Then u has
a unique extension to C ∞ (Ω) that satisfies u(φ) = 0 if supp u and supp φ are
disjoint.
If K is a compact set that contains a neighborhood of supp u, then there is a k
so that X
|u(φ)| ≤ C ∥∂ α φ∥K , φ ∈ C ∞ (Ω). (6.1)
|α|≤k

Proof. Take χ ∈ C0∞ (K) with χ = 1 in a neighborhood of supp u. If φ ∈ C0∞ ,


then according to Theorem 2.12
u(φ) = u(χφ + (1 − χ)φ) = u(χφ) + u((1 − χ)φ) = u(χφ).
Thus
u(φ) = u(χφ), φ ∈ C ∞
defines an extension of u. (1) follows by Leibnitz’ rule.
Assume on the other hand that u1 is an extension to C ∞ . The condition on the
support implies that u1 ((1 − χ)φ) = 0, and consequently u1 (φ) = u1 (χφ) = u(χφ)
and thus the extension is unique.

Remark 6.2. Exercise 2.6 shows that it is not always possible to take K =
supp u in (1). 2
Exercise 6.1. State and prove a converse of Theorem 1.
Thus by Theorem 1 we can identify distributions with compact support with
the linear functionals on C ∞ (Ω) that satisfies (1). These distributions are denoted
E ′ (Ω).

33
CHAPTER 7

Convergence of distributions

Definition 7.1. A sequence uj ∈ D ′ (Ω) converges to u ∈ D ′ (Ω) if


uj (φ) → u(φ), j → ∞,
for every test function φ ∈ D(Ω). We denote this by uj → u in D ′ . 2
If uj →Pu in D ′ , we also have ∂ α uj → ∂ α u in D ′ for every multiindex α. We
write u = uj in D ′ if the partialP
sums converges in D ′ . If the series converges,
it is differentiable and we have ∂ α ( uj ) = ∂ α uj .
P

Remark 7.2. Convergence in D ′ is a ”weak” condition, if for instance fj → f


in Lp then fj → f i D ′ . 2
Exercise 7.1. Prove that.

Definition 7.3. uj ∈ D ′ (Ω) is a Cauchy sequence in D ′ (Ω) if uj (φ) is a


Cauchy sequence in C for every φ ∈ D(Ω).
Theorem 7.4. D ′ (Ω) is complete.
Since uj (φ) is a Cauchy sequence in C, the following limit exist
u(φ) = lim uj (φ),
j→∞

and defines a linear functional on D(Ω). The difficulty is to show that u is a distri-
bution, i.e. that u satisfies the norm inequality (2.1), or the equivalent formulation
in Theorem 2.4. This is a consequence of the Banach-Steinhaus theorem. The
Banach-Steinhaus theorem is a standard result in functional analysis. But for the
readers benefit(?), we prove it here.
Let K be a compact set in Ω. We shall study the space X = XK =
{φ ∈ C ∞ (Ω); supp φ ⊂ K}. We introduce a metric on X by
X ∥φ1 − φ2 ∥k
d(φ1 , φ2 ) = 2−k ,
k
1 + ∥φ 1 − φ2 ∥k

α
P
where ∥φ∥k = |α|≤k supK |∂ φ| and put ∥φ∥ = d(φ, 0).
Observe that if ϵ > 0, and we take N = Nϵ so that ∞ −k
< 2ϵ , then
P
N +1 2
N N
X ϵ X −k ϵ ϵ
∥φ∥ ≤ 2−k ∥φ∥k + ≤ 2 ∥φ∥N + ≤ ∥φ∥N + < ϵ ,
k=1
2 k=1
2 2
35
36 7. CONVERGENCE OF DISTRIBUTIONS

if ∥φ∥N < 2ϵ .

Exercise 7.2. Show that d is a metric on X.

Exercise 7.3. Show that X is complete.

Exercise 7.4. Show that φj → 0 in D(K) if and only if ∥φj ∥ → 0.

Exercise 7.5. Show that if ∥φj ∥ → 0 in D(K), there are positive numbers cj with cj → ∞ but
∥cj φj ∥ → 0.

The Banach-Steinhaus theorem. Let Λα be a family of linear functionals on


X with |Λα φ| ≤ Cα ∥φ∥. Then, either
1) there are r > 0 and C < ∞ with
sup |Λα φ| ≤ C
α

for all φ ∈ X with ∥φ∥ ≤ r ,

or

2) supα |Λα φ| = ∞ for some φ ∈ X.

Now we can complete the proof of Theorem 4. Take φ with support in K.


Since uj (φ) converges, 2) can not hold. Thus 1) holds, ie.
|u(φ)| ≤ sup |uj (φ)| ≤ C if ∥φ∥ ≤ r .
j

Hence if φk → 0 in D(K), k → ∞, Exercise 7.5 implies that there are ck → ∞ and


|u(ck φk )| ≤ C if k is large enough. Thus |u(φk )| ≤ cCk → 0, k → ∞, ie. u ∈ D ′ .
2
The Banach-Steinhaus theorem is a consequence of

Baire’s theorem. Assume that X is a complete metric space. Let V1 , V2 , . . . be


open dense sets in X. Then ∩i Vi is non-empty.

Proof. Let Br (ϕ) = {φ ∈ X; d(φ, ϕ) < r}. Since Vi are open and dense
we can successivly choose ϕi and ri with ri < 1i such that Br1 (ϕ1 ) ⊂ V1 and
Bri (ϕi ) ⊂ Vi ∩ Bri−1 (ϕi−1 ), i = 1, 2, 3, . . ..
If i, j ≥ n, then ϕi , ϕj ∈ Brn (ϕn ), and therefore d(ϕ1 , ϕj ) < n2 . Thus ϕn is a
Cauchy sequence, and ϕn → ϕ0 for some ϕ0 ∈ X. But ϕi ∈ Brn (ϕn ) if i ≥ n.
Hence ϕ0 ∈ Brn (ϕn ) ⊂ Vn for all n and ϕ0 ∈ ∩Vi . □
7. CONVERGENCE OF DISTRIBUTIONS 37

Proof of the Banach-Steinhaus theorem. Let ϕ(φ) = sup |Λα φ|. ϕ is


lower semi-continuous, and hence Vn = {φ; ϕ(φ) > n} is open. If some VN isn’t
dense, then there are φ0 , r with Br (φ0 ) ⊂ VNc ie.
{φ; ∥φ − φ0 ∥ < r} ⊂ VNc .
Thus if ∥φ∥ < r, then |Λα (φ0 + φ)| ≤ N . This implies
|Λα φ| ≤ |Λα (φ + φ0 )| + |Λα φ0 | ≤ 2N = C if ∥φ∥ < r.

On the other hand if all Vn are dense, then there are φ ∈ ∩Vn , ie. ϕ(φ) = ∞
or supα |Λα φ| = ∞. □

Theorem 7.5. Assume that uj → u0 in D ′ (Ω) and that uj ≥ 0. Then uj


converges weakly to a positive measure u0 .
Proof. Since u0 is the limit of positive distributions, u0 is a positve distribu-
tion. By Theorem 2.7, u0 is a positive measure. If χ ∈ C0∞ is equal to 1 on K, the
proof of Theorem 2.7 gave the estimate
|uj (φ)| ≤ 2uj (χ)∥φ∥∞ ,
when φ ∈ C0∞ is supported in K.
Since uj (χ) → u0 (χ), we have supj |uj (χ)| ≤ C, and we obtain
|uj (φ)| ≤ C∥φ∥∞ , φ ∈ C0∞ , j = 0, 1, 2, . . .
By taking limits, compare Theorem 2.5, this also holds when φ ∈ C0 .
Now let φ ∈ C0 . We have to prove that uj (φ) → u0 (φ), j → ∞. Take φn ∈ C0∞
whit φn → φ uniformly. Then
|uj (φ) − u0 (φ)| ≤ |uj (φ) − uj (φn )| + |uj (φn ) − u0 (φn )| + |u0 (φn ) − u0 (φ)|
= |uj (φ − φn )| + |uj (φn ) − u0 (φn )| + |u0 (φn − φ)|
≤ 2C∥φ − φn ∥ + |uj (φn ) − u0 (φn )| .
Hence
lim |uj (φ) − u0 (φ)| ≤ 2C∥φ − φn ∥∞ < ϵ
j→∞
if n is large enough.

Exercise 7.6. Assume that f is analytic in Ω = I × (0, δ) ⊂ C, where I is an open interval.
Show that if |f (z)| ≤ C|Imz|−N , then f (x + i0) = limy→0 f (x + iy) exists in the distribution
sense and f (x + i0) ∈ DN

+1 (I).

Exercise 7.7. Compute


1 1
a) x+i0 + x−i0
and
1 1
b) x+i0 − x+i0 .
Exercise 7.8. H 2.5
38 7. CONVERGENCE OF DISTRIBUTIONS

Exercise 7.9. H 2.6


Exercise 7.10. H 2.7
Exercise 7.11. H 2.9
Exercise 7.12. H 2.16
CHAPTER 8

Convolution of distributions

If u ∈ L1loc and φ ∈ C0∞ , then u ∗ φ(x) =


R
u(y)φ(x − y)dy. This motivates the
following
Definition 8.1. If u ∈ D ′ (Rn ) and φ ∈ D(Rn ), then
u ∗ φ(x) = ⟨uy , φ(x − y)⟩.
2
The notation ⟨uy , φ(x − y)⟩ means that the distribution u acts on the test
function y 7→ φ(x − y). Sometimes we also write ⟨u, φ(x − ·)⟩.
Remark 8.2. This definition can also be used in the case where u ∈ E ′ (Rn ),
φ ∈ C ∞ (Rn ). 2
Theorem 8.1. If u ∈ D ′ (Rn ) and φ ∈ D(Rn ), then
a) u ∗ φ ∈ C ∞ (Rn )
b) supp(u ∗ φ) ⊂ supp u + supp φ
c) ∂ α (u ∗ φ) = u ∗ ∂ α φ = (∂ α u) ∗ φ
Proof. We first show that u ∗ φ is continuous. Let x → x0 . If |x − x0 | ≤ 1,
then y 7→ φ(x − y) has support in a fixed compact set. Furthermore
∂yα (φ(x − y) − φ(x0 − y)) → 0, x → x0 , uniformly. Hence φ(x − y) → φ(x0 − y),
in D when x → x0 , and we get u ∗ φ(x) = ⟨uy , φ(x − y)⟩ → ⟨uy , φ(x0 − y)⟩ =
u ∗ φ(x0 ), x → x0 .
Since u ∗ φ is continuous, to prove b) it is enough to show that if x ∈
/ supp u +
supp φ, then u ∗ φ(x) = 0. But if x ∈ / supp u + supp φ, there are no y ∈ supp u
with x − y ∈ supp φ. So there is no y with y ∈ supp u and y ∈ supp φ(x − ·).
Hence supp u ∩ supp φ(x − ·) = ∅ och u ∗ φ(x) = 0.
The proof of the second equality in c) is simple.
∂ α u ∗ φ(x) = ⟨∂ α uy , φ(x − y)⟩ = (−1)|α| ⟨uy , ∂yα φ(x − y)⟩ =
= ⟨uy , φ(α) (x − y)⟩ = u ∗ (∂ α φ)(x).
The first equality follows by induction if we can prove it in the special case
α = (1, 0, . . . , 0). Thus it is enough to show that
1
lim (u ∗ φ(x + he1 ) − u ∗ φ(x)) = u ∗ ∂1 φ(x).
h→0 h
39
40 8. CONVOLUTION OF DISTRIBUTIONS

Let ϕx,h (y) = h1 (φ(x+he1 −y)−φ(x−y)). Then h1 (u∗φ(x+he1 )−u∗φ(x)) = u(ϕx,h ).


But ϕx,h (y) → ∂x ∂φ
1
(x − y) i D. Hence ∂ α (u ∗ φ)(x) = limh→0 u(ϕx,h ) = uy ( ∂x
∂φ
1
(x −
y)) = u ∗ ∂1 φ(x).
Since a) follows from c) the theorem is proved. □
Exercise 8.1. Show that ϕx,h (y) → ∂φ
∂x1 (x − y) i D
Exercise 8.2. Show that the convolution of functions is associative.

Theorem 8.2. If u ∈ D ′ (Rn ) and φ, ψ ∈ D(Rn ), then (u ∗ φ) ∗ ψ = u ∗ (φ ∗ ψ).


Remark 8.3. If u ∈ E ′ (Rn ), it is enough that one of φ, ψ has compact
support. 2
Proof. We have
Z
?
u ∗ (φ ∗ ψ)(x) = ⟨uy , φ ∗ ψ(x − y)⟩ = ⟨uy , φ(x − y − t)ψ(t)dt⟩ =
Rn
Z Z
?
= ⟨uy , φ(x − y − t)⟩ψ(t)dt = u ∗ φ(x − t)ψ(t)dt
Rn Rn
= (u ∗ φ) ∗ ψ(x).
?
To prove that = holds, we approximate the integral with its Riemann sum. By
?
Lemma 4 below the Riemann sum converges to the convolution in D and thus =
holds. □
Lemma 8.4. If φ ∈ C0j (Rn ) and ψ ∈ C0 (Rn ), then
X
φ(x − kh)ψ(kh)hn −→ φ ∗ ψ(x) i C0j ,
k∈Zn

as h → 0.
Proof of Lemma 4. The sum is supported in supp φ + supp ψ. The function
(x, y) 7→ φ(x − y)ψ(y) is uniformly continuous. Hence the Riemann sum converges
uniformly to φ ∗ ψ(x). Since ∂ α (φ ∗ ψ) = ∂ α φ ∗ ψ om |α| ≤ j, this also holds for
the derivatives. □
Theorem 8.5 (Regularisation of distributions.). Let u ∈ D ′ (Rn ) and φδ be an
approximate identity. Then u ∗ φδ → u in D ′ (Rn ), δ → 0 .
∨ ∨ ∨

Proof. Define ψ by ψ (x) = ψ(−x). Then u(ψ) = u∗ ψ (0). By Theorem 2,


this implies

uδ (ψ) = u ∗ φδ (ψ) = (u ∗ φδ )∗ ψ (0) =


= u ∗ (φδ ∗ ψ (0) .
8. CONVOLUTION OF DISTRIBUTIONS 41
∨ ∨

But, since φδ is an approximate identity, φδ ∗ ψ→ψ in D(Rn ), δ → 0. Hence


∨ ∨

lim uδ (ψ) = lim u ∗ (φδ ∗ ψ)(0) = u∗ ψ (0) = u(ψ).


δ→0 δ→0


Exercise 8.3. Let u ∈ D ′ (Ω). Show that there are C0∞ functions un with un → u in D ′ (Ω), n →
∞.

Exampel 8.6. An alternative proof that u is constant if u′ = 0.


Let uδ = u ∗ φδ ∈ C ∞ . Then u′δ = u′ ∗ φδ = 0 ∗ φδ = 0. Hence uδ = Cδ . But
uδ → u in D ′ , and Cδ → C for some constant C and u = C. 2
Exercise 8.4. Let u ∈ D ′ (R). Show that
a) If u′ ≥ 0, then u is an increasing function.
b) If u′′ ≥ 0, then u is a convex function.

Exampel 8.7. Harmonic functions.


If u ∈ C 2 (Rn ) satisfies ∆u = 0, we say that u is a harmonic function. Harmonic
functions satisfies the mean value property.
Z
1
u(x) = u(y)dσ(y).
|Sr (x)| Sr (x)
In R2 , this follows from the Cauchy integral formula since a harmonic function
locally is the real part of a holomorphic function. The general case follows from
Exercise 8.5. Prove the mean value property.
Hint. We may assume that x = 0. First apply Green’s identity to the functions u and 1 on
Br = {|x| ≤ r}, and then on u and E (E is the fundamental solution of ∆) on Ωϵ = {ϵ ≤ |x| ≤ 1}.
Let ϵ → 0. (A different proof is given in Section 17.2.)

Theorem 8.8 (Weyl’s lemma). If u ∈ D ′ (Rn ) and ∆u = 0, then u ∈ C ∞ and


∆u = 0 classically.
R
Proof. Let φδ be an approximate identity, φ(x) = φ(|x|), φ ≥ 0 and φ = 1.
Put uδ = u ∗ φδ . Then uδ ∈ C ∞ and ∆uδ = (∆u) ∗ φδ = 0 ∗ φδ = 0. So uδ satisfies
the mean value property. Hence
Z ∞ Z
n−1
uδ ∗ φ(x) = r φ(r)dr uδ (x − rω)dσ(ω)
0 S n−1
Z ∞ Z
n−1
= ωn uδ (x) r φ(r)dr = uδ (x) φ(y)dy = uδ (x).
0 Rn
Thus uδ = uδ ∗ φ. Now let δ → 0. We get u = u ∗ φ ∈ C ∞ and ∆u = ∆u ∗ φ =
0 ∗ φ = 0. □
Next we will define the convolution of two distributions. We want to do it in
such a way that the associativity is preserved. To be able to do that we assume
that at least one of the distributions has compact support.
42 8. CONVOLUTION OF DISTRIBUTIONS

Definition 8.9. Assume that u, v ∈ D ′ (Rn ), and at least one of them has
compact support. Then u ∗ v is the (uniquely determined) distribution that satisfies
(u ∗ v) ∗ φ = u ∗ (v ∗ φ), φ ∈ D(Rn ).
2
Is this a definition?
We first observe that u ∗ (v ∗ φ) is well-defined. If v has compact support, then
v ∗ φ ∈ D and u ∗ (v ∗ φ) is well-defined by Definition 1.1. On the other hand, if u
has compact support, then v ∗φ ∈ C ∞ and u∗(v ∗φ) is well-defined by Remark 1.2.
That there is at most one U = u ∗ v is also clear. Namely, if there were two

such distribution U and U e , then U ∗ φ = u ∗ (v ∗ φ) = U e ∗ φ and U (φ) = U ∗ φ (0)

=U e ∗ φ (0) = Ue (φ).
To show the existence we will study the map φ 7→ u ∗ φ.
Proposition 8.10. Let T φ = u ∗ φ. Then we have
a) If u ∈ D ′ (Rn ), then T is a continuous linear map D(Rn ) → C ∞ (Rn ).
b) If u ∈ E ′ (Rn ), then T is a continuous linear map D(Rn ) → D(Rn ) and
C ∞ (Rn ) → C ∞ (Rn ).
Proof. We prove a) and leave b) as an exercise. Thus we assume that φj → 0
i D(Rn ) and shall prove that ∂ α (u ∗ φj ) → 0 uniformly on compact sets. Since
∂ α φj → 0 in D(Rn ) if φj → 0 in D(Rn ), and ∂ α (u ∗ φj ) = u ∗ ∂ α φj , we may
assume that α = 0. If x is contained in a compact set and if all φj are supported
in another compact set, then also y 7→ φj (x − y) is supported in a fix compact set.
Thus
X
|u ∗ φj (x)| = |u(φj (x − ·)| ≤ C ∥∂ α φj (x − ·)∥∞ −→ 0, j → ∞.
|α|≤k


Exercise 8.6. Prove Proposition 10 b).
Let τh be the translation operator, τh φ(x) = φ(x − h). Then we have
Proposition 8.11. Convolution and translation commutes.
Proof.
u ∗ τh φ(x) = ⟨uy , τh φ(x − y)⟩ = ⟨uy , φ(x − h − y)⟩
= u ∗ φ(x − h) = τh (u ∗ φ)(x).

An important converse of this is
Theorem 8.12. Assume that T is a continuous linear map from D(Rn ) into
C (Rn ) that commutes with translations. Then there is a distribution u ∈ D ′ (Rn )

with
T φ = u ∗ φ, φ ∈ D(Rn ).
8. CONVOLUTION OF DISTRIBUTIONS 43
∨ ∨
Proof. If T φ = u ∗ φ, then in particular u(φ) = u∗ φ (0) = T φ (0). We
therefore define u by

u(φ) = T φ (0).
The continuity assumption implies that u is a distribution. Furthermore we have
∨ ∨
u ∗ φ(h) = ⟨u, φ(h − x)⟩ = ⟨u, τh φ⟩ = T ((τh φ)∨ )(0)
= T (τ−h φ)(0) = τ−h T (φ)(0) = T φ(h) .

The above results implies that Definition 9 is a definition. Proposition 10
shows that φ 7→ u ∗ (v ∗ φ) satisfies the conditions in Theorem 12 and, u ∗ v is this
distribution. 2
Remark 8.13.
a) If v ∈ D(Rn ), Definitions 1 and 2 coincides.
b) If both u and v have compact support, then (u ∗ v) ∗ φ = u ∗ (v ∗ φ) for
all φ ∈ C ∞ (Rn ).
2
Exampel 8.14. u ∗ δ = u since (u ∗ δ) ∗ φ = u ∗ (δ ∗ φ) = u ∗ φ. 2
Theorem 8.15.
a) u ∗ v = v ∗ u
b) supp (u ∗ v) ⊂ supp u + supp v
c) u ∗ (v ∗ w) = (u ∗ v) ∗ w
if at least two of the distributions have compact support.
Proof. a) To show that two distributions U and V coincides, it is enough
to show that U ∗ (φ ∗ ψ) = V ∗ (φ ∗ ψ), if φ, ψ ∈ D(Rn ). Namely, in that case,
(U ∗ φ) ∗ ψ = U ∗ (φ ∗ ψ) = V ∗ (φ ∗ ψ) = (V ∗ φ) ∗ ψ, according to Theorem 2.
This implies U ∗ φ = V ∗ φ, and U = V .
Now
(u ∗ v) ∗ (φ ∗ ψ) = u ∗ (v ∗ (φ ∗ ψ)) = u ∗ ((v ∗ φ) ∗ ψ)
= u ∗ (ψ ∗ (v ∗ φ)) = (u ∗ ψ) ∗ (v ∗ φ).
If v has compact support, the last equality follows by Theorem 8.2. If v does not
have compact support it follows from the next exercise.
We also have
(v ∗ u) ∗ (φ ∗ ψ) = (v ∗ u) ∗ (ψ ∗ φ) = (v ∗ φ) ∗ (u ∗ ψ) = (u ∗ ψ) ∗ (v ∗ φ),
and a) is proved.
b) By the commutativity we may assume that v has compact support. Define
∨ ∨ ∨
v by ⟨v, φ⟩ = ⟨v, φ⟩. If x ∈ supp (u ∗ v), there is to every ϵ > 0 a function

φ ∈ D(Rn ), supp φ ⊂ {y; |x − y| < ϵ} = Oϵ , with 0 ̸= u ∗ v(φ) = u ∗ v∗ φ (0)
44 8. CONVOLUTION OF DISTRIBUTIONS
∨ ∨ ∨
= u((v∗ φ)∨ ) = u(v ∗φ). So E = supp u ∩ supp (v ∗φ) ̸= ∅. Let y ∈ E. Then

y ∈ supp u och y ∈ supp v ∗φ, or y = −z + x + δ, where z ∈ supp v and |δ| < ϵ.
Thus x = y + z − δ ∈ supp u + supp v + Oϵ . Now let ϵ → 0.
c) Assume first that w has compact support. Then w ∗ φ ∈ D, and we get
((u ∗ v) ∗ w) ∗ φ = (u ∗ v) ∗ (w ∗ φ) = u ∗ (v ∗ (w ∗ φ)).
But also,
(u ∗ (v ∗ w)) ∗ φ = u ∗ ((v ∗ w) ∗ φ) = u ∗ (v ∗ (w ∗ φ))
and hence u ∗ (v ∗ w) = (u ∗ v) ∗ w.
If w does not have compact support, both u and v have, and a) implies
u ∗ (v ∗ w) = (v ∗ w) ∗ u = v ∗ (w ∗ u) = (w ∗ u) ∗ v
= w ∗ (u ∗ v) = (u ∗ v) ∗ w.

Exercise 8.7. Show that u ∗ (ψ ∗ φ) = (u ∗ ψ) ∗ φ if u ∈ E ′ , ψ ∈ D and φ ∈ C ∞ .

Theorem 8.16. ∂ α (u∗v) = ∂ α u∗v = u∗∂ α v if at least one of the distributions


have compact support.
Proof. If u ∈ D ′ (Rn ), we have ∂ α u = ∂ α δ ∗ u, since
∂ α u ∗ φ = u ∗ ∂ α φ = u ∗ (δ ∗ ∂ α φ) = u ∗ (∂ α δ ∗ φ) = (u ∗ ∂ α δ) ∗ φ.
Using this we get
∂ α (u ∗ v) = ∂ α δ ∗ (u ∗ v) = (∂ α δ ∗ u) ∗ v = ∂ α u ∗ v.
The second equality follows from Theorem 15 a). □
Theorem 8.17. Assume that u ∈ Dk′ and v ∈ C0k (or u ∈ Ek′ , v ∈ C k ). Then
u ∗ v is the continuous function x 7→ ⟨uy , v(x − y)⟩.
Proof. If x → x0 , then v(x − ·) → v(x0 − ·) in C0k . But u is continuous on
C0k ,
and we get ⟨uy , v(x − y)⟩ → ⟨uy , v(x0 − y)⟩. Thus h(x) = ⟨uy , v(x − y)⟩ is a
continuous function.
According to Definition 9, (u ∗ v) ∗ ψ = u ∗ (v ∗ ψ). As in the proof of Theorem 2
one can show that h ∗ ψ = u ∗ (v ∗ ψ). Hence h = u ∗ v.

Exercise 8.8. Let u, v ∈ D ′ (R) with support in {x ≥ 0}. Define u ∗ v.


Exercise 8.9. H 4.2.1
Exercise 8.10. H 4.2.2
Exercise 8.11. H 4.2.3
Exercise 8.12. H 4.2.4
CHAPTER 9

Fundamental solutions

Let X
P = aα ∂ α
|α|≤N
be a differential operator with constant coefficients and E a fundamental solution
to P , i.e. E ∈ D ′ (Rn ) and P E = δ. Then
P (E ∗ f ) = f, f ∈ E ′ (Rn ), (9.1)
and
E ∗ P u = u, u ∈ E ′ (Rn ). (9.2)
Thus E is both a left and a right invers to P on E . (But on different domains,

so it does not imply that P is bijective.) So (1) gives a solution u = E ∗ f of the


equation P u = f if f has compact support. (2) can be used to study regularity of
solutions of P u = f .
Remark 9.1. In Chapter 14 we will show that every differential operator with
constant coefficients has a fundamental solution. 2
In Chapter 5, we obtained fundamental solutions to the Laplace and heat
equations. Another example is that

 (x . . . x )k /(k!)n , all x > 0
1 n i
Ek (x) =
 0 otherwise,
is a fundamental solution to Pk+1 = ∂1k+1 . . . ∂nk+1 . Using this we can prove
Theorem 9.2. If u ∈ Em′ (Rn ), there is a continuous function f with
∂1m+2 . . . ∂nm+2 f = u.
Proof. Em+1 is a fundamental solution to Pm+2 . Thus f = Em+1 ∗ u satisfies
Pm+2 f = u. By Theorem 8.17, f is continuous. □
A corollary of Theorem 2 is the following representation theorem for distri-
butions.
Theorem 9.3. If u ∈ D ′ (Ω), there are functions fα ∈ C(Ω) with
X
u= ∂ α fα
in D ′ . The sum is locally finite, and if u has finite order the sum is finite.
45
46 9. FUNDAMENTAL SOLUTIONS

Proof. Choose a partition of unity ψi ∈ C0∞ and χi ∈ C0∞ with χi = 1 on


supp ψi . This can be done in such a way that Σχi is locally finite. We get
X X
u(φ) = ψi u(φ) = χi u(ψi φ).
i i

The distribution χi u has compact support, and hence finite order. Theorem 2
implies that χi u = ∂ αi fi , fi ∈ C. Hence
X X Z
αi |αi |
u(φ) = ∂ fi (ψi φ) = (−1) fi ∂ αi (ψi φ) dx .
i i Rn

If we compute ∂ αi (ψi φ), we get


XX Z XX
|α|
u(φ) = (−1) fi,α ∂ α φ dx = ∂ α fi,α (φ) .
i α Rn α i
P
Finally, we let fα = i fi,α . □
To study the regularity of solutions of the equation P u = f , we want to study
the set where u is not C ∞ .
Definition 9.4. The singular support of a distribution u ∈ D ′ (Ω) is denoted
sing supp u, and consists of those points in Ω that have no neighborhood where u
is C ∞ . 2
sing supp u is the smallest closed set such that u is C ∞ in its complement. It
is clear that sing supp u ⊂ supp u.
Theorem 9.5. If u, v ∈ D ′ (Rn ), and at least one of them have compact support,
then
sing supp(u ∗ v) ⊂ sing supp u + sing supp v. (9.3)
Proof. Put u1 = u and u2 = v. Let us first assume that both distributions
have compact support. Let Ki = sing supp ui , and Ωi a neighborhood of Ki and
take ψi ∈ C0∞ (Ωi ) with ψi = 1 on Ki . Then
u1 ∗ u2 = (ψ1 u1 + (1 − ψ1 )u1 ) ∗ (ψ2 u2 + (1 − ψ2 )u2 )
= ψ1 u1 ∗ ψ2 u2 + ψ1 u1 ∗ (1 − ψ2 )u2
+ (1 − ψ1 )u1 ∗ ψ2 u2 + (1 − ψ1 )u1 ∗ (1 − ψ2 )u2 .
Since (1 − ψi )ui ∈ C0∞ , Theorem 8.1 implies that the last three terms are C ∞ .
Thus
sing supp(u1 ∗ u2 ) = sing supp(ψ1 u1 ∗ ψ2 u2 )
⊂ supp(ψ1 u1 ∗ ψ2 u2 ) ⊂ supp ψ1 + supp ψ2 ⊂ Ω1 + Ω2 .
If we let Ωi ↓ Ki , we obtain (9.3).
9. FUNDAMENTAL SOLUTIONS 47

If only one of the distributions have compact support, we can by Theorem 8.15
asssume that v ∈ E ′ .
To show that sing supp u ∗ v ⊂ sing supp u + sing supp v, it is enough to show
that sing supp u ∗ v ∩ B1 (x) ⊂ (sing supp u + sing supp v) ∩ B1 (x) for each x ∈ Rn .
Take R ≥ 1 so large that supp v ⊂ BR (0) and |x| ≤ R, and choose χ ∈

C0 (B6R (0)) with χ = 1 on B5R (0). Put u1 = χu and u2 = (1 − χ)u. Thus
C
u = u1 + u2 where u1 has compact support and supp u2 ⊂ B5R (0). Then supp u2 ∗
C C
v ⊂ B5R (0) + BR (0) ⊂ B4R (0) and B1 (x) ⊂ B2R (0). Hence u2 ∗ v = 0 on B1 (x).
Since both u1 and v have compact support, we get
sing supp (u ∗ v) ∩ B1 (x) = sing supp (u1 ∗ v) ∩ B1 (x)
⊂ (sing supp u1 + sing supp v) ∩ B1 (x),
and the theorem is proved. □
Theorem 9.6. If P has a fundamental solution with supp E = {0}, then
sing supp u = sing supp P u, u ∈ D ′ .
Remark 9.7. The converse is also true. Thus if there is a fundamental solution
with singular support at the origin, then this is true for all fundamental solutions.
2
Proof. sing supp P u ⊂ sing supp u allways holds since P u is C ∞ if u is.
For the other inclusion, we first observe that if u has compact support, then
u = E ∗ P u and by Theorem 5
sing supp u ⊂ sing supp E + sing supp P u = sing supp P u.
If u is not compactly supported, take ψ ∈ C0∞ with ψ = 1 on an open set Ω. Then
sing supp ψu ⊂ sing supp P (ψu).
But on Ω we have P (ψu) = P u and ψu = u, and the result follows. □
A differential operator P is called hypoelliptic if every solution u of P u = f is

C if f is. Theorem 6 thus implies that P is hypoelliptic if P has a fundamental
solution E with sing supp E = {0}.
The Laplace and the heat operators are hypoelliptic. In Chapter 12, we will
show that all elliptic operators are hypoelliptic. The Laplace operator is elliptic
but not the heat operator.
d
Exercise 9.1. Let P be a polynomial of one variable. Show that P ( dx ) has a fundamental
solution.
Exercise 9.2. H 4.4.3
Exercise 9.3. H 4.4.4
Exercise 9.4. H 4.4.5
Exercise 9.5. H 4.4.6
Exercise 9.6. H 4.4.9
CHAPTER 10

The Fourier transform

If u is a ”nice” periodic function with period T , u can be written as


X
u(x) = cm e2πimx/T . (10.1)
m

Then X
u(x)e−2πiνx/T = cν e2πi(m−ν)x/T ,
ν

and integration over [− T2 , T2 ]gives formally that


Z T
2
T cν = u(x)e−2πiνx/T dx
− T2
or
Z T
1 2
cν = u(x)e−2πiνx/T dx
T − T2

cν are the Fourier coefficients of u. (1) is the inversion theorem. We also have
Parseval’s identity
X 1
|cν |2 = ∥u∥22 .

How can this be generalised to Rn ? Let us first consider the case n = 1. Let
u ∈ C0∞ (R) and choose T so that supp u ⊂ (− T2 , T2 ). Let uT be the periodic
extension of u, X
uT (x) = u(x − kT ).
k∈Z
Then we have
Z T
1 2
cT (ν) = u(x)e−2πiνx/T dx.
T − T2

Thus for |x| < T2 , (1) implies


X
u(x) = uT (x) = cT (ν)e2πiνx/T .
ν

Define Z
u
b(ξ) = u(x)e−iξx dx, ξ ∈ R.
R
49
50 10. THE FOURIER TRANSFORM

b( 2πν
We observe that cT (ν) = T1 u T
), and we can write
X 1 2πν 1 X 2π 2πν ix 2πν
u(x) = u
b( )e2πiνx/T = u
b( )e T .
ν
T T 2π ν
T T
This is a Riemann sum of the integral
Z
1
b(ξ)eiξx dξ.
u
2π R
So, if we let T → ∞, we obtain
Z
1
u(x) = b(ξ)eiξx dξ.
u (10.2)
2π R

With some care, the above argument can be used to prove (2) when u is a nice
function. We will not do this, but instead prove (2) (and its generalization to Rn )
directly. The theory for Fourier series will then be a corollary of the theory of the
Fourier transform.
Definition 10.1. Assume that f ∈ L1 (Rn ). The Fourier transform of f is
defined by Z
fb(ξ) = f (x)e−ixξ dx,
Rn
Pn
where xξ = 1 xi ξi . We sometimes write Ff instead of fb.
2
We will prove the following important properties of the Fourier transform.
I. The inversion formula. If f and fb ∈ L1 , then
Z
1
f (x) = fb(ξ)eixξ dξ.
(2π)n Rn
1
II. Parseval’s identity. If f ∈ L1 ∩ L2 , then fb ∈ L2 and ∥f ∥2 = (2π)n
∥fb∥2 .
III. If f, g ∈ L1 , then (f ∗ g)∧ = fb gb.
IV. F(P (D)f )(ξ) = P (ξ)fb(ξ) where Dj = −i∂j .
Exercise 10.1. Prove the Riemann-Lebesgue lemma: If f ∈ L1 , then fb is continuous and
fb(ξ) → 0 when |ξ| → ∞.
Exercise 10.2. Prove III and IV.

To solve the constant coefficient differential equation P (D)u = f , we can use


I and IV. By Fourier transformation, we get P (ξ)b u(ξ) = fb(ξ). Thus u b(ξ) =
−1 b
f (ξ)/P (ξ) and u = F (f /P ). To be able to use this method ”often”, we want to
b
extend the Fourier transform to distributions. As a motivation for the definition,
we observe that, by Fubini’s theorem we have
10. THE FOURIER TRANSFORM 51
Z Z
1
Proposition 10.2. If f, g ∈ L , then f gb dx = fbg dx.
Rn Rn

Exercise 10.3. Prove Proposition 2.


Hence for a L1 function we Z have Z
⟨fb, φ⟩ = fbφ dx = b dx = ⟨f, φ⟩.
fφ b
Rn Rn
b when u ∈ D ′ by
It is therefore natural to try to define u
⟨b
u, φ⟩ = ⟨u, φ⟩.
b (10.3)
But if φ ∈ D, φ ̸≡ 0, then φ
b can not have compact support and hence φ b∈/ D. So
we can not define u
b by (3).
So what to do? Well, we will consider a different class of test functions, that is
preserved by the Fourier transform. This is the Schwartz space S , that consists
of C ∞ rapidly decreasing functions.
Definition 10.3.
(a) φ ∈ S (Rn ) if φ ∈ C ∞ and supx∈Rn (1 + |x|2 )k |∂ α φ(x)| < ∞ for all k and
α.
(b) φj → 0 in S (Rn ) if
sup (1 + |x|2 )k |∂ α φj (x)| → 0
x∈Rn
for all k and α. 2
Definition 10.4.
(a) A tempered distribution on Rn is a linear functional on S , such that
u(φj ) → 0 when φj → 0 in S . We write u ∈ S ′ .
(b) A sequence uj ∈ S ′ converges to u ∈ S ′ if
uj (φ) → u(φ),
for every testfunction φ ∈ S .

2
To show that (3) works as a definition of the Fourier transform if u ∈ S ′ we
need to study the Fourier transform on S . We start with the following
Proposition 10.5. If f, g ∈ S , then
(a) F(xα f (x)) = iα ∂ α fb
(b) (∂ α f )∧ (ξ) = (iξ)α fb(ξ)
(c) (τh f )∧ (ξ) = e−iξh fb(ξ)
(d) F(eixh f (x)) = τh fb
(e) (fa )∧ (ξ) = fb(aξ)
(f) (f (ax))∧ = (fb)a
52 10. THE FOURIER TRANSFORM
Z Z
(g) f gb = fbg
Rn Rn
(h) (f ∗ g)∧ = fb gb
and
(i) fb ∈ S .
Exercise 10.4. Prove Proposition 5.

Theorem 10.6 (The inversion theorem). If f ∈ S , then


Z
1
f (x) = n
eixξ fb(ξ)dξ. (10.4)
(2π) Rn
To prove this we need to find one function that satisfies (4). Then (4) follows
2
in general by Proposition 5. We make the choice G(x) = e−|x| /2 .
b = (2π)n/2 G.
Lemma 10.7. G
Proof. By Fubinin’s theorem it is enough to consider the case n = 1. G
satisfies the differential equation
G′ (x) + xG(x) = 0.
If we take the Fourier transform of this equation, Proposition 5 (a) and (b) implies
iξ G(ξ) b′ (ξ) = 0,
b + iG
or
b′ (ξ) + ξ G(ξ)
G b = 0.
Hence G(ξ)
b = CG(ξ). If we let ξ = 0, we get
Z
2 √
C = G(0)
b = e−x /2 dx = 2π.
R

Exercise 10.5.
a) Prove Lemma 7 using the Cauchy theorem.
b) Prove Lemma 7, letting ξ = ζ ∈ C, and compute G(iη).
b
1
Proof of Theorem 6. (2π) n (G)δ is an approximate identity (but without
b
compact support). Proposition 5 f) and g) implies that
Z Z
1 1
f (x)(G)
b δ (x)dx = fb(ξ)G(δξ)dξ.
(2π)n Rn (2π)n Rn
Letting δ → 0, we get
Z Z
1 1
f (0) = G(0) fb(ξ)dξ = fb(ξ)dξ.
(2π)n Rn (2π)n Rn
Exercise 10.6. Prove this.
10. THE FOURIER TRANSFORM 53

If we apply this to τ−x f , we get


Z Z
1 ∧ 1
f (x) = τ−x f (0) = (τ−x f ) (ξ)dξ = fb(ξ)eixξ dξ.
(2π)n Rn (2π)n Rn

Remark 10.8. If we only assume that f ∈ L1 , then
Z Z
1 1 1 2 2

n
f ∗ (G)δ (x) =
b
n
f (x + y)Gδ (y)dy =
b
n
fb(ξ)eixξ e−δ |ξ| /2 dξ.
(2π) (2π) Rn (2π) Rn
This implies
Z
1 2 2
f (x) = lim fb(ξ)eixξ e−δ |ξ| /2 dξ,
δ→0 (2π)n Rn

with convergence in L1 .
In particular, if fb ∈ L1 , then
Z
1
f (x) = fb(ξ)eixξ dξ a. e.
(2π)n Rn
2
Theorem 10.9 (Plancherel). If ϕ, ψ ∈ S , then
Z Z
1
ϕψ̄ dx = ϕb ψb dξ .
Rn (2π)n Rn
Corollary 10.10 (Parseval). If ϕ ∈ S , then
1
∥ϕ∥2 = ∥ϕ∥
b 2
(2π)n/2
.
Proof. Proposition 2 g) implies
Z Z
ϕψ0 dx =
b ϕψ
b 0 dx .
Rn Rn

Let ψb0 = ψ̄. By the inversion theorem,


Z Z
1 ixξ 1
ψ0 (x) = e ψb0 (ξ)dξ = eixξ ψ̄(ξ)dξ
(2π)n Rn (2π)n Rn
Z
1 1 b
= n
e−ixξ ψ(ξ)dξ = ψ(x).
(2π) Rn (2π)n
Z Z
1
Thus ϕψ̄ = ϕb ψ.
b The corollary follows by taking ψ = ϕ. □
Rn (2π)n Rn
54 10. THE FOURIER TRANSFORM

Remark 10.11. The Parseval formula also holds if ϕ, ψ ∈ L2 . We will prove


this in the next chapter. 2
To prove that ub, defined by u
b(φ) = u(φ),
b is a tempered distribution we need
the following
Lemma 10.12. F : S → S continuously, i.e., if φj → 0 in S , then φ bj → 0
in S .

Proof. Proposition 5 a) and b) implies ξ β ∂ α φ bj (ξ) = cF ∂ β (xα φj (x)) (ξ) .
Hence
Z
β α
sup |ξ ∂ φbj (ξ)| ≤ c sup e−ixξ ∂ β (xα φj (x))dx
ξ ξ R n
Z
≤c |∂ β (xα φj (x))|dx → 0,
Rn

as φj → 0 in S . □

We are now ready to make the following definition.


Definition 10.13. If u ∈ S ′ , then u
b is the tempered distribution given by
u
b(φ) = u(φ).
b
2
Remark 10.14. We observe that the two definitions of fb when f ∈ L1 coincides.2
Theorem 10.15. The Fourier transform is a continuous linear bijection from

S to S ′ with u

b = (2π)n u.
b
∨ ∨ ∨ ∨
Proof. We remind the reader that u is defined by u(φ) = u(φ) and φ(x) =
φ(−x), and that uj → u in S ′ means that uj (φ) → u(φ) for all φ ∈ S . The
theorem is an easy consequence of the corresponding properties on S :

u
b(φ) = u
b b(φ) bb = (2π)n u(φ
b = u(φ) )
and
u b → u(φ)
bj (φ) = uj (φ) b =u
b(φ)
if uj → u i S ′ . □
Exampel 10.16. a) A measure µ with Rn (1 + |x|2 )−k dµ(x) < ∞ for some k
R
is a tempered distribution.
b) If f ∈ Lp , 1 ≤ p ≤ ∞, then f ∈ S ′ . (Proof. Hölder’s inequality.)
c) δb = 1 and b1 = (2π)n δ.
x
d) e is not a tempered distribution. 2
10. THE FOURIER TRANSFORM 55

Proposition 10.17. If u ∈ S ′ , then


a) (xj u)∧ = −Dj u b
b) (Dj u)∧ = ξj ub
c) (τh u)∧ (ξ) = exp(−ihξ)b
u(ξ)
and
d) F(exp(ixh)u) = τh u b.
Proof. It is easy to see that Dj u, xj u, . . . are tempered distributions. Then
the formulas follows from Proposition 5. (Remember that Dj = −i∂j .) □
Exercise 10.7. Show that ex cos(ex ) ∈ S ′ .
Exercise 10.8. Show that u ∈ S ′ if and only if
X
|u(φ)| ≤ C sup(1 + |x|2 )k |∂ α φ(x)|
k+|α|≤N

for some N .
Exercise 10.9. H 7.1.10
Exercise 10.10. H 7.1.19
Exercise 10.11. H 7.1.20
Exercise 10.12. H 7.1.21
Exercise 10.13. H 7.1.22
Exercise 10.14. H 7.6.1
CHAPTER 11

The Fourier transform on L2

According to Exempel 10.16b), a function f ∈ L2 has a Fourier transform


defined as a tempered distribution. In fact we have the following result.
Theorem 11.1. If f ∈ L2 (R2 ), then fˆ ∈ L2 (Rn ) and
1
∥f ∥2 = ∥fˆ∥2 .
(2π)n/2
Furthermore fˆ is given by
Z
ˆ
f (ξ) = lim e−iξx f (x)dx ,
N →∞ |x|≤N
2
with convergence in L .
Proof. Take fn ∈ C0∞ , fn → f in L2 . Then fn is a Cauchy sequence in L2 .
By the Plancherel theorem, we have
∥fˆn − fˆm ∥2 = c∥fn − fm ∥2 → 0, n, m → ∞.
Hence fˆn is a Cauchy sequence in L2 . By the completeness of L2 , we have fˆn → g
in L2 for some g ∈ L2 . This implies that fˆn → g in S ′ . Furthermore fn → f in
S ′ , and since the Fourier transform is continuous on S ′ , we have fˆn → fˆ in S ′ .
Hence fˆ = g ∈ L2 and we get
1 1
∥fˆ∥2 = lim ∥fˆn ∥2 = n/2
lim ∥fn ∥2 = ∥f ∥2 ,
n→∞ (2π) n→∞ (2π)n/2
and the first part is proved.
Put fN = f χ{|x|≤N } . Then fN → f in L2 and fN ∈ L1 . Hence
Z
ˆ
fN (ξ) = e−ixξ f (x)dx,
|x|≤N

and by the Plancherel theorem, we obtain


∥fˆ − fˆN ∥2 = c∥f − fN ∥2 → 0, N → ∞,
and the proof is complete.

57
CHAPTER 12

The Fourier transform and convolutions

We shall show that under suitable conditions (u ∗ v)∧ = u


b vb.
First we observe that D ⊂ S ⊂ C ∞ . The inclusions are continuous, ie. if
φj → 0 i D, then φj → 0 in S , and this in turn implies that φj → 0 in C ∞ .
Furthermore, D is dense in S , and S is dense in C ∞ . (Show that!) Hence

E ′ (Rn ) ⊂ S ′ (Rn ) ⊂ D ′ (Rn ).

Definition 12.1. If u ∈ S ′ and ϕ ∈ S , we define the convolution u ∗ ϕ by


u ∗ ϕ(x) = ⟨uy , ϕ(x − y)⟩. 2

Theorem 12.2. If u ∈ S ′ and ϕ ∈ S , then


(a) u ∗ ϕ ∈ C ∞ och ∂ α (u ∗ ϕ) = ∂ α u ∗ ϕ = u ∗ ∂ α ϕ
(b) u∗ϕ is bounded by a polynomial (and hence u∗ϕ ∈ S ′ ), and (u∗ϕ)∧ = ϕb u
b.
(c) u ∗ (ϕ ∗ ψ) = (u ∗ ϕ) ∗ ψ (ψ ∈ S )
and
b ∗ ϕb = (2π)n (ϕu)∧ .
(d) u

Sketch of proof. (a) We assume that n = 1. The second equality is proved


in the same way as in Theorem 8.1. As in the proof of Theorem 8.1, the first
equality folows if we can prove that
ϕ(x + h) − ϕ(x)
−→ ϕ′ (x) in S .
h
To do this is elementary but tedious. The simplest way is (probably) to use the
Fourier transform.
(b) By Exercise 10.8, we have

|u ∗ ϕ(x)| = |⟨uy , ϕ(x − y)⟩|


X
≤ C sup (1 + |y|2 )k |∂ α ϕ(x − y)|
y
k+|α|≤N
X
≤ C sup (1 + |x|2 )k (1 + |x − y|2 )k |∂ α ϕ(x − y)|
y
k+|α|≤N

≤ C(1 + |x|2 )N .
59
60 12. THE FOURIER TRANSFORM AND CONVOLUTIONS

If ψ ∈ D, we also have

Z
∧ n n
b = (u ∗ ϕ)(ψb ) = (2π) (u ∗ ϕ)(ψ) = (2π)
(u ∗ ϕ) (ψ) u ∗ ϕ(x)ψ(−x)dx
b
Rn
Z
n
= (2π) ⟨uy , ψ(−x)ϕ(x − y)⟩dx = Approximate with a Riemann sum =
−K
Z Z
n n
= (2π) ⟨uy , ψ(−x)ϕ(x − y)dx⟩ = (2π) ⟨uy , ψ(x)ϕ(−y − x)dx⟩
Rn Rn

= (2π)n ⟨uy , (ϕ ∗ ψ)∨ )⟩ = ⟨uy , (ϕ ∗ ψ)∧ ⟩ = u
b((ϕ ∗ ψ)∧ ) = u
b(ϕbψ)
b = ϕb u
b(ψ).
b

But Db is dense in S , and (b) follows.


(c) From the proof of (b), we get

u ∗ ϕ(ψ) = u((ϕ ∗ ψ)∨ ),


first for ψ ∈ D, and by contiuity also for ψ ∈ S . This can be written
(u ∗ ϕ) ∗ ψ(0) = u ∗ (ϕ ∗ ψ)(0).
The general case follows if we replace ψ with τ−x ψ.


(d) By (b), we have (b b ∧ = ϕb
u ∗ ϕ)
bb
b = (2π)2n ϕu= (2π)2n (ϕu)∨ = (2π)n (ϕu)b.
u
b

Thus, by the inversion theorem we get u b ∗ ϕb = (2π)n (ϕu)b.



Theorem 12.3. If u ∈ E ′ (Rn ), then u
b ∈ C ∞ and u
b(ξ) = ux (e−ixξ ).
Proof. Let ψ ∈ C0∞ be 1 on a neighborhood of supp u. Then u = ψu and by
b = (ψu)b = (2π)−n u
Theorem 2(d) we get u b ∗ ψb ∈ C ∞ . Thus

−n b = (2π)−n ⟨b b − x)⟩ = (2π)−n ⟨b
u
b(ξ) = (2π) b ∗ ψ(ξ)
u ux , ψ(ξ ux , ψb (x − ξ)⟩
= (2π)−2n ⟨b
ux , F 3 ψ(x − ξ)⟩ = (2π)−2n ⟨b
ux , τξ F 3 ψ(x)⟩
= (2π)−2n ⟨ux , e−ixξ F 4 ψ(x)⟩ = ux (e−iξx ψ(x)) = ux (e−iξx ).

Remark 12.4. In the next chapter we will prove the Paley-Wiener theorem
b when u ∈ E ′ .
that gives much more precise information of u 2
Exampel 12.5. Determine the Fourier transform of pv x1 .
Method 1. Let u = pv x1 . Then u is the sum of a distribution in E ′ and an L2
function. Thus Z
dx
u
b(ξ) = lim e−ixξ .
ϵ→0
N →∞ ϵ<|x|<N
x
12. THE FOURIER TRANSFORM AND CONVOLUTIONS 61

If ξ > 0, the change of variables y = xξ implies,


Z ∞
e−iy
Z
sin y
u
b(ξ) = lim dy = −i dy = −iπ.
ϵ→0
N →∞ ϵ<|x|<N
y −∞ y

When ξ < 0, we instead get u


b(ξ) = iπ. Hence
b(ξ) = −π sgn ξ.
u
R∞sin y
Exercise 12.1. Prove that −∞ y
dy = π.

1
Method 2. We have x pv = 1. This implies ibu′ = 2πδ, u
b′ = −2πiδ and
x
b. Hence c = − 12 och
b = −2πi(H + c). Since u is odd, so is u
u
b(ξ) = −iπ sgn ξ.
u
2
In the last argument we used that if a distribution u is odd, then u
b is also odd.
This is clear if u ∈ L1 (a simple change of variables).
∨ ∨
Definition 12.6. A distribution is even if u= u, and odd if u= −u. 2
Proposition 12.7. If u is an odd tempered distribution, then its Fourier trans-
form is also odd.
Proof. By Theorem 10.15
∨  ∧ b

b= (2π)−n u
u b = (2π)−n u
b = u = (u is odd) = −b
u.
b
b b


In the same way we see that the Fourier transform of an even distribution is
even.
1
Remark 12.8. The map Hφ = pv ∗ φ, is called the the Hilbert transform.
x
The Hilbert transform is an important example of a so called singular integral
operator. The Hilbert transform is bounded on Lp , 1 < p < ∞, and of weak-type
(1, 1).
When p = 2, this follows from Exampel 5 and the Plancherel theorem. 2
Next we will study invariance properties of the Fourier transform. Let F :
Rn → Rn be a diffeomorphism (i.e. a C ∞ bijection). If u is a function, we have
Z Z
φ
u ◦ F (x)φ(x)dx = u(y) ′ ◦ F −1 (y)dy.
Rn Rn |F |
Therefore, if u ∈ D ′ , we define u ◦ F by
φ
⟨u ◦ F, φ⟩ = ⟨u, ◦ F −1 ⟩
|F ′ |
62 12. THE FOURIER TRANSFORM AND CONVOLUTIONS

In particular, if F = Λ is linear, then


⟨u ◦ Λ, φ⟩ = | det Λ|−1 ⟨u, φ ◦ Λ−1 ⟩
2
Definition 12.9. A distribution u is radial if u ◦ O = u for all orthogonal
maps O.
Theorem 12.10. If u is a radial tempered distribution, then u
b is radial.
Proof. First, we observe that if φ ∈ S , then
Z
b ◦ O(ξ) = φ(Oξ)
φ b = e−ixOξ φ(x)dx
Rn
 

y = O x
Z

= e−iO xξ φ(x)dx =  
Rn x = Oy
Z
= e−iyξ φ(Oy)dy = (φ ◦ O)∧ (ξ).
Rn
This implies
⟨b u, φ ◦ O−1 ⟩ = ⟨u, (φ ◦ O∗ )∧ ⟩
u ◦ O, φ⟩ = ⟨b
b ◦ O∗ ⟩ = ⟨u, φ
= ⟨u, φ b ◦ O−1 ⟩ = ⟨u ◦ O, φ
b ⟩ = ⟨u, φ
b ⟩ = ⟨b
u, φ⟩.

Theorem 12.11. If u is a tempered distribution that is homogeneous of degree
b is homogeneous of degree −n − α.
α, then u
Proof. By Definition 4.4, u ∈ S ′ is homogeneous of degree α if ⟨u, φt ⟩ =
tα ⟨u, φ⟩. Therefore,
⟨b
u, φt ⟩ = ⟨u, φ
bt ⟩ = ⟨uξ , φ(tξ)⟩
b = t−n ⟨u, (φ)
b 1/t ⟩
= t−(n+α) ⟨u, φ⟩
b = t−(n+α) ⟨b
u, φ⟩.

Exampel 12.12. A fundamental solution of the Laplace operator when n ≥ 3.
By Fourier transformation of
∆u = δ,
we get
−|ξ|2 u
b(ξ) = 1.
One solution is
1
b(ξ) = −
u .
|ξ|2
12. THE FOURIER TRANSFORM AND CONVOLUTIONS 63

Observe that |ξ|12 ∈ L1loc (Rn ) if n ≥ 3, and that |ξ|12 is radial and homogeneous of
degree−2. Hence u is radial and homogeneous of degree 2 − n. This implies
cn
u(x) = n−2 .
|x|
Argument. If n = 3, then |ξ|12 ∈ L1 + L2 , and thus u is a function. If n > 4,
1
then |x|n−2 ∈ L1 + L2 , and we can argue as above using the inversion theorem.
1
When n = 4, we have uϵ = ∈ L1 + L2 , and thus its Fourier transform is a
|x|2+ϵ
1
constant times 2−ϵ and the statement follows by letting ϵ → 0.
|ξ|
An alternative way is to use Exercise 4 below.
Exercise 12.2. What is cn
Exercise 12.3. What happens if n = 2?
Exercise 12.4. Determine all radial distributions in Rn that are homogeneous of degree α.
Hint. Consider first n = 1. Compute the derivate of ⟨u, φt ⟩ = tα ⟨u, φ⟩ with respect to t.
Warning. Be careful when −α = n, n + 2, n + 4, . . ..
Exercise 12.5. What is the Fourier transform of fp |x|α in R?
Exercise 12.6. What is a reasonably definition of fp |x|α in Rn ? What is its Fourier transform?
Exercise 12.7. Determine a fundamental solution to the heat equation.
Hint. Determine FE(x, t), where F is the Fourier transform with respect to x ∈ Rn .

Theorem 12.13. If u ∈ S ′ and v ∈ E ′ , then u ∗ v ∈ S ′ and


(u ∗ v)∧ = vb u
b.
Proof. If φ ∈ C0∞ , then
∨ ∨ ∨ ∨
u ∗ v(φ) = (u ∗ v)∗ φ (0) = u ∗ (v∗ φ)(0) = u((v∗ φ)∨ ) = u(v ∗φ).

To see that u ∗ v ∈ S ′ , we need to show that v ∗φj → 0 in S when φj → 0 in
S . Let K be a compact neigborhood of av supp v and k the order of v. Then
∨ X
|∂ β (v ∗φj )(x)| = |⟨vy , ∂ β φj (y − x)⟩| ≤ C sup |∂ β+γ φj (x − y)|.
y∈K
|γ|≤k

Thus
∨ X
(1 + |x|2 )ℓ |∂ β (v ∗φj )(x)| ≤ C (1 + |x|2 )ℓ sup |∂ β+γ φj (x − y)| → 0, j → ∞.
y∈K
|γ|≤k

To compute the Fourier transform, we observe that


(u ∗ v)∧ (φ) = u ∗ v(φ)
b = u(v ∗φ) b = (2π)−n u(b
vb ∗ φ)
b

= u((b
v φ) ) = u
b(bv φ) = vb u
b(φ).
64 12. THE FOURIER TRANSFORM AND CONVOLUTIONS

2

∗n
Exercise 12.8. 1
Compute ( 1+x 2) and (e−x )∗n .
Exercise 12.9. H 7.1.6
Exercise 12.10. H 7.1.7
Exercise 12.11. H 7.1.9
Exercise 12.12. H 7.1.11
Exercise 12.13. H 7.1.18
Exercise 12.14. H 7.1.28
CHAPTER 13

The Paley-Wiener theorem

If u ∈ E ′ (Rn ), we know that û ∈ C ∞ and that


û(ξ) = u(e−ixξ ), ξ ∈ Rn .
We shall show that û can be extended to an analytic function in Cn , i.e.
û(ζ1 , . . . , ζn ) is an analytic function in each variable ζ1 , . . . , ζn . We start with
a version of the theorem for test functions.
Proposition 13.1.

(a) If ϕ ∈ C0∞ and supp ϕ ⊂ {x; |x| ≤ R}, then


Z
ϕ̂(ζ) = e−ixζ ϕ(x)dx
Rn
is an entire function with
|ϕ̂(ζ)| ≤ CN (1 + |ζ|)−N eR|Im ζ| (13.1)
for all N .
(b) Conversely, if ϕ̂ is entire and satifies (1), then ϕ ∈ C0∞ and
supp ϕ ⊂ {x; |x| ≤ R}.
Proof. (a) By differentiation under the integral sign, we see that ϕ̂ is analytic.
Furthermore, if ζ = ξ + iη,
Z
|ϕ̂(ζ)| ≤ exη |ϕ(x)|dx ≤ CeR|η| . (13.2)
|x|≤R
α
If we apply (2) to D ϕ, (1) follows.
(b) Since ϕ̂ is rapidly decreasing we can differentiate under the integral sign in
the Fourier inversion formula. Hence ϕ ∈ C ∞ .
Again, by the rapid decrease of ϕ̂ we can use the Cauchy theorem to change
the contour of integration and integrate along {ζ; Im ζi = ηi }. We get
Z
−n
|ϕ(x)| = (2π) eix(ξ+iη) ϕ̂(ξ + iη)dξ ≤ Ce−xη eR|η| .
Rn
If we let η = tx, we obtain
|ϕ(x)| ≤ Ce−t|x|(|x|−R) .
65
66 13. THE PALEY-WIENER THEOREM

Thus if |x| > R, and we let t → ∞, we obtain ϕ(x) = 0. Hence


supp ϕ ⊂ {x; |x| ≤ R}. □

For distributions we have


Theorem 13.2 (The Paley-Wiener thorem).

(a) If u is a distribution of order N with support in {x; |x| ≤ R}, then û is


an entire function and
|û(ζ)| ≤ C(1 + |ζ|)N eR |Im ζ| . (13.3)
(b) Conversely, if û is an entire function that satisfies (3) for some N , then
u is a distribution that is supported in {x; |x| ≤ R}.
Proof. (a) That û is entire follows since
∂ ∂  ∂ 
û(ζ) = u(e−ixζ ) = u (e−ixζ ) .
∂ζi ∂ζi ∂ζi
The last equality holds as
e−ix(ζ+ωi ) − e−ixζ ∂ −ixζ
→ (e ) in C ∞ , ωi → 0.
ωi ∂ζi
To prove (3), we fix χδ ∈ C0∞ with χδ = 1 in a neighborhood of {x; |x| ≤ R}
and supp χδ ⊂ {x; |x| < R + δ}. We can choose χδ such that ∥Dα χδ ∥∞ ≤ Cδ −|α| .
We obtain
|û(ζ)| = |u(e−ixζ )| = |u(χδ (x)e−ixζ )|
X
≤ CN sup |Dxα (χδ (x)e−ixζ )|
|α|≤N
X
≤ Ce(R+δ)| Im ζ| δ −|β| (1 + |ζ|)N −β .
|β|≤N
1
If we let δ = ,
(3) follows.
1+|ζ|
(b) The polynomial growth of û implies that û, and hence also u, is in S ′ . Let
φδ ∈ S be an approximative identity and let uδ = u ∗ φδ . Then uδ ∈ C ∞ , uδ → u
as δ → 0, and
|ûδ (ζ)| = |û(ζ)φ̂(δζ)| ≤ CM,δ (1 + |ζ|)−M exp((R + cδ)| Im ζ|).
Here we have used (3) and (1) in Proposition 1. If we apply Proposition 1 to
uδ , we get supp uδ ⊂ {x; |x| ≤ (R + cδ)}. If we let δ → 0, we obtain supp
u ⊂ {x; |x| ≤ R}. □
Exercise 13.1. Assume that u, v ∈ E ′ (Rn ) and that u ∗ v = 0. Show that then u = 0 or v = 0.
What happens if only one of u and v have compact support?
Exercise 13.2. H 7.1.40.
CHAPTER 14

Existence of fundamental solutions

Let P (D) be a differential operator with constant coefficients in Rn . We shall


show that P (D) has a fundamental solution E.
Let us first make a formal computation. By Fourier transformation of P (D)E =
δ, we get P (ξ)E(ξ)
b = 1 and E(ξ)
b = P (ξ)−1 . Now

∨ ∨
b ∨
⟨E, φ⟩ = ⟨E, ⟩ = (2π)−n ⟨E, φ⟩
φ b = (2π)−n ⟨E,
b φ⟩
b .
Hence it is natural to define E by
Z
−n
⟨E, φ⟩ = (2π) P (ξ)−1 φ(−ξ)dξ.
b
Rn

Then (formally)
Z
−n
⟨P (D)E, φ⟩ = ⟨E, P (−D)φ⟩ = (2π) P (ξ)−1 (P (−D)φ)b(−ξ)dξ
Rn
Z
= (2π)−n P (ξ)−1 P (ξ)φ(−ξ)dξ
b = φ(0) = ⟨δ, φ⟩.
Rn

However, this does not always work since P (ξ) may vanish. Therefore, we will
change the contour of integration and define ⟨E, φ⟩ by an integral along a set in
Cn that contains no zero of P .
Theorem 14.1. Every linear differential operator with constant coefficients has
a fundamental solution E ∈ D ′ .
Proof. Let m = grad P . After a linear change of variables P is of the form
P (ξ) = Pξ′ (ξn ) = ξnm + Pm−1 (ξ ′ )ξnm−1 + . . . + P0 (ξ ′ )
= (ξn − α1 (ξ ′ )) . . . (ξn − αm (ξ ′ )).
Here ξ = (ξ1 , . . . , ξn ) = (ξ ′ , ξn ) and αi (ξ ′ ) are the zeros of Pξ′ (ξn ). We can choose
ϕ(ξ ′ ) ∈ R such that |ϕ(ξ ′ )| ≤ m + 1 and |ϕ(ξ ′ ) − αi (ξ ′ )| ≥ |ϕ(ξ ′ ) − Im αi (ξ ′ )| ≥ 1
for i = 1, 2, . . . , m. Define ⟨E, φ⟩, when φ ∈ D, by
Z Z
−n ′
⟨E, φ⟩ = (2π) dξ P (ζ)−1 φ(−ζ)dζ
b n.
Rn−1 Im ζn =ϕ(ξ ′ )
67
68 14. EXISTENCE OF FUNDAMENTAL SOLUTIONS

By the Paley-Wiener theorem, φ(ζ)


b is an entire function and
C X
|φ(ζ)| ≤ N
∥Dα φ∥∞ .
(1 + |ζ|)
b
|α|≤N

Furthermore |P (ζ)−1 | ≤ 1, and hence, if N is large enough, we get


X
|⟨E, φ⟩| ≤ C ∥Dα φ∥∞ .
|α|≤N

Thus E ∈ D . Finally, we see that


⟨P (D)E, φ⟩ = ⟨E, P (−D)φ⟩


Z Z
−n ′
= (2π) dξ P (ζ)−1 (P (−D)φ)∧ (−ζ)dζn
Rn−1 Im ζ =ϕ(ξ )′
Z Z n
= (2π)−n dξ ′ φ(−ζ)dζ
b n
Rn−1 Im ζn =ϕ(ξ ′ )
Z Z
−n ′
= the Cauchy theorem = (2π) dξ φ(−ξ)dξ
b n
Rn−1 R
Z
= (2π)−n φ(ξ)dξ
b = φ(0) = ⟨δ, φ⟩.
Rn

Exercise 14.1. Determine a fundamental solution to the Schrödinger equation
Xn
(Dt − Dx2i )E = δ.
1
(D = −i∂)
Hint. See Exercise 10.14 and the hint to Excercise 12.7
CHAPTER 15

Fundamental solutions of elliptic differential operators

Let P (D) be a differential operator with constant coefficients. We write the


polynomial P as
P = Pm + Pm−1 + . . . + P0 ,
where Pk is a homogeneous polynomial of degree k. The operator P (D) is called
elliptic if Pm (ξ) ̸= 0 for ξ ̸= 0, ξ ∈ Rn .
Exampel 15.1. ∆ and ∂¯ are elliptic. The heat and wave operators are not
elliptic. 2
Theorem 15.2. Let P (D) be an elliptic differential operator. Then there is a
distribution E ∈ S ′ (Rn ) such that sing supp E = {0} and P (D)E = δ − ω, for
some ω ∈ S (Rn ).
E is called a parametrix to P (D).
Corollary 15.3. If P is elliptic, then P is hypoelliptic.
Proof of the corollary. We shall show that u is C ∞ if P (D)u is. If u has
compact support, we have u = δ ∗ u = (P (D)E + ω) ∗ u = E ∗ P (D)u + ω ∗ u ∈ C ∞ .
The general case follows by considering ψn u where ψn ∈ C0∞ with ψn = 1 on
{|x| ≤ n} (compare Theorem 9.6) □
Proof of the theorem. Since P is elliptic, |Pm (ξ)| ≥ δ > 0 when |ξ| = 1.
By homogenity this implies
|Pm (ξ)| ≥ δ|ξ|m .
Hence, if |ξ| > R where R is large enough,
|P (ξ)| ≥ c|ξ|m .
Take χ ∈ C0∞ (Rn ) with χ(ξ) = 1 if |ξ| ≤ R. Then (1 − χ)P −1 is bounded and
hence a tempered distribution. Thus we can define E ∈ S ′ (Rn ) by

b = 1 − χ.
E
P
Then,
b = P 1 − χ = 1 − χ = δb − χ.
(P (D)E)∧ = P E
P
69
70 15. FUNDAMENTAL SOLUTIONS OF ELLIPTIC DIFFERENTIAL OPERATORS

b = χ, then ω ∈ D
If we define ω by ω b ⊂ S and P (D)E = δ − ω. It remains to
∞ n
show that E ∈ C (R \ {0}). Observe that
1 − χ(ξ)
(xβ Dα E)∧ (ξ) = cDβ (ξ α ) = O(|ξ|−|β|−m+|α| ), |ξ| → ∞.
P (ξ)
If we choose |β| large enough, we get (xβ Dα E)∧ ∈ L1 . Thus xβ Dα E ∈ C and
hence Dα E ∈ C(Rn \ {0}), and the proof is complete. □
CHAPTER 16

Fourier series

Let u be a distribution that is periodic with period 2π in each variable, i.e.


⟨u, τ2πk φ⟩ = ⟨u, φ⟩,
if k ∈ Zn . Intuitively u is determined by its ”values” on
T = {x; 0 ≤ xi < 2π}.
Lemma 16.1. If u is periodic, then u ∈ S ′ .
Proof. Let ψ ∈ C0∞ with 0 ≤ ψ ≤ 1 and ψ = 1 on T . Put
X
ψ(x)
e = ψ(x − 2πk).
k∈Zn

Then ψe is a 2π-periodic C ∞ -function with ψe ≥ 1. Thus ϕ = ψ/ψe ∈ C0∞ and


X
ϕ(x − 2πk) = 1.
k
If φ ∈ D, then
X
⟨u, φ⟩ = ⟨ux , ϕ(x − 2πk)φ(x)⟩ = a finite sum =
k
X
= ⟨ux , ϕ(x − 2πk)φ(x)⟩ = periodicity =
k
X X
= ⟨ux , ϕ(x)φ(x + 2πk)⟩ = ⟨ux , ϕ(x) φ(x + 2πk)⟩.
k k

But if φj → 0 in S , then ϕ(x) k φj (x + 2πk) → 0 i D. (Prove that!) Hence the


P
right hand side defines an extension of u to S ′ . □
To compute u
b, we first show the following result.
Theorem 16.2 (The Poisson summation formula). If φ ∈ S , then
X X
φ(2πk)
b = φ(k).
k∈Zn k∈Zn
P
Proof. Let u = k∈Zn δ2πk . Then δ2πl ∗ u = u, since δ2πl ∗ δ2πk = δ2π(k+l) .
(Prove that!)
Hence
(e−2πilξ − 1)b
u = 0.
71
72 16. FOURIER SERIES

But e−2πilξ − 1 ̸= 0 if ξ ∈
/ Zn , and consequently ub is supported on Zn . By choosing
different l, we see that close to the origin we have ξi u b = 0, i = 1, 2, . . . , n. Thus
b = cδ0 there. Furthermore e−ikx u = u, and hence u
u b is invariant under translation
by integers. From this, we obtain
X
ub=c δk .
k∈ZN
This means that X X
φ(2πk)
b =c φ(k).
k∈Zn k∈Zn
If we replace φ with a translation of φ, we get
X X
2πikx
φ(2πk)e
b =c φ(k + x).
k∈Zn k∈Zn

Integration over {x; 0 ≤ xi < 1} gives


Z
φ(0)
b =c φ(x)dx = cφ(0)
b .
Rn
Thus c = 1 and the proof is complete. □
Let us return to the computation of ub when u is periodic. Using the Poisson
−ixy
summation formula on φ(y) = ψ(y)e , we get, as φ(ξ)
b = ψ(xb + ξ),
X X X X
ψ(x
b + 2πk) = φ(2πk)
b = φ(k) = e−ixk ψ(k).
k k k k
From the proof of Lemma 1, we have
X
⟨b
u, ψ⟩ = ⟨u, ψ⟩
b = ⟨u, ϕ(x) ψ(x
b + k)⟩
k
X
−ixk
= ⟨u, ϕ(x) e ψ(k)⟩
k
X
= ψ(k)⟨u, ϕ(x)e−ixk ⟩.
k
P
Hence u
b= k ck δk ,where
ck = ⟨u, ϕ(x)e−ixk ⟩.
In particular, if u is an integrable function on T , we have
Z
−ixk
ck = ⟨u, ϕ(x)e ⟩= u(x)ϕ(x)e−ixk dx
Rn
XZ
= u(x − 2πj)ϕ(x − 2πj)e−i(x−2πj)k dx
j T
Z X Z
−ixk
= u(x)e ϕ(x − 2πj)dx = u(x)e−ixk dx.
T j T
16. FOURIER SERIES 73

Hence ck are ”our old” Fourier coefficients. The inversion theorem implies that
1 X ikx
u(x) = ck e in S ′ .
(2π)n k
If u ∈ C l , then ck = O(|k|−l ), |k| → ∞, and the sum is uniformly convergent if
l > n. Thus we have proved
Theorem 16.3. If u ∈ C l (Rn ), l > n, and u is periodic with period 2π in each
variable, then
1 X ixk
u(x) = ck e ,
(2π)n k
where the series is uniformly convergent.
We finish this chapter by proving
Theorem 16.4 (The Plancherel theorem). If u ∈ L2 (T ) with Fourier coeffi-
cients ck , then
1 X ixk
u(x) = n
ck e in L2 ,
(2π)
and Z
1 X
|u|2 dx = n
|ck |2 .
T (2π)
1
|ck |2 < ∞, then u(x) = (2π) ixk
is a function in L2 (T )
P P
Conversely, if n k ck e
with Fourier coefficients ck .
Proof. If u ∈ C n+1 , the series is uniformly convergent, and we get
Z Z
2 1 X ix(k−l) 1 X
|u| dx = c k c̄ l e dx = |ck |2 .
T (2π)2n k,l T (2π)n

As C n+1 is dense in L2 , we can extend this to u ∈ L2 : Take un ∈ C n+1 , un → u in


L2 . Then, also un → u in S ′ and u bn → ub in S ′ . But also, by the isometry, u bn is
2 2
a Cauchy sequence in l . This implies u bn → ub in l . Hence
Z Z
1 X 1 X
|u|2 dx = lim |un |2 dx = lim n
|c (u
k n )|2
= n
|ck |2 .
T n→∞ T n→∞ (2π) k
(2π)
2
P
Conversely, if |ck | < ∞, let
1 X ixk 1 X
u(x) = n
c k e and uN (x) = n
ck eixk .
(2π) k (2π)
|k|≤N

Then, uN → u in L2 and S ′ , and we get


X
u
b = lim u
bN = c k δk .
N →∞

74 16. FOURIER SERIES

Remark 16.5. If u is a function with period t, then ut (x) = u( 2πxt


) has period
2π. Using this, we can generalise Fourier series to functions with arbitrary period.
2
Exercise 16.1. H 7.2.1
Exercise 16.2. H 7.2.5
Exercise 16.3. H 7.2.8
P∞ 1
P+∞ 1
P∞ (−1)n
Exercise 16.4. Compute a) −∞ 1+n2 b) −∞ (n+a)2 and c) n=0 (2n+1)3 .
CHAPTER 17

Some applications

17.1. The central limit theorem


Theorem 17.1 (The central limit theorem). Let X, X1 , X2 , . . . be independent
identically distributed stochastic variables with E[X] = m and Var[X] = σ 2 . Then
 X + X + . . . + X − nm Z x
1 2 n
 1 1 2
lim P √ ≤x = √ e− 2 y dy. (17.1)
n→∞ σ n 2π −∞
Some background: To a stochastic variable X we associate a probability mea-
sure µ on R (we write X ∼ µ) by
Z x
P (X ≤ x) = dµ(y).
−∞
If µ1 are µ2 probability measures, we define a new probability measure µ1 ∗ µ2 by
ZZ
⟨µ1 ∗ µ2 , φ⟩ = φ(x + y)dµ1 (x)dµ2 (y).
R2

Then (µ1 ∗ µ2 ) = µ b2 . (Show that!) If X ∼ µ1 and Y ∼ µ2 are independent,
b1 µ
then X + Y ∼ µ1 ∗ µ2 .
Proof. We may assume that m = 0 and σ = 1. Let
X1 + . . . + X n
Sn = √
n
and
µn∗ = µ ∗ . . . ∗ µ .
| {z }
n times
Then Sn ∼ µn , where
Z  
x
⟨µn , φ⟩ = φ √ dµn∗ (x),
R n
and   n
ξ
µ b √
cn (ξ) = µ .
n
Since Var[X] < ∞, Z
µ
b(ξ) = e−ixξ dµ(x)
R
75
76 17. SOME APPLICATIONS

is a C 2 -function with
b ′ (0) = −im = 0 and µ
µ b ′′ (0) = −σ 2 = −1.
Thus
1
b(ξ) = 1 − ξ 2 + o(ξ 2 ),
µ ξ → 0,
2
and
  n   2
ξ 1 2 ξ n 1 2
µ b √
cn (ξ) = µ = 1− ξ +o → e− 2 ξ , n → ∞,
n 2n n
for each fixed ξ. But, since |b
µ(ξ)| ≤ 1, we get, by dominated convergence, that
1 2
cn (ξ) → e− 2 ξ
µ in S ′ .
Hence Fourier inversion implies that
1 1 2
µn → √ e − 2 x

in S , and hence also in D . But µn are positive measures, and by Theorem 7.0.5
′ ′

1 1 2
µn → √ e − 2 x

weakly, and hence we obtain (1). □

17.2. The mean value property for harmonic functions


Theorem 17.2. If u ∈ C ∞ is harmonic in a neighborhood of {|x| ≤ 1}, then
Z
1
u(0) = u(y)dσy).
ωn S n−1
Remark 17.3. By Weyl’s lemma, the assumption that u ∈ C ∞ is unnecessary.
2
Proof. Define a distribution Λ by
Z
⟨Λ, φ⟩ = φ(y)dσ(y) − ωn φ(0).
S n−1

Then Λ ∈ E ′ , and hence Λ b is an entire function. Furthermore, Λ, and therefore


also Λ, is radial. Hence Λ(ζ) = G(|ζ|), where G(t) = Λ(t,
b b b 0, . . . , 0) is holomorphic.
Also, G is even, so G(z) = F (z 2 ) for some entire function F . Since F (0) = Λ(0)
b =0

and F (0) = 0,
Λ(ξ)
b F (|ξ|2 )
=
|ξ|2 |ξ|2
is the restriction of an entire function. By the Paley-Wiener theorem, there is a
distribution µ ∈ E ′ with µ
b(ξ) = −F (|ξ|2 )/|ξ|2 , and so
(∆µ)b(ξ) = −|ξ|2 µ b(ξ) = Λ(ξ).
b
17.3. THE HEISENBERG UNCERTAINTY PRINCIPLE 77

Hence ∆µ = Λ, which gives


⟨Λ, u⟩ = ⟨∆µ, u⟩ = ⟨µ, ∆u⟩ = ⟨µ, 0⟩ = 0.

17.3. The Heisenberg uncertainty principle


Theorem 17.4. If f ∈ L2 (R), then
r
π
∥xf (x)∥2 ∥ξ fb(ξ)∥2 ≥ ∥f ∥22 , (17.2)
2
with equality only if f (x) = exp(−kx2 ), k > 0.
Quantum mechanical background: The state of a particle is described by a
function ψ ∈ L2 (R) with ∥ψ∥2 = 1. We interprete
Z
|ψ|2
E
as the probability that the particle is in the set E. An observable quantity A is
a symmetric operator on a suitable subspace of L2 . The mean value of A in the
state ψ is Z
E[A] = Aψ · ψ̄ = ⟨Aψ, ψ⟩.
That A is symmetric means that A = A∗ and hence we have
⟨Aψ, ψ⟩ = ⟨ψ, A∗ ψ⟩ = ⟨ψ, Aψ⟩ = ⟨Aψ, ψ⟩.
Thus the mean value is real.
Exampel 17.5.
a) Position. Aψ(x) = xψ(x)
b) Momentum. Bψ = 2πiψ ′ . 2

We have
Z Z Z Z
′ b 2 dξ.
E[B] = Bψ · ψ̄ = 2πi ψ ψ̄ = Plancherel = ξ ψ(ξ)ψ(ξ) = ξ|ψ(ξ)|
b

b 2 as the density of the momentum.


Hence we can interprete |ψ(ξ)|
The general form of the Heisenberg uncertainty principle is
1 2
E[(A − E[A])2 ] E[(B − E[B])2 ] ≥ E[AB − BA] (17.3)
4
for arbitrary A and B.
Exercise 17.1. Show that if A and B are position and momentum, then AB − BA = −2πi.
Exercise 17.2. Prove that (1) implies (2), when A and B are position and momentum.
78 17. SOME APPLICATIONS

Proof. If f ∈ S , then
∥xf (x)∥2 ∥ξ fb(ξ)∥2 = ∥xf (x)∥2 ∥fb′ (ξ)∥2 = Parseval =
√ √ Z
= 2π∥xf (x)∥2 ∥f (x)∥2 ≥ Schwartz ≥ 2π |xf (x)f ′ (x)|dx

√ Z 1 ′


≥ (|xz̄w| ≥ x Re z̄w) ≥ 2π x f (x)f (x) + f (x)f (x) dx
2
r Z
π
= x(|f (x)|2 )′ dx = Integration by parts
2
r Z r
π 2 π
= |f | dx = ∥f ∥22 .
2 2
The proof that the theorem holds for functions in L2 , and the statement of equality
is left to the reader. □

17.4. A primer on Sobolev inequalities


A benefit of the theory of distributions is that we can find solutions to problems
that has no classical solutions. But we often want our solutions to be nice functions.
Therefore it is natural to ask the question
When is a distributional solution a function?
The theory of Sobolev spaces gives us a method to answer that question.
We start with the simplest result in Sobolev theory,
Theorem 17.6 (The Sobolev L1 -inequality). Let f be an integrable function
on Rn . Assume that the distributional derivatives ∂ α f also are integrable for all
|α| ≤ n. Then f is a bounded continuous function and
X
∥f ∥∞ ≤ ∥∂ α f ∥1 . (17.4)
|α|≤n

If furthermore ∂ α f are integrable for all |α| ≤ n + k, then f is a C k -function.


Proof. We start with the case n = 1 where we shall show that
∥f ∥∞ ≤ ∥f ∥1 + ∥f ′ ∥1 . (17.5)
If φ ∈ C0∞ , then
Z x Z x Z ∞
′ ′
|φ(x)| = φ (t)dt ≤ |φ (t)|dt ≤ |φ′ (t)|dt .
−∞ −∞ −∞
This implies that
∥φ∥∞ ≤ ∥φ′ ∥1 . (17.6)
This inequality is sharper than (5), but we have obtained it asssuming two strong
extra conditions, C ∞ and compact support. (The function 1 shows that (6) can
not be true in general.)
17.4. A PRIMER ON SOBOLEV INEQUALITIES 79

If f ∈ C ∞ is not compactly supported, we choose a sequence of cut off functions


χn ∈ C ∞ , with χn = 1 as |x| ≤ n and ∥χ′n ∥∞ ≤ 1. If we apply (6) to φ = χn f ,we
get
∥χn f ∥∞ ≤ ∥(χn f )′ ∥1 ≤ ∥χ′n f ∥1 + ∥χn f ′ ∥1 ≤ ∥f ∥1 + ∥f ′ ∥1 .
Since n is arbitrary, (5) is proved for C ∞ -functions.
If f is not C ∞ , we let ϕδ be an approximative identity. Then fδ = ϕδ ∗ f ∈ C ∞
and we can apply (5) to fδ . We get, as ∥ϕδ ∗ f ∥1 ≤ ∥f ∥1 och ∥(ϕδ ∗ f )′ ∥1 =
∥ϕδ ∗ f ′ ∥1 ≤ ∥f ′ ∥1 , that
∥ϕδ ∗ f ∥∞ ≤ ∥f ∥1 + ∥f ′ ∥1 .
But ϕδ ∗ f → f a.e. and we have proved (5) in the general case.
To finish the proof, we apply (5) to f − fδ , to obtain
∥f − fδ ∥∞ ≤ ∥f − ϕδ ∗ f ∥1 + ∥f ′ − ϕδ ∗ f ′ ∥1 → 0, δ → 0 .
Thus fδ → f uniformly, and f is a continuous function.
The last claim follows by applying this argument to the functions ∂ i f , i =
1, 2, . . . , k.
The argument when n ≥ 2 is similar. We only give the details when n = 2.
The case n = 2 is general enough to show how to extend from the case n = 1, but
without too cumbersome notation. So let n = 2 and φ ∈ C0∞ . We now get
Z xZ y Z ∞Z ∞
(1,1)
|φ(x, y)| = ∂ φ(s, t)dsdt ≤ |∂ (1,1) φ(s, t)|dsdt
∞ ∞ ∞ ∞

and hence
∥φ∥∞ ≤ ∥∂ (1,1) φ∥1 .
When we apply this to χn f , f ∈ C ∞ , we get, as ∂ (1,1) (χn f ) = ∂ (1,1) χn f +
∂ (1,0) χn ∂ 0,1 f + ∂ (0,1) χn ∂ 1,0 f + χn ∂ (1,1) f , that
∥f ∥∞ ≤ ∥f ∥1 + ∥∂ 1,0 f ∥1 + ∥∂ 0,1 f ∥1 + ∥∂ (1,1) f ∥1 , f ∈ C ∞ .
The rest of the argument works exactly the same as in the case n = 1.

Remark 17.7. The proof shows that it is enough to consider α = (α1 , . . . , αn ),
where each αi is either 0 or 1, in the sum (1). 2
Theorem 17.8 (The Sobolev L2 -inequality). Let f ∈ D ′ (Ω), where Ω is an
open set in Rn , and let r and k ≥ 0 be integers. If ∂ α f ∈ L2loc for all α, 0 ≤ |α| ≤ r
where r > k + n2 , then f ∈ C k (Ω).
Proof when n = 1 and k = 0.
The assumption means that f ∈ L2loc and f ′ ∈ L2loc . Let ω be an open set,
ω ⊂⊂ Ω, and take χ ∈ C0∞ (Ω) with ∥χ∥∞ ≤ 1, ∥χ′ ∥∞ ≤ C and χ = 1 in a
neighborhood of ω. Define F (x) = Fω (x) = χ(x)f (x). (F = 0 outside the support
80 17. SOME APPLICATIONS

of χ.) Since F ∈ L2 (R) and F ′ = χ′ f + χf ′ ∈ L2 (R), the Parseval identity implies


that Z Z
2
|F | dξ < ∞ and
b ξ 2 |Fb|2 dξ < ∞ .
R R
Hence Z
(1 + |ξ|)2 |Fb|2 dξ < ∞ .
R
The Cauchy-Schwartz inequality implies
Z 2 Z 2

|Fb|dξ = (1 + |ξ|)|Fb|
R R 1 + |ξ|
Z Z

≤ (1 + |ξ|)2 |Fb|2 dξ 2
<∞.
R R (1 + |ξ|)

Thus Fb is integrable and hence F is continuous. Since f = F on ω, f is continuous


on ω. As ω is an arbitrary open subset of Ω, it follows that f ∈ C(Ω). □
The general case.
If Zn = 1 and k = 1, we also know that F ′′ ∈ L2 . Thus ξ 2 Fb(ξ) ∈ L2 and we
have (1 + |ξ|)4 |Fb|2 dξ < ∞. Using the Cauchy-Schwartz inequality as in the case
R Z
k = 0, this gives (1 + |ξ|)|Fb|dξ < ∞. In particular, ξ Fb(ξ) is integrable and F ′
R
is continuous. The case for arbitrary k follows in the same way.
If n > 1, the condition on ∂ α f implies that ξil Fb(ξ) ∈ L2 , l ≤ r. Using the
inequality (1+|ξ|)2l ≤ Cl (1+ξ12l +. . .+ξn2l ), and the Cauchy-Schwartz
Z inequality as

above, we obtain (1+|ξ|)k Fb ∈ L1 and hence F ∈ C k . Note that s
<∞
Rn (1 + |ξ|)
if s > n.

17.4.1. Sobolev spaces. Let us abstract the ideas in the proof of the L2 -
inequality. We saw that if f and its derivatives up to order r are in L2 , then
(1 + |ξ|)r fb ∈ L2 or equivalently (1 + |ξ|2 )r/2 fb ∈ L2 . In this condition, r may be an
arbitary real number, and we can make the following definition.
Definition 17.9. A distribution f ∈ D ′ (Rn ), is in the Sobolev space H s (Rn ),
s ∈ R, if
∥f ∥H s = ∥(1 + |ξ|2 )s/2 fb(ξ)∥L2 < +∞
Proposition 17.10. If f ∈ H r (Rn ), where r > s + n2 , then
(1 + |ξ|)s fb ∈ L1 (Rn ).
Remark 17.11. If s = k is a non-negative integer, this implies that f ∈ C k .
17.5. MINKOWSKI’S THEOREM 81

Proof. We have (1 + |ξ|2 )r/2 fb(ξ) ∈ L2 . Hence, by the Cauchy-Schwartz in-


equality, we get
Z
(1 + |ξ|2 )s/2 fd
(ξ)dξ
Rn
Z

= (1 + |ξ|2 )r/2 fb(ξ) ≤ C∥f ∥H r ,
Rn (1 + |ξ|2 )(r−s)/2
as (1 + |ξ|2 )−(s−r) is integrable when r − s > n2 . □

17.5. Minkowski’s theorem


Let B be a convex set in Rn that is symmetric at the origin. If |B| ≥ 2n , then
B contains more than one lattice point.
Proof. We assume that 0 is the only lattice point in B, and show that this
implies that |B| < 2n . Let f = χB ∗ χB . Since B is symmetric, χ
bB is real. Hence
2 2
χB ) = |b
f = (b
b χB | .
We observe that if f (2k) ̸= 0, ie.
Z
f (2k) = χB (2k − x)χB (x)dx ̸= 0 ,

then there is x ∈ B with 2k − x ∈ B. But then we have k = 12 (2k − x) + 12 x ∈ B,


as B is convex. Hence if f (2k) ̸= 0 we have k = 0. Furthermore
Z Z
f (0) = χB (−x)χB (x)dx = |χB |2 dx = |B|.

The Poisson summation formula, applied to the lattices (2Z)n and (πZ)n , gives
X X
f (2j) = 2−n fb(πj) .
j∈Zn j∈Zn

Hence
X X
|B| = f (0) = f (2j) = 2−n fb(πj)
j j
!
X X
= 2−n χB (πj)|2 = 2−n |B|2 +
|b χB (πj)|2
|b .
j j̸=0

If we can show that X


χB (πj)|2 > 0,
|b
j̸=0
−n
we obtain |B| > 2 |B| , or |B| < 2n , and we are done.
2

bB (πj) = 0 when j ̸= 0, then


But if χ
X
χ(x) = χB (x + 2j)
j
82 17. SOME APPLICATIONS

is constant. This follows from the Poisson summation formula since


X X
χ(x) = τ−x χB (2j) = 2−n bB (πj) = 2−n χ
eiπxj χ bB (0).
j j

But this is a contradiction as


χ(0) = 1 ̸= 0 = χ(1, 0, . . . , 0).

Exercise 17.3. The proof is ”wrong”. Why? Correct it!
Index

C0∞ , 7 hypoelliptic, 47, 69


L2 , 57
D ′ (Ω), 11 inversion theorem, 52
D(Ω), 11 Laplace operator, 29, 62
E (Ω), 33

mean value property, 76
S , 51 measure, 11
S ′ , 51 Minkowski’s theorem, 81
approximate identities, 8 odd, 61
Cauchy sequence, 35 Paley-Wiener theorem, 65, 66
Central limit theorem, 75 parametrix, 69
compact support, 33 Parseval formula, 53, 57
complete, 35 partition of unity, 13
convergence, 35, 51 periodic, 71
convolution, 9, 39 Plancherel theorem, 53, 73
Poisson summations formula, 71
derivative, 17 positive distribution, 12
Dirac measure, 17 principal value, 24
distribution, 11
division problem, 24 rapidly decreasing functions, 51
regularisation, 40
elliptic, 69 regularization, 9
even, 61
Schwartz space, 51
finite part, 21 singular support, 46
Fourier coefficient, 49, 73 Sobolev inequalities, 78
Fourier series, 50 Sobolev spaces, 80
Fourier transform, 49, 50 support of a distribution, 13
fundamental solution, 29, 45, 67
tempered distribution, 51
harmonic function, 41 translation, 42
harmonic functions, 76
heat eqution, 30 Weyl’s lemma, 41
Heaviside function, 17
Heisenberg uncertainty principle, 77
Hilbert transform, 61
homogeneous, 23, 62
83

You might also like