8 Fourier Series

Our aim is to show that under reasonable assumptions a given 2-periodic

function f can be represented as convergent series
f(x) =
a
0
2
+

n=1
(a
n
cos nx +b
n
sin nx) . (8.1)
By denition, the convergence of the series means that the sequence (s
n
(x))
of partial sums, dened by
s
n
(x) =
a
0
2
+
n

k=1
(a
k
cos kx +b
k
sinkx) ,
converges at a given point x to f(x), s
n
(x) f(x).
We start with some denitions. A real valued function dened on R is
said to be periodic with period L, or L-periodic, is f(x + L) = f(x) for
all x. In this case f is completely determined by its values on any interval
[a, a+L). A function f dened on [a, a+L) can be extended in a unique way
to a periodic function on R. Indeed, given x, there exists a unique integer
n so that x belongs to [a + nL, a + (n + 1)L) so that x nL [a, a + P)
and we set f(x) = f(x nL). A function f on [a, b] is piecewise continuous
if f is continuous except a nite numbers of points and at each such point
the one-sided limits f(x
+
) = lim
0
f(x + ) and f(x

) = lim
0
f(x +
) exist and are nite. We say that f is piecewise smooth (or piecewise
dierentiable) on [a, b] if f and f

are piecewise continuous on [a, b]. If f

is piecewise continuous, then it is Riemann integrable over any bounded
interval contained in its domain. In addition, if f is L-periodic, then
_
a+L
a
f(x) dx =
_
L
0
fx) dx
for every a. In what follows we focus on functions which are 2-periodic.
Assume
f(x) =
a
0
2
+

n=1
(a
n
cos nx +b
n
sin nx) .
We also assume that the convergence is well-behaved so that we can integrate
term-by-term. We want to compute the coecients a
k
and b
k
. We use
66
integral identities
_

cos mxcos nx dx =
_

_
0, m = n
2, m = n = 0
, m = n = 0
_

sin mxsinnx dx =
_
0, m = n
, m = n = 0
_

cos mxsinnx dx = 0.
To compute a
0
, we multiply both sides of (8.1) by the constant function 1
and integrate over [, ] and get
_

f(x) dx =
a
0
2
_

1 dx +

n=1
_
a
n
_

cos nx dx +b
n
_

sin nx dx
_
=
a
0
2
(2) = a
0
To compute a
m
with m 1 we we multiply both sides of (8.1) by cos mx
and integrate over [, ] to get
_

f(x) cos mx dx =
a
0
2
_

cos mx dx
+

n=1
_
a
n
_

cos nxcos mx dx +b
n
_

sin nxcos mx dx
_
= a
m
_

cos mxcos mx dx = a
m
.
Similarly, multiplying both sides of (8.1) by sinmx and integrating over
[, ], we nd that
_

f(x) sin mx dx = b
m
.
Summing up,
a
n
=
1

f(x) cos nx dx, n 0

b
n
=
1

f(x) sin nx dx, n 1.

(8.2)
67
The coecients a
n
and b
n
are called the Fourier coecients of f and the
series
a
0
2
+

n=1
[a
n
cos nx +b
n
sin nx]
the Fourier series of f. We denote this fact by
f(x)
a
0
2
+

n=1
[a
n
cos nx +b
n
sin nx] .
The symbol should be read as f has Fourier serier.
Example 8.1. Let f be a 2-periodic function given by f(x) = x for x (, ].
Then integrating by parts
a
n
=
1

xcos nx dx =
1

_
xsin nx
n
_

1
n
_

sin nx dx = 0
and
b
n
=
1

xsin nx dx =
1

xcos nx
n
_

+
1
n
_

cos nx dx
=
2 cos n
n
=
2(1)
n+1
n
.
Consequently,
f(x) 2

n1
(1)
n+1
n
sin nx.
The series converges for every x in view of the following Dirichlets test. If (a
n
) is
a decreasing sequence converging to 0 and the sequence (s
n
) of partial sums of the
series

b
n
is bounded, then the series

a
n
b
n
converges. In our case, a
n
=
1
n
is
decreasing and converges to 0, (1)
n
sin nx = sin n(x +), and
sin

2

n

k=1
sin k = sin
(n + 1)
2
sin
n
2
which implies that

k=1
sin k

sin

2

if = 2k for all k Z.
Example 8.2. Consider f be a 2-periodic function dened by f(x) = |x| for
x [, ]. Note that f is even. Since the product of even function with the odd
function is odd, it follows that
_

f(x) sin mxdx = 0.

68
Hence b
n
= 0 for all n 1. To compute b
n
note that the product of an even
function with an even function is even so that
a
n
=
1

f(x) cos nx dx =
2

_

0
f(x) cos nxdx =
2

_

0
xcos nx dx.
If n = 0, then
a
0
=
2

_

0
x dx = ,
and if n 1, then integrating by parts, one nds that
2

_

0
xcos nx dx =
2

_
xsin nx
n
_

2
n
_

0
sin nx dx
=
2
n
2

_
cos nx
_

0
=
2
n
2

[(1)
n
1] .
Hence a
n
= 0 if n is even and a
n
=
4
n
2

when n is odd, and hence

f(x)

2

4

n1
cos(2n 1)x
(2n 1)
2
.
Since the series

1
n
2
converges, the M-Weierstarss test implies that the above
series converges.
In the above examples we have used the following identities,
_
L
L
g(x) dx =
_
2
_
L
0
g(x) dx if g is even
0 if g is odd.
Using that the sum of two even functions is even and the sum of two odd
functions is odd, that the product of two even functions or two odd functions
is even, and the product of an even function and an odd function is odd, we
have the following proposition.
Proposition 8.3. (i) If f is even, then its Fourier sine coecients b
n
are equal to 0 and f is represented by Fourier cosine series
f(x)
a
0
2
+

n=1
a
n
cos nx
where
a
n
=
2

f(x) cos nx dx for n 0.

69
(ii) If f is odd, then its Fourier cosine coecients a
n
are equal to 0 and
f is represented by Fourier sine series
f(x)

n=1
a
n
sin nx
where
b
n
=
2

f(x) sin nx dx for n 1.

8.0.1 Convergence Theorem
Theorem 8.4. Assume that f is a 2-periodic function which is piecewise
smooth. Then
1
2
_
f(x

) +f(x
+
)
_
=
a
0
2
+
n

k=1
[a
k
cos kx +b
k
sin kx]
where f(x

) = lim
yx
f(y). In particular, if f is continuous at x, then
f(x) =
a
0
2
+
n

k=1
[a
k
cos kx +b
k
sin kx] .
Example 8.5. Consider the function f from Example 8.1. Then f is smooth at
all points except points m where m is odd. The one sided limits at these points
are
f(m

) = lim
xm

f(x) = and f(m

+
) = lim
xm
+
f(x) =
provided that m is odd. Consequently, the Fourier series converges of f converges
to f at every point except points m with m odd. At these points the Fourier series
converges to
1
2
[f(m

) +f(m
+
)] = 0. In particular,
x
2
=

n=1
(1)
n+1
n
sin nx, for x (, ).
Taking x = /2, we nd that

4
=

n=1
(1)
n1
2n 1
.
70
Example 8.6. Consider the function 2-periodic function f given by f(x) = |x|
for x (, ). (See Example 8.2.) Then f is continuous at every point and it is
smooth except points m with m odd. Hence the Fourier series of f converges to
f at every point. In particular,
|x| =

2

4

n=1
cos(2n 1)x
(2n 1)
2
for x [, ].
Substituting x = 0, we nd that

2
8
=

n=0
1
(2n + 1)
2
.
Set S =

n=1
1
n
2
. Then we note that
S
4
=
1
4

n=1
1
n
2
=

n=1
1
(2n)
2
and hence
3
4
S = S
S
4
=

n=1
1
n
2

n=1
1
(2n)
2
=

n=0
1
(2n + 1)
2
=

2
8
.
From this we conclude that
S =

n=1
1
n
2
=

2
6
.
Integration of Fourier series
We start with the following observation. Assume that f is 2-periodic,
piecewise continuous, and let F(x) =
_
x
0
f(y) dy. Then F is 2 periodic if
and only if
_

f(y)dy = 0. Indeed, we have

F(x + 2) F(x) =
_
x+2
x
f(y)dy =
_

f(y)dy.
This means that the constant term in the Fourier series of f is equal to 0.
Since
a
0
2
=
1
2
_

f(y) dy, the number a

0
/2 is the mean of the function f
over the interval [.].
Theorem 8.7. Assume that f is 2-periodic and piecewise continuous and
its mean is equal to 0. Then its Fourier series
f(x)

n1
[a
n
cos nx +b
n
sin nx]
71
can be integrated term by term and produce the Fourier series
F(x) =
_
x
0
f(y) dy C
0
+

n1
_

b
n
n
cos nx +
a
n
n
sin nx
_
where the constant C
0
=
1
2
_

F(y) dy.
Example 8.8. Consider a 2-periodic function f given by f(x) = x on (, ].
(See Example 8.1.) Since f is odd, its mean value over [, ] is equal to 0. Its
Fourier series is given by
x = 2

n1
(1)
n1
n
sin nx, for x (, ].
Integrating term by term we nd that
x
2
2
=

2
6
2

n1
(1)
n1
n
2
cos nx for x (, ]
where
2
/6 =
1
2
_

(x
2
/2) dx.
Dierentiation of Fourier series
Proposition 8.9. Assume that f is 2-periodic, continuous, and piecewise
smooth. Abbreviate by a
n
, b
n
the Fourier coecients of f and by a

n
, b

n
the
Fourier coecients of f

. Then
a

n
= nb
n
and b

n
= na
n
.
Proof. Integrating by parts,
a

n
=
1

(x) cos nx dx =
1

_
f(x) cos nx
_

+
n

f(x) sin nx dx = nb
n
.
Similarly, b

n
= na
n
.
As a consequence, we get:
Theorem 8.10. Let f be 2-periodic, continuous, and piecewise smooth.
In addition, assume that f

is piecewise smooth. If
f(x) =
a
0
2
+

n=1
[a
n
cos nx +b
n
sin nx] ,
then
1
2
_
f

(x

) +f

(x
+
)
_
=

n=1
[nb
n
cos nx na
n
sin nx] .
In particular, at points x at which f

(x) exists, the series converges to f

(x).
72
Proof. By Proposition 8.9,
f

(x)

n=1
[nb
n
cos nx na
n
sin nx] ,
and since f

is piecewise smooth, the theorem follows from Theorem 8.4.

Example 8.11. Recall the function 2-periodic function f given by f(x) = |x|
for x (, ). (See Example 8.2 and Example 8.6.) The function f is continuous
and piecewise smooth, and
|x| =

2

4

n=1
cos(2n 1)x
(2n 1)
2
, x [, ].
The derivative of f is given by
d
dx
|x| =
_
1, x > 0,
1, x < 0.
In view of Theorem 8.10,
4

n=1
sin(2n 1)x
(2n 1)
=
_
1, 0 < x < ,
1, < x < 0.
At x = 0, the series converges to
1
2
[f

(x

) +f

(x
+
)] =
1
2
[1 + 1] = 0.
8.0.2 Absolute and Uniform Convergence
Theorem 8.12. Let f be 2-periodic, continuous, and piecewise smooth.
Then its Fourier series converges absolutely and uniformly.
Proof. We prove the result under additional assumption that f

is piecewise smooth.
This means that f

is piecewise continuous. Abbreviate by a

n
, b
n
the Fourier co-
ecients of f and by a

n
, b

n
the Fourier coecients of f

. Since f is continuous,
f(x) =
a
0
2
+

n=1
[a
n
cos nx +b
n
sin nx] .
To prove the absolute convergence it suces to show that |a
n
| , |b
n
| M/n
2
for
some constant M independent of n. Then, by the comparison test,

n1
|a
n
| and

n1
|b
n
| converge which imply the absolute convergence of the Fourier series.
Also, the Weierstrass M-test implies the uniform convergence. By Proposition 8.9,
a
n
= b

n
/n. That is,
a
n
=
1

f(x) cos nx dx =
1
n
_

(x) sin nx dx.

73
By assumption, f

is piecewise continuous and so f

is continuous except a nite

number of point at which it has a jump. If (a, b) is an interval on which f

is
continuous, then
1
n
_

(x) sin nx dx =
1
n
2

_
f

(x) cos nx
_
b
a

1
n
2

_
b
a
f

(x) cos nx dx
=
1
n
2

_
f(b

) cos nx f(a
+
) cos nx
_
+
1
n
2

_
b
a
f

(x) cos nx dx.

Since |f

(x)| , |f

(x)| K for some constant K and all x (at points x where f

and
f

has a jump discontinuity, f

(x) one has to take left- and hand-side derivatives).

So,
1
n

(x) sin nx dx

K(2 + (b a)
n
2

2K(1 +
n
2

4K
n
2
.
Since f

has a nite number of discontinuities, the integral

_

(x) sin nx dx can

be written as the nite number, say m, of integrals
_
b
a
f

(x) sin nx dx over intervals

on which f

is continuous. Then
|a
n
| =
1
n

(x) sin nx dx

4Km
n
2
.
Similarly, |b
n
|
4Km
n
2
.
Theorem 8.13. Let f be 2-periodic. Assume that f is of class C
k1
and
that f
(k1)
is piecewise smooth. If the Fourier coecients satisfy
|a
n
|
C
n
k+
and |b
n
|
C
n
k+
for some > 1 and C > 0, then f is of class C
k
.
Proof. Let a
(j)
n
and b
(k)
n
are coecients of the Fourier series of f
(j)
. Then using
Proposition 8.9

a
(j)
n

= n
j
|a
n
| if j is even and

a
(j)
n

= n
k
|b
n
| if j is odd. Similarly,
for b
(k)
n
. Observe, using the assumption, that if j k, then

n1

n
j
a
n

n1
1
n
kj+
C

n1
1
n

for j k. By the Weierstrass M-test, the series

n1

n
j
a
n

converges absolutely
and uniformly for j k. Similarly, the series

n1

n
j
b
n

converges absolutely
and uniformly for j k Hence, f
(j)
=

n1
[a
j)
n
cos nx + b
j)
n
sin nx] for j k and
f
(k)
is continuous.
74
8.1 Changing a scale
So far we have considered 2-periodic functions and their Fourier series over
the interval [, ]. Now, given a function f dened on [L, L], we dene
F(y) = f
_
L

y
_
. Then F is dened on [, ] and its Fourier series is given
by
F(y)
a
0
2
+

n1
_
a
n
cos ny +b
n
sin ny

where
a
n
=
1

F(y) cos ny dy and b

n
=
1

F(y) sin ny dy. (8.3)

Since f(x) = F(

L
x) for x [L, L], we deduce that
f(x)
a
0
2
+

n1
_
a
n
cos
n
L
x +b
n
sin
n
L
x

The coecients a
n
and b
n
can be computed by integrating (8.3) by substi-
tution x =
L

y. We get
a
n
=
1

F(y) cos ny dy =
1

f
_
L

y
_
cos ny dy
=
1
L
_
L
L
f(x) cos
nx
L
dx
(8.4)
and similarly
b
n
=
1

F(y) sin ny dy =
1
L
_
L
L
f(x) cos
nx
L
dx. (8.5)
75