FOURIER SERIES AND FOURIER INTEGRALS

DIRICHLET CONDITIONS
Let f(x) be defined in the interval < < + 2 with period 2 and satisfy the following
conditions.
(i) f(x) is single valued except possibly at a finite number of points in a given period
(ii) f(x) is periodic with period 2
(iii) () and () are piecewise continuous in a given interval and
it has a finite number of maxima and minima in a given period
then the above series converges to
(a) f(x) if x is a point of continuity
f(x+0) + f(x0)
(b) if x is a point of discontinuity
2
(1) Fourier series for f(x) in < < + 2

0
f(x) = + cos + sin
2
=1 =1
c+2
c+2
1 1 1 c+2
Where a0 = f(x)dx , an = f(x) cos nx dx , bn = f(x) sin nx dx
c c
c
The values of a0 , an and bn are known as Eulers Formulae (Eulers coefficients or Fourier
coefficients)
Result: 1
If c =0, then the interval of definition for f(x) is (0, 2) and in this case,
2
1 1 2 1 2
a0 = f(x)dx , an = f(x) cos nx dx , bn = f(x) sin nx dx
0 0
0
Result: 2
If c = -, then the interval of definition for f(x) is (, ) and in this case,

1 1 1
a0 = f(x)dx , an = f(x) cos nx dx , bn = f(x) sin nx dx


Change of interval:
(2) Fourier series for f(x) of period 2l(0 < x < 2l)

0
f(x) = + cos + sin where
2
=1 =1
2
1 1 2 nx 1 2 nx
a0 = f(x)dx , an = f(x) cos dx , bn = f(x) sin dx
0 0
0
Suppose < < ,

1 1 nx 1 nx
a0 = f(x)dx , an = f(x) cos dx , bn = f(x) sin dx


RESULTS:
c+2

(1) sin nx dx = 0
c
c+2

(2) cos nx dx = 0
c
c+2
, n = 1,2,3
(3) sin2 nx dx = {
0 ,n = 0
c
c+2
n = 1,2,3,
(4) cos 2 nx dx = {
2 n=0
c
c+2
0 ,m n
(5) cos mx cos nx dx = { ,m = n
c 2 , m = n = 0
c+2
0 ,m n
(6) sin mx sin nx dx = { ,m = n
c 0 ,m = n = 0
c+2

(7) sin mx cos nx = 0

c
L

(8) dx = 2L
L
L
nx
(9) cos ( ) dx = 0
L
L
L
nx
(10) sin ( ) dx = 0
L
L
L
mx nx 0 if m n
(11) cos ( ) cos ( ) dx = {
L L L if m = n
L
L
mx nx 0 if m n
(12) sin ( ) sin ( ) dx = {
L L L if m = n
L
L
nx nx
(13) sin ( ) cos ( ) dx = 0
L L
L

PERIODIC FUNCTION:
A function f(x) is said to be periodic function iff ( + ) = ()with period
NOTE:
1. All trigonometric function are periodic function
2. sin , cos , sec , csc are periodic function with period 2
3. tan , cot with period
EVEN AND ODD FUNCTION:
(1) If () = (), then f(x) is even
(2) If () = (), then f(x) is odd
EXAMPLES:
1. cos x, sec x are even function
2. sin x , csc x , tan x and cot x are odd function
3. x 2 , sin2 x, |x|, x sin x are even function
4. x 3 , sin3 x, tan3 x are odd function

5. For odd function , () = 0

a a

6. For even function , f(x)dx = 2 f(x)dx

a 0
7. Sum of even function is even and sum of odd function is odd
8. The product of two even or two odd functions is even.
9. The product of an even and an odd function is odd.
10. If f is an even function and g is an odd function, then fg is an odd function.
11. The product fg is even function if either f and g are both even or if f and g are both odd.
12. Suppose [, ] and

a0
f(x) = + an cos nx + bn sin nx then if
2
n=1 n=1
(i) f is odd, a0 = a1 = a2 = = 0, (ii) f is even, b1 = b2 = b3 = = 0.
13. If f is any real valued function on [, ] and () = () + (),
() = () () then g is an even function and h is an odd function.
2
1 + , < <

14. The function f(x) defined as () = { 2 is an even function.
1 , 0 < <

< < 0
15. The function f(x) defined as () = {
0 < <
is neither odd nor an even function
16. sin + cos is neither an odd nor an even function.

THE COSINE SERIES (FOURIER SERIES FOR EVEN FUNCTION)

0
f(x) = + cos where
2
=1

2 2
a0 = f(x)dx , an = f(x) cos nx dx
0
0
THE SINE SERIES (FOURIER FOR ODD FUNCTION)

f(x) = sin where

=1
1
bn = f(x) sin nx dx

DEDUCTION
1 1 1 2
1. + + + ..=
12 22 3 2 6
1 1 1 2
2. + ..=
12 22 3 2 12
1 1 1 2
3. + + + ..=
12 32 5 2 8

1 1
4. 2 =
4 1 2
=1
1 1 1 1
5. + + + =
1.3 3.5 5.7 2
1 1 1 2
6. + =
1.3 3.5 5.7 4
1 1 1 1
7. + + .=
1 3 5 7 4
1 1 1 1 3
8. + + ..=
13 33 53 73 32
1 1 1 4
9. + + + ..=
14 24 3 4 90
1 1 1 4
10. + + + . . =
14 34 5 4 96
1 1 1 6
11. + + + ..=
16 26 36 945
1 1 1 6
12. + + + ..=
16 36 56 960

L2 NORM:
L2 () set of all real- valued measurable functions f on I such that 2 ().
i.e., The functions in L2 () are said to be SQUARE INTEGRABLE
The non- negative number = (, )12 L2 - norm of f

INNER PRODUCT:

The inner product (f , g) of two such functions defined by ( , ) = ()()

ORTHOGONAL SYSTEM OF FUNCTIONS

Let = {0 , 1 , 3 , . }be a collection of functions in L2 ().
(i) If ( , ) = 0 where , the collection S is said to be an orthogonal system on I
(ii) if ( , ) = 1 then the collections S is said be an orthonormal on I
EXAMPLES
1 cos sin cos 2 sin 2 1
1. = { , , , , , } or = {0 , 1 , 3 , . } Where 0 () = ,
2

2
cos sin
21 () = , 2 () = for = 1, 2 , 3 , . This system consists of

real- valued functions. T is orthonormal family in L2 [ , ].
2 + 1 1
2. = {0 , 1 , 2 , . }of Legendre functions () = ( 2 1) ,
2 2 !
2 [1
1 1 , = 0,1,2, .. Then L is an orthonormal family in L , 1]
3. = {0 , 1 , 2 , . }of Rademache functions. Here () = (2 )
0 1 , = 0,1,2, .. where the sgn functions is defined as = 1 ( > 0) ,
= 1 ( < 0) , = 0 , = 0 R is an orthonormal family in L2 [1 , 1].

RESULTS
1. An orthonormal system of complex valued functions on every interval of length 2 is
cos + sin
given by () = = , = 0,1,2, .
2 2
2
2. lim 0 () = 0

lim = 0 , lim = 0 , = 0,1,2, ..

DEFINITION
Let = {0 , 1 , 3 , . }be an orthonormal family in L2 (). If L2 [ , ] and
if = ( , ) , we call the generalized Fourier coefficients of f.

The series is called the generalized Fourier series for f ,

=0

and we write f . (1)

=0

RESULT

Equation (1)can be written as () ( cos + sin )

=1
and coefficients being given by ,
1 cos sin cos 2 sin 2
If [ , ] = [ , ] and we take = = { , , , , , }then
2

generalised Fourier seriers for f L2 f ck k , ck = (f , k ) then

k=0
1 1 cos cos sin sin
( , ) + (, ) + (, ) +
2 2
1 cos
Then generalized Fourier coefficients: ( , ) , (, ) , ie., 0 2 , 1 , 1 , ..
2
 PROPERTIES OF THE FOURIER COEFFICIENTS

1) BESSELS INEQUALITY
If L2 [ , ] has Fourier coefficients , then

a20 1
+ (a2k + b2k ) f2
2
=1

In particular the series (2 + 2 ) converges

=1

2) BESSELS INEQUALITY (USING ORTHONORMAL SYSTEM)

Let = {0 , 1 , 3 , . }be an orthonormal family in L2 [ , ].

If f L2 [a , b] and if ck = (f , k ), and f(x) then the series

=0

|ck |2 converges and satisfies the inequality |ck |2 f2

k=0

3) PARSEVALS FORMULA

1 a20
f = + (a2k + b2k )
2
2
k=1

4) PARSEVALS FORMULA (USING ORTHONORMAL SYSTEM)

The equation |ck |2 = f2 holds iff we also have lim f sk = 0

n
k=0
n

{sn } is the sequence of partial sums defined by sn (x) = ck k (x)

k=0
5) THEOREM
If L2 [ , ], then Fourier series of f converges to f in matric for L2 [ , ].
ie. , if f L2 [ , ], then lim sn f2 = 0 where
n

0
sn () = + ( cos + sin ) ,
2
=1

6) COROLLARY
If the Fourier coefficients ak , (k = 0,1,2, ) and bk (k = 1,2, )of L2 [ , ]
are all 0, then f = 0

7) COROLLARY
If f and g are two functions in L2 [ , ]that have the same Fourier coefficients,
then f = g
 8) PROPERTY

0
If f L2 [ , ] and f + (an cos nx + bn sin nx) then lim an = 0 and lim bn = 0
2 n n
n=1

9) RIESZ FISHER THEOREM (CONVERSE OF BESSELS INEQUALITY)

If { }
=0 and { }=0 are any sequences of real numbers such that

a20
+ (a2k + b2k ) < , then there exists f L2 [ , ] whose Fourier
2
k=1
Coefficients are precisely the and

10) RIESZ FISHER THEOREM (USING ORTHONORMAL SYSTEM)

Assume {0 , 1 , 3 , . }be an orthonormal on I. Let { } be any sequence of
complex number such that |ck |2 converges. Then there is a function f in L2 (I) such that
(i) (f , k ) = ck for each k 0, and

(ii) f2 = |ck |2
k=0

11) THEOREM
Let {0 , 1 , 3 , . }be an orthonormal on I, and assume that f L2 (I). Define two
sequences of functions { } and { } on I as follows
n n

sn (x) = ck k (x) , t n (x) = bk k (x) , where ck = (f , k )for k = 0,1,2, . and b0 , b1 ,

k=0 k=0
are arbitrary complex numbers. Then for each n we have sn t n
More over equality holds iff bk = ck for = 0,1,2,

NOTE:
The norm t n used as a measure of the error made in approximating f by t n

THE CONVERGENCE FOR TRIGONOMETRIC SERIES

1) Consider the trigonometric Fourier series generated by a function f which is
Lebesgue integral on I = [0 ,2 ] (say)

0
() + ( cos + sin )
2
=1
2) There exists Lebesgue integral functions whose Fourier series diverges everywhere.
3) There exist continuous functions whose Fourier series diverge on an uncountable set.
4) Fourier series of a function in L2 (I)continuous at most everywhere on I
s (x)+s1 (x)++sn1(x)
5) The sequence of A.M. { }, where n (x) = 0 is converges and its
n
f(x+ ) + f(x )
limit is at every point in [0 ,2 ] where f(x and f(x ) exist, the only
+)
2
restriction on f being that it be Lebesgue integrable on [0 ,2 ]
DEFINITION: (C, 1) SUMMABILITY OF SERIES
Let an be a series of real numbers with partial sums sn = 1 + 2 + + , we shall
say that an is (C , 1) summable to A
lim sn = A
n

RIEMANN LEBESGE LEMMA

Assume that f L(I) then , for each real , we have
lim f(t) sin(t + ) dt = 0
+
NOTE

If = 0 and = , then if f L(I),
2

lim f(t) sin(t) dt = lim f(t) cos(t) dt = 0

+ +

COROLLARY
If L[0, ] then

1
lim (t) sin (k + ) t dt = 0
+ 2

THOREM
If f L(, ) we have

1 cos t f(t) f(t)
lim f(t) dt = dt
+ t t
0
wherever the Lebesgue integral on the right exists.

DIRICHLET INTEGRALS

sin t
g(t) dt
t
0
NOTE
g(0+ ) = lim+ g(t)
t0

JORDANS THEOREM
If g is of bounded variation on [0 , ] then

2 sin t
lim g(t) dt = g(0+ )
+ t
0
NOTE
Dirichlet used to prove above theorem as g is continuous on [0 , ]

DINI THEOREM
Assume that g(0+ ) exists and suppose that for some > 0 the Lebesgue integral
g(t)g(0+ ) 2 sin t
0 dt exists. Then we have lim 0 g(t) t dt = g(0+ )
t +
DIRICHLETS KERNEL
1
sin (n + ) t
n 2 if t 2m
1 t
Dn (t) = + cos kt = 2 sin Where m an integer
2 2
k= 1 1
{ n + if t = 2m
2
THEOREM
Assume that f L[0, 2] and suppose that f is periodic with period 2. Let { }denote the
sequence of partial sums of Fourier series generated by f (say)

0
sn () = + ( cos + sin ) , = 1,2,
2
=1
Then we have the integral representation

2 f(x + t) + f(x t)
sn (x) = Dn (t) dt
2
0
RIEMANNS LOCALIZATION THEOREM
Assume that f L[0, 2] and suppose that f has period 2 then the Fourier series generated
by f will converges for a given valve of x iff for some positive < the following limit exists
1
2 f(x + t) + f(x t) sin (n + 2) t
lim dt in which case the value of this limit is the
+ 2 t
0
Sum of the Fourier series
NOTE
Sufficient conditions for convergence of a Fourier series at a particular point
JORDANS TEST
If f is of bounded variation on the compact interval [x , x + ] for some < then the
limit () exists and the Fourier series generated by f converges to ()
DINIS TEST
g(t)s(x)
If the limit () exists and if the Lebesgue integral 0 dt exists for some < ,
t
then the Fourier series generated by f converges to ()
NOTE
f(x + t) + f(x t) f(x + t) + f(x t)
g(x) = if t [0 , ], and let s(x) = g(0+ ) = lim+
2 t0 2
Here f L[0 , 2] and x [0 , 2]
CEASARO SUMMABILITY OF FOURIER SERIES
THEOREM
Assume that f L[0 , 2] and suppose that f is periodic with period2. Let sn denote the
s (x)+s1 (x)++sn1 (x)
nth partial sum of the Fourier series generated by f and let n (x) = 0 , = 1,2,.
n
Then we have the integral representation
2 n
1 f(x + t) + f(x t) sin 2
n (x) = t dt
n 2 sin 2
0 2
FEJER THEOREM
Assume that f L[0 , 2] and suppose that f is periodic with period2. Define a function is
f(x + t) + f(x t)
by following equation s(x) = lim+ whenever the limit exists. Then for
t0 2
each x for which () is defined the Fourier series generated by f is ceasaro summable and has
( , 1) sum ()
s0 (x) + s1 (x) + + sn1 (x)
i. e. , we have lim n (x) = s(x) , where n (x) =
n n
NOTE
f is continuous on [0 , 2], then the sequence {n } converges uniformly to f on [0 , 2]

THEOREM
Let f be continuous on [0 , 2] and periodic with period 2. Let {sn } denote the sequence of
partial sums of the Fourier series generated by f , (say)

() 0 + ( cos + sin ) (1) , then we have
2
(a) lim sn = f on [0 , 2]
n
2
1 2
a20
(b) |f(x)| dx = + (a2n + b2n )
2
0
(c) The Fourier series can be integrated term by term
x x
a0 x
i. e. , x we have f(t)dt = + (an cos nt + bn sin nt) dt
2
0 n= 10
the integrated series being uniformly convergent on every interval,
even if the Fourier series (1) diverges
(d) If the Fourier series in (1) converges for some x, then it converges to f(x)

WEIERSTRASS APPROXIMATION THEOREM

Let f be real valued and continuous on a compact interval [a , b]. Then for every > 0,
there is a polynomial p (which may depend on ) such that |f(x) p(x)| < for every x in [a , b]

i.e., every continuous function on a compact interval can be uniformly approximated by a

polynomial

NOTE
a
If f L[ , ] and if f is continuous at x [ , ] then 0 + (an cos nx + bn sin nx) = f(x)
2
It is known as Fejer theorem (condition is sufficient)

OTHER FORMS OF FOURIER SERIES

a0 a0
1) f(x) + (an cos nx + bn sin nx) = + (n einx + n einx )
2 2
a n i bn a n + i bn
where einx = cos nx + i sin nx, n = , n =
2 2

a0
2) If 0 = and n = n then f(x) = n einx ,
2
n=
2
1
n = f(t) eint dt if n = 0, 1, 2, .
2
0
3) If f L[0 , p] and if f has period p , we write

a0 2nx 2nx 2inx
f(x) + (an cos + bn sinA ) ~ n e ,
2
=1 =

FOURIER INTEGRAL THEOREM

Assume that f L(, ) suppose there is a point x in R and an interval
[x , x + ] about x such that either
(a) f is of bounded interval on [x , x + ] or else
(b) both limits f(x + ) and f(x ) exist and both Lebesgue integrals

f(x + t) f(x + ) f(x t) f(x )
dt and dt
t t
0 0
Exist. Then we have the formula

f(x + ) + f(x ) 1
= [ f(t) cos s(t x) dt] ds = f(x)
2
0

the integral being an improper Riemann integral

0

EXPONENTIAL FORM OF THE FOURIER INTEGRAL THEOREM

If f satisfies the hypothesis of the Fourier integral theorem , then

f(x + ) + f(x ) 1
= f(x) = lim [ f(t) eis(tx) dt] ds
2 2

DEFINITION FOR COMPLETE

If = {0 , 1 , 2 , . }be an orthonormal family in L2 [ , ], then is said to be
complete if (h , k ) = 0, k I, (orthogonal) = 0

EXAMPLE
1 cos sin cos 2 sin 2
1) The family = { , , ,
, } is complete,
2
2) The family L of Legendre function is also complete in L2 [1 , 1]
3) The family R of Rudemacher function is not complete in L2 [0 , 1]
THEOREM
If = {0 , 1 , 2 , . }be an orthonormal family in L2 [ , ], then if has only one of
the following properties it has then all.
(a) is complete
(b) The set of all polynomials (of all degree) is dense in L2 [ , ]
(c) For any f L2 [ , ], the generalized Fourier series for f converges to f in L2 [ ,
i. e. , lim sn f2 = 0
n

(d) If f L2 [a , b], then f22 = ck2 , ck = ( f , k ) parseval s equality

k=1

CONVOLUTIONS
Given two functions f and g both Lebesgue integrable on (, ), Let S denote the set of x

for which the Lebesgue integral h(x) = f(t) g(x t)dt exist . This integral defines a function
h on S called the convolution of f and g. we also write h = f g to denote this function

RESULT
f g = g f whenever the integral exists

EXAMPLS
1 1
(1) Let f(t) = and g(t) = , if 0 < t < 1 and
t 1 t
let f(t) = g(t) = 0 if t 0 or if t 1 then
(i) f has an infinite discontinuity at = 0
(ii) g has an infinite discontinuity at = 1
1
(iii) Lebesgue integral f(t) = 0 1/2 exists
1
(iv) Lebesgue integral g(t) dt = 0 (1 t)1/2 dt exists
(v) convolution integral of f and g (corresponding to x = 1)

h(x) = f(t) g(x t)dt does not exist

1 1 1
1 1 dt
h(1) = f(t) g(1 t)dt = dt = does not exist
t t t
0 0 0
THEOREM
Let R = (, ) Assume that f L(R) , g L(R), and that either f or g is bounded on R

Then the convolution integral h(x) = f(t) g(x t)dt exist for every x in R, and the function
h so defined is bounded on R

NOTE
The function f or g is continuous on R, then h is also continuous on R and h L(R)
THEOREM
Let R = (, ) Assume that f L2 (R) , g L2 (R). Then the convolution integral

h(x) = f(t) g(x t)dt exist for every x in R, and the function h is bounded on R
CONVOLUTION THEOREM FOR FOURIER TRANSFORMS OR FALTING THEOREM
The Fourier transform of a convolution f g is the product of the Fourier transforms of f and g
i. e., F(f g) = F(f). F(g)
THEOREM
Let R = (, ) Assume that f L(R) , g L(R), and that at least one of for g is
continuous and bounded on R. Let h denote the convolution h = f g. Then for every s we have

h(x) eixs dx = ( f(t) eits dt) ( g(y) eiys dy)

The integral on the left exist both as a Lebesgue integral and as an improper Riemann integral.

EXAMPLE
1
m n
If m > 0 and n > 0, x m1 (1 x)n1 dx =
(m + n)
0
The integral on the left is called the Beta function and is denote by B(m , n)

POISSON SUMMATION FORMULA

Let f be a non negative function such that the integral f(x)dx exist as an improper
Riemann integral. Assume also that f increases on ( , 0 ] and decreases on [0 , ). Then

f(m+ ) + f(m )
= f(t) e2int dt each series being absolutety convergent
2
m = m =

NOTE
If is continuous at each integer , then

f(m) = f(t) e2int dt

m = m =
RESULTS
(1) Transformation formula for the theta function is defined x > 0 by

2 1 1
(x) = en x and (x) = ( )
n =
x x
(2) Partial fraction decomposition of cot hx

e2x + 1 1 1
cot hx = 2x , x 0 also cot hx = + 2x 2 for x > 0
e 1 x x + 2 n2
n= 1
PARSEVALS THEOREM
If the Fourier series corresponding to f(x) converges uniformly to f(x) in ( , ) then

1 2
02
|()| = + (2 + 2 )
2
=1

1 2
02 1
. ., |()| = + (2 + 2 )
2 4 2
=1
COMPLEX OR EXPONENTIAL FORM OF FOURIER SERIES

1
in < x < , f(x) = cn einx where cn = f(x) einx dx
2
n =

FOURIER TRANSFORMS (OR) FOURIER INTEGRALS

FOURIER INTEGRAL THEOREM OR FOURIER INTEGRAL FORMULA

1
f(x) = f(t) cos s(t x) dt dx

0

1
OR f(x) = f(t) cos s(t x) dt dx
2

FOURIER SINE INTEGRAL

2
f(x) = sin sx f(t) sin st dt ds

0 0
FOURIER COSINE INTEGRAL

2
f(x) = cos sx f(t) cos st dt ds

0 0
COMPLEX FORM OF FOURIER INTEGRAL

1
f(x) = eisx f(t) eist dt ds
2

COMPLEX FOURIER TRANSFORM OF f(x)

1
F[f(x)] = F[f(t)] = f(t) eist dt
2

INVERSION FORMULA FOR COMPLEX FOURIER TRANSFORM OF F[f(x)]

1
() = F[f(x)] eisx ds
2

FOURIER SINE TRANSFORM OF f(x)

2
FS [f(x)] = f(t) sin st dt

0
INVERSION FORMULA FOR FOURIER SINE TRANSFORM

2
() = FS [f(x)] sin sx ds

0
FOURIER COSINE TRANSFORM OF f(x)

2
FC [f(x)] = f(t) cos st dt

0
INVERSION FORMULA FOR FOURIER COSINE TRANSFORM

2
() = FC [f(x)] cos sx ds

0

PROPERTIES OF FOURIER TRANSFORM

1) LINEAR PROPERTY
F[a f(x) + b g(x) = a F(s) + b G(s)]

2) CHANGE OF SCALE PROPERTY

1 s
F[f(ax)] = F ( )
a a

3) SHIFITING PROPERTY
(i) F[f(x a)] = eias F(s)
(ii) F[f(x + a)] = eias F(s)
(iii) F[eiax f(x)] = F(s + a)
(iv) F[eiax f(x)] = F(s a)

4) MODULATION THEOREM
1
IF F[f(x)] = F(s) then F[f(x) cos ax] = [F(s + a) + F(s a)]
2

5) FOURIER TRANSFORM OF DERIVATIVES

n (x)]
dn
(i) If F[f(x) = F(s) then F[f = F [ n (f(x))] = (is)n F(s)
dx

dn
(ii)If F[f(x) = F(s) then F[ f(x)] = ()

ds n
F(s)
(iii) If F[f(x) = F(s) then F [ () ] =
(is)

(iv) F[f(x)] = F(s)
] = F(s)
(v) F[f(x)
] = F(s)
(vi) F[f(x)
 PROPERTIES OF FOURIER SINE TRANSFORM AND COSINE TRANSFORM
(1) FS [a f(x) + b g(x) = a FS (s) + b G (s)]
FC [a f(x) + b g(x) = a FC (s) + b G (s)]

1
(2) FS [f(x) cos ax] = [FS (s + a) + FS (s a)]
2
1
FC [f(x) cos ax] = [FC (s + a) + FC (s a)]
2
1
(3) FS [f(x) sin ax] = [FC (s a) FC (s + a)]
2
1
FC [f(x) sin ax] = [FS (a + s) + FS (a s)]
2
1 s
(4) FS [f(ax)] = FS ( )
a a
1 s
FC [f(ax)] = FC ( )
a a
(5)
FS [f (x)] = sFC (s)
2
FC [f (x)] = (0) + FS (s)

d
(6) FS [x f(x)] = (FC [f(x)])
ds
d
FC [x f(x)] = (FS [f(x)])
ds
(7) If f(x) and g(x) are given functions of x and FC [f(x)] and G [g(x)] are their Fourier
cosine transforms and FS [f(x)] and G [g(x)] are their Fourier sine transforms then

(a) FC [f(x)] . GC [g(x)] ds = f(x) g(x) dx

0 0

(b) FS [f(x)] . GS [g(x)] ds = f(x) g(x) dx

0 0

CONVOLUTION THEOREM
F[f g] = F(s). G(s)

PARSAVELS IDENTITY

|f(x)|2 dx = |f(s)|2 ds


1
OR |f(x)|2 dx = |f(s)|2 ds
2

NOTE
Lebesgue integrals or improper Riemann integrals of the form

g(y) = k(x, y) f(x) dx

where g is integral transform of f and function K is kernel of the transform

SOME TRANSFORM

(1) Exponential Fourier transform eixy f(x)dx

(2) Laplace transform exy f(x)dx

(3) Mellin transform x y1 f(x)dx

(4) Fourier cosine transform cos xy f(x)dx

0

(5) Fourier sine transform sin xy f(x)dx

0
RESULTS
f (x) , < x < 0
1) Define f(x) = { 1 then f(x) is even if f1 (x) = f2 (x)
f2 (x) , 0 < x <
and f(x) is odd if f1 (x) = f2 (x)
b
sin x sin x
2) dx = = lim dx
x 2 b x
0 0

2
sin x sin2 x
3) ( ) dx = = 2 dx
x 2 x
0 0

sin x 4
4) ( ) dx =
x 3
0

dx
5) dx =
(x 2 + a2 )2 4a3
0

x 2 dx
6) 2 dx =
(x + a2 )2 4a
0

1
7) dx =
(x 2 + a2 )(x 2 + b 2 ) 2ab(a + b)
0
 n! sin[tan1 (s/a)]
8) FS [x n ax ]
e = n+1
(s 2 + a2 ) 2
n! cos[tan1 (s/a)]
9) FC [x n eax ] = n+1
(s 2 + a2 ) 2

10) FS [] = 0
1
11) FC [] = 2

n n
12) FS [x n1 ] = n sin , 0<<1
s 2
n n
13) FC [x n1 ] = n cos , 0<<1
s 2
2
14) FS [e ] =
s 2 + a2

2
15) FC [e ] =
s 2 + a2

1
16) If the Fourier transform is defined as F[f(x)] = eisx f(x)dx
2

2 /2 2 /2
Then the Fourier transform of is

2
17) =
2
0

a
18) eax cos bx dx =
a2 + b 2
0

b
19) eax sin bx dx =
a2 + b 2
0

20) n (1 n) =

sin n
1 2
21) ( ) = =
2

1
22) 22n1 n (n + ) = (2n) this is called duplication formula for gamma function
2
1 /2
m n
23) If m, n > 0 x m1 (1 x)n1 dx = 2 sin2m1 x cos 2n1 x dx = = B(m , n)
(m + n)
0 0
 PREVIOUS YEAR EXAM QUESTIONS
1) For the Fourier series of a function f L[ , ] to converge to f(x)
(A) continuity at a point x is not sufficient
(B) continuity at a point x is sufficient
(C) differentiability at a point x is not sufficient
(D) differentiability at a point x is sufficient
2) The Fourier transform of a convolution f g is equal to
(A) the sum of Fourier transforms of f and g
(B) the difference of Fourier transforms of f and g
(C) the product of the Fourier transforms of f and g
(D) the quotient of the Fourier transforms of f and g
3) Fourier series of f L[ , ] at x is (C , 1) summable to f(x) if
(A) f is continuous at x
(B) f is monotonic at x
(C) f is not necessarily continuous at x
(D) none of these
4) If f L2 [ , ] has Fourier coefficients , then

a20 1
(A) + (a2k + b2k ) f2
2 2
=1

a20 1
(B) + (a2k + b2k ) f2
2
=1

a20
(C) + (a2k + b2k ) f 2
2
=1
(D) none of these
5) If f L[ , ], then the trigonometric polynomial closest to f in the metric for L2 [ , ] is
(A) the Fourier series for f
(B) the nth partial sum of the Fourier series for f
(C) the Fourier series for f 2
(D) none the these
6) The Fourier coefficients , of a lebesque integrable functions must
(A) approach as k
(B) approach zero as k
(C) approach zero as k 0
(D) none of these
7) If f L[ , ] and L[ , ], then the Fourier series for f at x converges to
(A) f(x)
f(x+ ) + f(x )
(B)
2
f(x+ ) f(x )
(C)
2
(D) zero
Q.No 1 2 3 4 5 6 7
Ans B C A B B B B
8) f and g are absolutely integrable functions on (, ). Then the Fourier transform of the
convolution f g i. e. , F(f g) is
(A) F(f) + F(g)
(B) F(f) + F(g)
(C) F(f) . F(g)
(D) F(f) / F(g)
9) The Fourier series generated by f R(0 , 2) with period 2 is cesaro summable and has
(C , 1) sum s(x) which is equal to
f(x + t) + f(x t)
(A) lim+
t0 2
f(x + t) f(x t)
(B) lim+
t0 2
(C) lim+ f(x + t) + f(x t)
t0
(D) lim+ f(x + t) f(x t)
t0
10) If f is a continuous , complex 2 periodic functions on R[0 , 2], then the arithmetic means
of the partial sums of the Fourier series of f converge uniformly to f . it is according to
(A) Fejer
(B) Lebesgue
(C) Fischer
(D) Bessel
11)